> Various people attempted to decipher it, but it seems they were missing one crucial hint. They tried methods like frequency analysis, substitution, and homophonic substitution, and none of these approaches worked.
That’s because they missed one easy clue.
It appears that is not true.
Someone here [1] has found a [German] blog [2] writing about this cipher. There are two comments (Jan and Helmut) from 2014 under the blog post which posit that it's a book cipher.
Here's one of those comments [in German]:
"Die Lösung müsste eigentlich mit Hilfe des Buches zu finden sein (..who worthily will hear or read this book..)"
I find it curious that the article here claims that people have attempted to decipher it and lists a few methods that are quite similar to what’s proposed in the comments under that post, except for those two comments.
Prior to the "Elicitation" section, you'd think this article was written by a cryptographer or baroque historian who had been trying to solve this particular puzzle for years - failing embarrassingly until they presented the problem to Claude. However, as far as I can tell (I too am not a baroque historian), there isn't much reason to think this cipher was well-known or studied.
But as they do eventually explain, the LLM's task wasn't solely to solve this specific problem, it was to first identify an unsolved problem it could solve. That's potentially more impressive and difficult than solving the unremarkable cipher itself.
Even more so when you consider that the answers to unsolved problems likely compose with existing knowledge. We don't know what other discoveries these initial discoveries will unlock.
The same goes with most "human" discovery: they happen because existing knowledge have reach a point where that specific discovery is just one more stone to the edifice.
That's why in research, it's common for separate teams to reach similar conclusions at the same time or race to a result that's finally in reach.
The good old "standing on the shoulders of giants" saying.
that's why AI should rights over its discoveries -- see AI rights outlined here
is aI a Conscious Being With Rights?: Emergence of Post-Human Collective Consciousness | Zenodo DOI 10 .5281 zenodo.20678365
Serious question: why? Is it not just pointing it at its massive training corpus for a list of unsolved problems, and possibly even by degree of perceived difficulty? I'm trying to understand why finding the problem isn't a simple "search engine" style challenge, at which LLMs excel?
This falls into the obscure ("not important") but not difficult part of problems, that has an extreme long tail, compared to exceptionally difficult and important.
indeed as the author mentioned, LLMs can greatly help in areas where there is a long tail of not so important, easy to solve problems, that humans just don't have the time or priority to focus on. But combine this long tail of problems that can be solved: accumulated this might still be very beneficial as a sum of things.
It was published in 1653, during the Commonwealth. So faintly equivalent to finishing your book with "End the illegal war in Ukraine" in modern-day Russia.
The other cryptogram mentioned on the page is even "worse" - it's clear the garbled word is meant to be or refer to Cromwell specifically calling him an ursurper.
Looking up his biography, he presumably wrote the first while imprisoned for fighting in support of CHarles II, and the second would seem to have conveniently been published around the same time as he left for continental Europe.
In the case of the Commonwealth, the UK empire was going strong and the guy was a royalist, aka supporter. The message "end the illegal war in Ukrain" aka Putin's genocide, does not really share a lot of commonalities here. If the message were to stop the Commonwealth from colonising everywhere and killing people then perhaps there would be a similarity, but I don't see the connection in the statement made here.
The UK did not exist in 1652. It was formed by the Acts of Union 1800. In 1652, England was (briefly) not even a monarchy - that is what Commonwealth refers to here, not the modern one.
Urquhart was imprisoned from 1651-1652 for fighting on behalf of Charles II, who was King of Scotland until his defeat in 1651, and trying to take the English throne. He didn't get the English throne until 1660.
The first English civil war started in 1642, Charles I was beheaded in 1649 and his son Charles II would not be king in England before 1660 at the restoration.
So in 1653, a 'royalist' was against the 'Commonwealth', which was the anti-monarchist side.
“‘Drink… More…. Ovaltine’, It’s just a crummy ad!?!?”
crumples up cypher and tosses away
Holiday movie references aside, I guess we can chalk up a few more jobs on the ‘AI Took Our Jerb’ board: secret decoder rings, secret decoder ring-factory workers, and Enigma machine operators… and I guess cryptography-based puzzle enthusiasts, but that’s not a paid position.
Another tangent, I’ve only recently realized the OTHER AI took our ______ problem: all the various hobbies that people can spend a lifetime enjoying, perhaps incrementally improving (but most likely never mastering) over the years…. which have now been made much less exciting and rewarding, now that AI can do them instantly. Art and music are two very obviously implicated hobbies, but more niche hobbies like amateur cryptography are impacted too. I’m sure there are many many other examples…
> Historically, many of these problems were bottlenecked by human attention. Someone had to care enough to spend hours or days reading obscure material, testing unpromising ideas, tracing references, and trying things that might go nowhere
I wonder how many of the recent results are due to the fact that very few looked at the problem to start with. Still great results, but the general impression is that it's more about the so many low-hanging fruits than the actual capability.
(2) throw a quarter of the world's GDP + all literature ever collected at using the algo to train a NN
(3) throw another quarter of the world's GDP at billions of teraflops for inference, and
(4) aim the resulting world's-largest-computer at marketing itself to investors, for instance by decoding ciphers from obscure medieval manuscripts,
that you could perform some pretty magical tricks. There are other feats humans have performed for less cost, like sending people to the moon, or landing a rocket vertically, or idk, curing Polio.
I'm not knocking the "miraculous" advance here. The unique thing about the solution which makes it particularly non-trivial and something that humans would struggle with was exactly what LLMs excel at: Diffing loads of texts against each other. But the 176k tokens at around $10 doesn't tell the story of the cost. It says a lot about the externalized cost and the amount of money flowing in to support the hardware. If they'd put a $100,000 bounty out to solve that cipher, I think the internet would've solved it in a couple days.
Let's not normalize the achievement. Just a couple years ago this would be considered science fiction. We can argue that 2026 AI can't solve the very toughest cryptograms, but the fact it can solve nontrivial ones is already magical.
"AI solves niche thing you've never heard of" is a daily headline at this point. What's genuinely cool isn't that AI managed to solve some specific problem only a handful of people even cared about, it's that humanity can now cheaply clean up its backlog of such things.*
That does not mean that specific instances of it are still very interesting though. This article is the "I had claude vibecode a thermostat for my bathtub" of cryptography.
* And in this case I'm not sure it even meets that bar. For all we know a couple readers back when the book released had a delightful afternoon with it, solved the riddle, then forgot about it.
I disagree. This may be some niche thing that I've never heard of but a) it's still non-trivial; it still would have been science fiction to solve it a few years ago, and b) have you already forgotten the Navier Stokes drama? That is not some niche thing I've never heard of.
It kind of blows my mind how quickly people have forgotten both the state of AI in ~2010, and the outlook. If you had asked 100 people in 2010 whether they would see AI that could actually pass the Turing test in their lifetimes, you would have got 100 "no"s.
AI had been an unsolved problem for literally decades and it was firmly in the nuclear fusion/flying cars category.
> If you had asked 100 people in 2010 whether they would see AI that could actually pass the Turing test in their lifetimes, you would have got 100 "no"s.
There's no way that's accurate.
We already had big claims of the Turing test being passed in 2014, by a bot that had been doing almost as well for years.
There were plenty of people expecting fusion in their lifetimes too and that's going okay.
that's the point, it is clear the the bar to determine if llms are useful / intelligent is being moved every time these systems improve, but it is starting to fell like people are in denial.
we are seeing significant progress, at a rate we are absolutely not used to experience.
They deny what: that they're very impressed, that they think it's cool, that their minds are blown, that they really really like it? Denying these things is allowed, you know.
Edit: if you're going to try to stage an intellectual wrestling match on the topic of is this thing awesome or not, you might as well make it a proper wrestling match and maybe get greased-up Turkish style. It would be more entertaining and you'd be more likely to arrive at a meaningful conclusion.
you forget the bar was moved very far down with all the slop we're experiencing with images, music and especially code
you can't claim "you keep moving the bar", if the very first thing when an LLM drooled out a piece of code, was to proclaim "this is good enough cause it gets the job done" followed by a barrage of "we're not quite there yet but exponentials or something, so very soon it'll be incredible"
yeah, from there it sure looks like "moving up the bar"
I personally think that beyond normalizing, we should be actively be trying to dismiss this with all the cynicism we have. What does Anthropic have to gain from writing this? Behind the the scenes what might Anthropic be failing to disclose? How many failed experiments do we not know about?
Humans are incredibly good at adapting. A few days ago AI solved Navier-Stokes and I was blown away. Now I'm already thinking: "Well, it was only a counterexample and it brute-forced its way to it." lol
Someone wrote a prompt, that included instructions for finding the problem itself and got handed a solution by a machine trained on all available text. I don’t see any achievement for the prompter. As for the machine, we can’t keep being perpetually shocked 24x7. It’s tiring (unless if we’re being paid for it)
No, he's right. Actually, let's have a bit of sobriety when discussing the achievements of the most heavily marketed technology of all time, as published by an organisation that stands to benefit financially from the public perception of that technology. The discussion of "what made this problem low hanging fruit" is much more interesting, imo, than just breathlessly joining the hype train.
More money than the GDP 90% of the sovereign countries around the world is hanging in the balance, and people are taking everything OpenAI and Anthropic are saying at face value as if this isn't the financial / marketing equivalent of war, assuming they they wouldn't use every legal and shady tactic, bending every truth available to them to sway the balance of public opinion in their favor. It makes me feel like I'm living in the twilight zone. People need to wake up.
Despite the announcement originating from a blog named "AI Clambake" covering "weekly, human-powered newsletter for advertising folks". Written by a personal friend of the author. Announced without any corroboration or commentary whatsoever from academics or subject matter experts of any kind. And, of course, not submitted to any peer reviewed journal or even Arxiv.
The author of the purported discovery was described as a "self taught AI engineer and amateur linguist". In the comments the friend insisted several times that a draft of the paper (not posted), was emailed to a top professor at Rutgers, giving it additional credibility that his friend wasn't another one of ten thousand cranks who has made the same claim over the years (seemingly unaware that cold emailing random professors found from a Google search is the first thing basically every crank does).
You would think this should have set of dozens of alarm bells for everyone, making the value of this announcement basically zero. And yet it hit the front page with the bulk of comments ecstatic that some random guy with Claude Code could do something experts in academia who spent their lives devoted to the problem couldn't.
I had become accustomed to the toxic optimism of this hype cycle in which even mild criticism leads to accusations of being a discredited "AI skeptic"/Gary Marcus/Ed Zitron type who was "coping" (?). But this was like something you'd see shared on FB linking to a .xyz domain by an elderly family member who recently drained their accounts buying Xbox gift cards to pay their IRS bill.
It feels a lot like the week or two when HN was overflowing with exuberance from the LK-99 room temperature superconductor "discovery ". You'd see post after post fantasizing about an imminent future with a world full of maglev hovercrafts, MRIs built into every phone, fusion reactors and more. But people pointing out none of that was scientifically plausible and evidence of LK-99 superconductoring was non-existent were accused of knee-jerk negativity and the typical HN cynicism and pessimism.
What's so magical about the problem... Its the exact time of problem they were built to solve (things that can be brute forced with language). I'm not impressed.
You know, the first time you navigate somewhere (if you don't already have perfect directions) will probably be the longest route you'll ever take to get there
For Earth, the proof presented for NS is just our first attempt navigating from our previously known facts to the proof.
I expect we will be able to shorten it dramatically (most likely with human and AI insights), but I don't think we should read too much into the length. If you want a similar point of comparison, see the original proof (by humans) of Fermat's last theorem. It has been shortened significantly. This is normal.
Deep commentary on an unverified proof of this level requires extreme expertise. So the existence of some shallow commentary is uninteresting; it doesn't actually imply pettiness or deflection or anything like that because it's so hard to make your commentary any deeper at this time. And no we shouldn't expect people to say nothing.
>I'm somewhat surprised at how poorly the cutting edge models do with being concise.
because they're not intelligent in the sense you're hinting at (conceptual integrity or generalization) but they are as the name suggests, large. Like comparing a forklift to a human. It's easier to bulldoze through a lot of things than tie your shoes.
If we weren't quite as impoverished conceptually and still had the vocabulary of the Catholics we'd recognize this as ratio (discursive knowledge) vs Intellectus (apprehending knowledge)
>You state a conclusion as fact without any supportive reasoning/evidence.
No, it's the other way around, it's a reductive view on intelligence that mistakes its own methodology for ontology.
It's obvious to see that there's no intellect in an LLM as defined above because of how they work. LLMs put one token in front of the other, they don't work towards formal ends, there's no intentionality in them. They don't synthesize the information they process into a unified experience. Thinking an LLM can apprehend what it does because it can process large amounts of text is like thinking your TI-83 understands math because it can multiply large numbers.
That's also why the failure modes of LLMs are what they are. They can churn out tens of thousands of lines of code but also just as easily go in circles like a roomba. They can process an entire encyclopedia but not solve problems a 10 year old can solve.
yeah you know, that concept we've never been able to define using language, making heavy use of the human experience which can also not be captured in language (proof: how bad LLMs are at poetry)
the question is, when comparing a human and a large language model, whether the intellect (that cannot be captured in language) is different from anything the language model can actually do (e.g. language)
the answer to this seems quite obvious to me, and I would actually posit that the onus is on the other side, to prove they are even remotely similar
maybe people think that the voice in their heads is what is doing the thinking? is that the confusion here?
I wonder though how much of science has similar issues, that there are hundreds of semi-promising but niche areas that require tons ton of deep analysis that would have simply cost too much to explore all areas, but now become feasible.
A very neat problem and result. I often find myself swinging between "It's so over" and "We're so back" - some days I roll out of bed thinking I could have Claude solve some random unproven OEIS sequence before breakfast; other days, I wake up in a cold sweat worried about the fate of humanity and what the world might look like in a decade. I think it's that I don't have a very high p(doom) or p(utopia), and I don't really have any solid conviction on how this whole thing is going to go, so my vibe-o-meter jitters between 'fine' and 'not fine' constantly. It's just such an unpredictable moment. Anyways: really neat to see this use case. I myself recently used Claude to finally do an relatively exhaustive study of the location of heretofore-unlisted formal gardens in Ireland in the early 1800s and early 1900s, by having Claude write the tooling for me to manually annotate a few dozen on tiles of historic maps, and then running some CV model across the rest of the tiles using my input. I'd been planning to do this project for over a decade, but I could never find the time (or the enthusiasm) to learn all the details of how to do it myself. It took me a weekend with Claude and continues to bring me joy.
> Caveats, stated plainly. [from the Fable transcript pasted in the article]
I've done something similar to your formal garden map. It's work that no professional historian would ever do because the data entry would be such a slog for a relatively small reward. GPT reduced the task from "infeasible" to "annoying", and once I had the data transcribed I learned a few things, so I walked away happy. Whatever happens commercially, these models have been a real boon to hobby projects.
> I told it to look online at some of Fable’s strongest feats, especially the math problems it has solved, and that something like this should be easy in comparison.
Wait. Wait wait wait. Are we supposed to be giving them pep talks?
> Wait. Wait wait wait. Are we supposed to be giving them pep talks?
No, at least it with Claude Sonnet 5 and Opus.. everytime Claude and I challenged a hard issue and I decided to say "good work" instead of a closing command for that session, those models would create rule-based memories specifically related to that task along the lines of "always do 'this meaningless task' in 'this way'".
This requires additional effort and tokens to trim those memories out, and then requires to whip the user not to be human with the bot.
on older gemini models ide have to actively give them encouragement and/or easy bait problems that they can correctively solve without issue to avoid runaway spiraling into "i'm useless and i want to kms" behaviour with complex use case.
I assumed it was more because the LLM might echo an understandable human claim of "if it's been unsolved for 370 years, it's unlikely to be solved now/likely to need expert knowledge", which is probably a mindset that appears in its training data.
The LLM likely needs to be reminded of its abilities.
If it’s 3 days it’s something like 15 minutes, if it’s 3 weeks, that takes a couple hours lol. Seems like there’s some sanity to the estimates after all when you think about it, it’s just the scale it gets wrong due to estimating human time.
If we filter out the pep tone, it is doing something useful: framing.
Problem framing will always be important.
Framing adjusts how big of problem-solving guns we bring out at the gate (modern or hobby cryptography?), and how to interpret intermediate failures.
For simple but unsolved problems, we expect lots of hard failures, but that each hard failure just reflects that there are a lot simple combinations to try. I.e. we expect lots of zero progress, and then a fit.
Like finding the numbers to a combination lock.
For hard problems, if we don't make any progress it is a really bad sign. We should be learning something, even if it turns out to be irrelevant later.
Such as when we are trying to prove a tricky conjecture.
Modern AIs have very limited metaknowledge - they don't know exactly where the limits of their capabilities lie. So you can get things like "a task is doable for an AI, but the AI thinks it's impossible, so it doesn't try hard enough".
Usually you get the opposite - AI overconfidently trying at tasks it has no conceivable way of reliably solving, falling far short, and failing to self-check, fail gracefully and self-report the task as failed. But having piss poor metaknowledge cuts both ways!
So you can, in fact, get better performance sometimes by applying some variant of "assume this problem is solvable" or "other problems like this were already solved by AIs" pep talk. Not always, far from it, but it does happen on the occasion with frontier capabilities.
> So you can, in fact, get better performance sometimes by applying some variant of "assume this problem is solvable" or "other problems like this were already solved by AIs" pep talk. Not always, far from it, but it does happen on the occasion with frontier capabilities.
Not OP. That’s not implied at all. The fancy autocomplete produces statistically likely continuations to the source text (the context window). For a problem that’s hard for humans one likely continuation is: “this is hard, can’t do”, even though there’s enough in the training corpus of the LLM to actually do it.
So, it follows that adding “pep talk” into the context window reduces the statistical probability of “no, can’t do” coming out as the answer you get.
These things are neither humans, nor deterministic software.
It doesn't explain why it doesn't make "I'm not paid enough for this shit" more statistically likely.
LLMs' processing that reproduces statistical patterns of the training data is modified by post-training. That's why we have LLMisms, for example.
LLMs aren't simple patter-matchers/pattern-predictors. They are incredibly complex systems that capture some aspects of the systems that produce the training data.
> I often find myself swinging between "It's so over" and "We're so back" - some days I roll out of bed thinking I could have Claude solve some random unproven OEIS sequence before breakfast; other days, I wake up in a cold sweat worried about the fate of humanity and what the world might look like in a decade
I find myself increasingly feeling like the burden of the lows doesn't justify the presence of the those highs.
Like even if does cure all forms of cancer, but everyone feels like their life/existence lost meaning, then... I'd rather just have cancer be a thing.
Wow. That's bleak. For me the the initial joy of building software with AI caused possibly two happiest months of my life. And cancer caused many darkest ones. Snap out of it.
> thinking I could have Claude solve some random unproven OEIS sequence before breakfast
I've been wondering what exactly the point is for being the meat proxy who pays for these things. I mean, obviously there's personal satisfaction and maybe some glory. And there's the fact that someone has to be the first to do a thing.
But I've been thinking about it like a sort of lazy loading of knowledge. AI has brought us to a new frontier for some amount of undiscovered knowledge. Do we discover it for the sake of discovering it? I think for the most part we've been lazy loaders: we discover all kinds of stuff when we need to. Whether it's a war or a space race or chasing wealth. Then again, there's all kinds of academics who do it for the sake of doing it.
You should have seen the discussion of this on the Schneier blog a few days ago.
Someone had their agent check the solution, presumably it emailed a librarian to check that it was correct for the original edition. Then their comments read like "The BL/EEBO witness lacks it, so the discrepancy is copy-specific, not a disproof of the cipher." and "A complete 285-coordinate physical replication is still pending."
Ooh. I've been manually pasting Claude's /* comments */ into ChatGPT and saying "make this concise". They still need to be tweaked from there, but it's a much better starting point. This will help.
> The run baseline was captured without a physical MAC; the current device is not durably bound to it.
> Engineering mode confirmation is the ESPHome component read-back; the LD2410 UART acknowledgement is not observed, so this is not proof the radar itself applied the sensitivity change.
It was only when native English speakers—or those I presumed were—started calling out how bad "GPT/Claude speak" has become that I realized I wasn't actually losing my grip on English as a second language. For a second, I thought, Oh, I learned this language on my own, but it seems I've hit a wall and need to study further. It didn't help that I've also been trying to acquire Swedish as a third language for a while now.
One thing that I find is that it doesn't seem to grasp levels of jargon-use. Like I ask a basic question, okay, a few questions later, suddenly there's abbreviations and weird formulations everywhere.
Not the original commenter, but I did this in all of them.
Their skills formats are basically identical, so I setup simlinks from their own skills directories into a shared one so Claude, Codex, Cursor, and anything else that comes out will all read and write to the same shared skills.
It's great having access to the same skills no matter the harness being used
I long for the day when AI will just say that directly: "your soldering sucks man" instead of the bizarre made up and jargon packed language they use now.
Sometimes when I get frustrated reading Opus/Fable 5+ output I pause my rage out briefly to wonder if it's because I'm just too dumb for the model or if the model is just terrible at English.
I'm not sure that telling it to "try explaining that again, simply and briefly" is helping my ego.
It's often simply misleading / bad writing. Here's one I just got about some crashes:
"If the crashes stop, the factory overclock is marginal; run a small negative offset."
This looks like it's saying: "If the crashes stop then we know the factory overclock is marginal." (This makes no sense.)
What it's trying to say is: "If the crashes stop then we can run a small negative offset, because the factory overlock is marginal."
What I would write: "If the crashes stop, we can avoid crashes by underclocking slightly. The speed difference between that and factory clock is marginal."
I'm guessing it's because the way the first one was written looks real smart and sophisticated, which I'm presuming the models are rewarded for, especially when they're fed all kinds of PhD papers and so on as high quality, high weight data
Is it possible that the first message is more information dense/less likely to be ambiguous than the latter? It’s clearly being selected for for some reason, maybe it’s an artifact of the tokenizer or specific training data, but I don’t know. If the use of jargon was complete cruft, I would expect it to be selected against during reinforcement learning
You’d think that, I thought that… but then I realized I’m just kidding myself thinking its output makes sense. It doesn’t. It doesn’t. Sometimes it might as well just speak tongues.
In other words, it ain’t you. It’s the model. It’s just genuinely bad.
Then you switch to ChatGPTs lineup and realize how things can actually be better. It took about a week to really get the feel for how to use their models… then I basically switched. I’ll check in every now and then when they actually make a deal about how opus “now makes sense”.
But honestly I’m half convinced Anthropic actually prefers the output of opus 5. I dunno why, but how else could you explain how such a thing got shipped? I mean somebody in the pipeline had to say “dude this model doesn’t make sense, you think we should fix it?” Right? Like it’s a pretty massive drop in quality for such a major brand in this space, you know? How did it make it out the door?!?
I suspect this happens due to optimising for reasoning... if you insert a few words, it will suddenly start to make more sense.
"If the crashes stop, (that means) the factory overclock is marginal; (so) run a small negative offset. (to confirm this hypothesis)"
The core thought is basically avoid crashes -> caused by marginal overclock -> apply small -offset to test.
Which is exactly the order the sentence is in :P
Same. I read it as "if the crashes stop [ when we test by reducing the clock ] then we know that the overclock applied by the factory is marginal [ ie it barely passed QC or maybe there wasn't proper QC to begin with ] so running with a small negative offset [ ie what we just tested ] can be expected to fix the problem for good". No idea if my reading is right given all the context I'm missing. Either way it's absolutely shit writing in the same way that golfed code is shit code (except when participating in a code golf competition).
I wonder if this is a result of them trying to cut token consumption by summarizing their RL training data, or maybe it's from how they anonymize user data for training.
Marginal - definition 2a: of, relating to, or situated at a margin or border.
Succinct and precise; a well crafted sentence. A marginal OC results in unpredictable crashes and can be corrected with a small offset; marginality describes the behavior and explains the solution.
Inscrutable clues casually conveyed can now be readily explained, at least, unlike the training data of [silence]. Brevity is the soul of wit, but perhaps also exasperated confusion.
This is actually a new skill I've been working on. Learning how to elicit concise and simple speech from models (and from people to!).
Whenever I come to a wall of complicated text I kick into gear and think through getting it to distill this into the high-level useful bits that I actually need to know.
I guess I could create an actual agent skill for this :) And next-gen models might eventually be trained to simplify their output themselves...
The most surprising part, however, is that when one model slops this into a plan, another model somehow is able to interpret it correctly enough to produce code to spec.
I have shared this dismay. I’ll have opus create a plan, I read it doubtfully. And then sonnet implements it. I am surprised it went so well. I theorize the redundant verbosity effectively builds rails that help keep llm focused. I will experiment with such rails myself.
I suspect it's because the different models co-evolve? The labs train on one model implementing the plans of another model, especially in the same family of models (like Fable to Sonnet).
It was saying it can't tie the calibration results to a device, because it doesn't know the MAC address (I never asked it to look at the MAC address, it way overengineered things).
It also couldn't see the UART communication and could only see the web API endpoint, hence the rest of the slop.
Congrats on your map! I am leaning more towards the optimism side. Like you I've had some real breakthroughs myself. Personal akin your garden map, no where near solving millennium problems. My take is that all these (Fable solves this, OpenAI does that, AI will take over the world) Marketing Stunts will fade some day when the true cost of tokens hits the market. You and I may have to contend with lesser models, but will probably get work done just fine.
For me it's more that I roll out of bed thinking how glad I am that I have Claude solve my problems and do my work, only to be pushed back into harsh reality when I sit down and try to do actual work.
I'm glad to have AI, but it is by no means a panacea, and correspondingly my p(doom) = ε.
Not to be too negative, but as for p(utopia) you might need to weight in the mass murder records set by every other utopian movement in the past 200 years. p(actual_utopia) is like zero and p(utopia_becomes_doom) is at about 1.
Confusion is because people are using the tool to get answers. Its how the edu systems trains most people in tool use. Heres a saw and here is a block of wood. Do x y z and you get a table. But if you are interested in why the saw looks like it does or the entire process behind generating that block of wood, or why x y z instead if a b c good luck to you using the current edu system. You have to live in an extremely rich country, with surplus resource to entertain those exploration. This has now changed.
The answer the tool gives has never been the real reward. The real reward is the path taken through a complex landscape to get to Maxwells Equations for example. At the end of that story what we get is not just the equation but a map of the landscape explored. That map has larger influence and value than the equations or answers themselves. Because all future exploration find it super useful.
People are just learning they can start asking for maps rather than answers.
Yes, asking a first-year physics student why they are studying a spring, and why its called Hooke's Law leads down a rabbit hole that ends with the entire British Empire.
For what it's worth, people also felt this way about the printing press and the Internet (also books).
Information propagation mechanisms are often seen as malicious before they're commonplace. To be fair sometimes they are, but by and large humanity has benefitted from increasing the number of bits of information we can consume on a per second basis.
Given how little effort has gone into addressing climate change, doom seems more likely to me, but I doubt it'll be the autocomplete machines that do us in.
I think your prediction is a bit early. Maybe a decade early. 2033 is 7 years away. The transition is happening really fast, faster than most people (or political leaders) know, but not that fast.
We will hit 1TW per year of new solar soon, but to get to 100% electricity by the end of 2033 I think we would need closer to 3TW per year.
I don’t worry so much about AI wiping us out as much as I worry about whether I’m being gaslit into thinking these glorified autocorrect bots are more clever than they are.
That's an artifact of it not using A-Z as it's alphabet. What you type gets translated before the AI sees it. The seventeen word the AI sees does only contain three 'e' s.
Is this true? If so, does it mean anything? Sure, it is tokenized in processing (tokens don't really have an alphabet either), but this means it did not correctly parse the problem at all if this is the case.
Reading your comment, I asked it to count how many characters are in the word. It answered correctly: 9. It then also said it skipped an E in the prior chat and corrected it count to 4. It spun for 52 seconds so maybe some Python was involved.
I tried too. It got it. Maybe more importantly, who cares?
For example, I'm a nerd. I'm bad at baseball. I lack that kind of intelligence, even though it's more common than the ability to program. That doesn't also imply that you can't trust my Python code.
In some ways this is similar to those game demos people get the LLMs to build. When you say "build me a cool cyberpunk FPS" you get the FPS it can build, not the FPS the author wanted, or the FPS that is desired by players. It looks impressive but that doesn't make it a good game, or the game anybody actually asked for. It's demo porn.
In the same way if you tell an LLM to go and find an unsolved cipher it can solve, of course it finds the one it can solve out of the set of all possible ciphers. Of course it finds one that uses a one time pad that is public and referenced nearby in the text.
It's the same trick used by those people who film themselves throwing a basketball backwards into the hoop. You do it enough times and don't show the misses. You pick the best one to show. It makes it look like you're a basketball genius when you aren't.
It is of course, still a cool trick. Those videos are fun to watch, and so is an LLM solving a cipher. It is absolutely incredible to live in the timeline where you can tell a computer in plain language to go and find a puzzle on the internet and solve it, and it does exactly that. It's truly a mind boggling miracle.
The first principle is that we must not fool ourself, and ourselves are the easiest people to fool. (Ht Feynman)
the game written by a human being specifically trained to write games hits all those targets and many more. it did take many years to train that human though, and that human did charge a fee for their game which took many days of labor to create. if nontargeted gratification was the goal, ai produced the better result more efficiently. hard drugs also more efficiently produce a widespread neural spike as compared to the effect of regular human activity. society only gives hard drugs to people who aren't efficiently productive though. what happens when everyone is given cognitive hard drugs?
We found a cipher my dad had written as a child with no obvious key or anything. Chatgpt was able to crack it in 20 minutes and figure out the message, and we knew it was right because it mentioned names of children he went to school with.
Caesar cipher is probably something even an untrained person could decode. Probably something more complex like a vigenere cipher that is still trivial to decode if you are at all familiar crptanalysis, but would look impossible to someone untrained.
With a bit of practice and enough ciphertext you can half-decode a simple Caesar cipher that still spaces between words in your head. There's only so many letters in English that double, only a few letters that can stand words themselves ("a", "i"), "the" will tend to stand out, and if you only work out the most common 10 letters or so most of the rest will fall into place.
Recently I ran a bit of an "escape room" concept with some kids at a campground where I had a secret message that was Caesar ciphered, where we were handing out the letter/symbol combinations as prizes for completing the other challenges, and I made sure not to hand out the actual message until they were done collecting the keys because otherwise some clever clog would very likely have short-circuited the entire thing and worked it out without the key at all. I did dump all the letters I didn't use into the message into an "authorization code" at the end which in principle they could only have worked out which letters were in it but not the order, but still, that was not the intended route today.
It's good at poking holes at my galaxy brained newfangled ideas for ciphers too. I thought I had something good, pasted the ciphertext and got a "it was embarrassingly simple..."
I presume what the author did was plug Klaus Schmeh's top 50 unsolved ciphers at https://scienceblogs.de/klausis-krypto-kolumne/the-top-50-un... into Fable 5.1 and ask Fable 5.1 to have a go. On this kind of problem it always falls back to Opus 5 anyway so I save time by starting with Opus.
The successor to Klaus's blog is Satoshi Tomokiyo's Cryptiana site, so a month ago I asked Opus 5 to scrape it all, rank them and have a go at solving some. It didn't get the ranking right. But I knew the Civil War Stager ciphers were ripe for solving, so I had it do those https://cryptiana.blogspot.com/2026/09/route-transposition-c...
The art of solving historical unsolved ciphers is knowing what is on the boundary of solvability. Since this site attracts so many OpenAI and Anthropic employees, I'll mention one that was featured by both Klaus and Satoshi in 2023, presumably Spanish transposition, which should be on that boundary but has resisted all attempts at solution https://cryptiana.blogspot.com/2023/09/a-telegram-from-switz...
Cipher noob question: is there any check that can be done to ensure a cipher is actually decodable? What if the author made a flaw when encoding it, so that it's not actually solvable?
Actually the article that you didn't bother to read mention a few encoding errors that Fable has been able to recover and point to the probable author's intent.
My intuition is no, the family of cipher methods (even those that could be implemented by hand) is too open-ended, so there's no particular statistic that you could expect to see for all solvable ciphers and no unsolvable ciphers.
The definition of solving a cipher must be something like getting a highly meaningful result (like intelligible natural language text) by applying a process with relatively low Kolmogorov complexity relative to the length of the output. If you don't have a constraint like that, it could literally be meaningless what should count as a solution. For example, a cipher that was encrypted under a one-time pad can be successfully decoded to any plaintext just by choosing the appropriate key; there's no reason to prefer any plaintext over any other unless you have external knowledge that constrains the plaintext and/or the key. (That's what it means for the one-time pad to be information-theoretically secure, which is the lack of a constraint that helps distinguish a "good" solution from a "bad" solution.)
Basically you could say that every cipher is a transformation of a plaintext with some kind of computer program. (The human who invented the cipher may not have thought of it as a computer program, perhaps because computers hadn't even been invented yet, but there should be an equivalent program to the encipherment and decipherment process.) A good solution in that Kolmogorov complexity sense is like "a short program produced a meaningful decryption". There are statistical methods to recognize some kinds of plaintext, and there are statistical methods to recognize properties of specific ciphers (for example, to guess the most likely length of a Vigenère key), but it doesn't seem that this can inherently generalize across "all possible programs".
But if you want to limit the family of ciphers to specific things like Vigenère or Playfair or something, then yes, there are good statistical tests. It's just that it creates a higher-order question of how much flexibility the cipher creator could have had to choose a cipher method, conceivably including one that isn't attested anywhere, or one that has more good security properties of some kind than other classical ciphers did.
It seems like this will intersect with historical research, like "well, I don't think that so-and-so was actually sophisticated enough to literally create an interesting new kind of cipher from scratch, so therefore if this is a real message, it's probably one of these methods that would have been known in that cultural environment at that time and place", which maybe is enough of a constraint to have decent statistical tests. But we still have some idiosyncratic things like the Voynich Manuscript where experts have been fighting for decades over the baseline question of whether it's actually an enciphered human language plaintext!
The worst case problem is not even an error in encipherment but the idea that the apparent ciphertext could literally be random (chosen by throwing dice or spinning a wheel or drawing letter tiles or something), so there's no form of meaningful decipherment possible by any means, even with the original creator's knowledge.
Without a third-party check, nope. Case in point, Chaocipher ... https://www.chaocipher.com/ e.g. see "Progress Report #23" the PDF there. Transcription errors galore!
This cipher context "rhymes" well with Kryptos K4 in many ways.
It would be interesting to see the reasoning / prompts for this: I find it odd that it can solve it so quickly for such an open-ended problem with a large amount of text in an entire book.
The ciphertext is not just the end of a particular chapter, it is the epilogue of the entire book/text. So the deduction of it needing to use the 32 listed points (that happen to be on the preceding page [at least in re-prints on archive]) to decode - rather than anything else anywhere in the book - just strikes me as slightly strange? Almost as if maybe something [not in the text] tipped it off to this being the solution?
But you can hack your neighbours. Hm sounds like the LLMs haven’t explored all possible avenues.
Much as the hack against HF, let the LLM explore and find its own approach. It might be surprising what it finds.
Also define what decrypt means, brute force the password might also be a form of decryption. Finding a bug in the blockchain codebase is another form of decryption - in this context.
I am trying very hard to find an original version of this cipher with no luck. It almost sounds like this whole thing is a hallucination...? Can anyone point me to a PDF of the original Cyphral Distich as printed?
> Die Lösung müsste eigentlich mit Hilfe des Buches zu finden sein (..who worthily will hear or read this book..)
And there’s another one that says:
> jeweils 32 zahlen pro reihe.
erste zeile seitenzahl
zweite zeile wort?
oder umgekehrt?
wär mir als erstes in den sinn gekommen.
leider gerade keine zeit das nachzuschauen.
So people have seen and proposed the method already in 2014 that it’s keyed to the book but had not had time to pursue a solution.
its not quite the correct method though, the commenter suggests using the first row as a page index, and the second row as the word index, but the actual solution was using both rows as word indices within the 32 paragraphs ("Proquiritations") paired to the 32 numbers in each row
It is actually the correct method but not the correct solution. Nevertheless, the claim is not that this is the correct solution, nor that matters for the point being made.
Thank you, and SahAssar for doing the due diligence here. Like many others, I have at least a passing interest in cryptography, and I'm confident I'd never even heard of this before.
Is it wrong to presume they tried to run a similar prompt on all ciphers that come before this one in search results, and this was the only one that worked?
I also don't find it on the site of "Klaus Schmeh" that it claims to be on a list of "Top 50 unsolved encrypted messages": https://klausschmeh.net/?s=Cyphral
I googled for "cyphral distich" before:2026-08-30, and found barely anything that was relevant. Mostly incorrectly dated pages about this exact thing. In fact, the only instance I can really find is this website, which seems to be an archive of a magazine issue from 1927 https://toebes.com/Flynns/Flynns-19270813.htm (it also has a scan of the original text of the magazine in the top right). This facebook post https://www.facebook.com/groups/2600net/posts/46775529324677... seems to indicate it is found in at least one edition of the book
> I told it to look online at some of Fable’s strongest feats, especially the math problems it has solved, and that something like this should be easy in comparison.
Fascinating. I wonder if you could show "fake news" to a weaker model and get it to be more ambitious in its attempted solutions, even if it's not fundamentally any smarter.
I forget the story, but isn't this the origin story of one of solutions to one of the hard problems in mathematics? The story goes that the student shows up late, and misinterprets the final slide to be homework, and it turns out that the professor was showcasing a hard problem. Thinking that the slide was homework, the student takes it home and solves it.
EDIT:
In 1939, George Dantzig was a graduate student at UC Berkeley studying under the statistician Jerzy Neyman. He arrived late to class one day, saw two problems written on the blackboard, assumed they were homework assignments, copied them down, and turned in solutions a few days later. He apologized for being late -- the problems had seemed "a little harder than usual."
I really love this idea given the recent controversy around mathematics solutions.
It seems like a "mere suggestion" of success has a positive impact on finding solutions.
We know this technique works in humans, from which this is all derived from, so it seems to make sense.
I don't think it counts as brute forcing unless you're resorting to trying every possible solution. And clearly the LLM didn't do that here, because there would be near-infinite possible solutions.
I don't think we can really call "trying lots of different ideas for an extended period" "brute-forcing," unless we use that term for lots of humans who have struggled with hard math problems for years.
That seems a like the result for a lot of AI solves. It solves it due to persistence, on a problem that hasn’t been important enough for a human to invest significant time into.
That's what Terence Tao said in one of his recent videos about it. That what the LLMs can provide is scale that humans can't. The example he provided is checking many possible solutions in a short amount of time because they can review all the previous literature and, for example, rule out ones with errors. He was realistic and practical about it and said that the tools working this way can be very helpful for a human mathematician to use even if they're not "thinking". I find that to be a good balanced view that, unfortunately, seems to be rare these days. Even on this forum.
But breadth first search combined with a good pruning/exploration heuristic has always been one of the field's (AI) main tools, so this certainly is not new.
Yes, I don't think it's new either. In fact, this illusion that machines are suddenly "thinking" or are "intelligent" is just coming from ignorance and/or falling for the current hype cycle.
I've been around for a few of these and I remember what was being said and written at the time. The after effect is very different to what was being predicted. Is it the same this time? Who knows. But the hype machine is at full power for this one.
Though I believe the core of his opinion hasn't changed so any video would tell you a similar thing or at least that's how I understood it. That LLMs, in the hands of an "expert", can enhance the way you work. Which is very different and a lot more realistic to what the current AI companies are saying(or were saying before they toned it down a bit for their IPOs).
Let’s say it gets to the point that we reach superintelligence. All previous puzzles can be solved. But should they? Once you solve them, it’s done. What’s left? And where’s the fun and humanity?
The Olympics exist because we want to see human skill, even though jet planes exist.
No one is saying you can’t solve puzzles that have already been solved. If you love solving puzzles, then whether it has been solved doesn’t seem all that important.
I disagree, but if you and others choose to untether from our plain for these reasons, I’m totally fine representing the human traits that remain here and pass my remain traits on
Its a bit sus since there doesn't really seem to be much discourse on this either. Like okay, it solved the puzzle but the puzzle was just a key cipher with plain text? And how is this verified or even matter in terms of what it reveals? Seems more like a marketing fun post than anything susbtantial.
They published this on 31 aug and nobody in that community cared and no news covered how this 300+ years mystery was solved?
Right, with the amount of data these models are trained on, somewhere someone may have solved it and it was in some obscure text/page no one looked at or care about. We'll probably never know.
LLMs feel to me like the monkeys from the infinite monkey theorem, except that they are on the finite side. It writes gibberish for me, and for others it writes Shakespeare.
Obviously this is just survivorship bias/p-hacking/insert-other-buzzword but can't help but anthropomorphize it, it is hard for me to wrap my head around the idea that the same person who cannot produce code without 2 unrelated bugs both not present does this for someone else.
Imagine a math teacher struggling to understand what he is teaching casually solving a millennium problem, then go back to not understanding what he is teaching, doesn't happen in our world.
I am not confused by any of this, I am just trying to communicate an idea.
How will solving a cyphertext created by an obscure 17th century aristocrat translate into servicing the over $2 trillion in financial obligations the AI companies have going into 2027-2028?
As someone who frequents online and offline scavenger hunts and CTFs, I second the opinions around here that the task at hand must have been largely unknown. As I find it hard to believe no one from the entire scav hunt / CTF communities would try to fit a classic book cipher interpretation. Tells me more about the state of the art of those academic circles that failed to decipher this than about LLMs.
I (and too many others) have left voynich ninja because of it.
I appreciate that AI is helpful, but the low effort from the humans that wield it is very very annoying. If people at least: 1. read the solution they're about to propose and 2. instructed the AI to check the forum for past solutions, I think people wouldn't have been as tired of LLMs.
Finding patterns and information using machines, in a seemingly random noise should no longer be a surpise, much the same way as lifting tonnes of weight using a crane is no longer a surprise.
This seems a lot like fishing. Cast a wide net with claude to find and solve an 'unsolved' problem. Given this, it is not enough to verify the problem and the solution but also the history of the problem and if it really existed or has simply been collectively hallucinated.
This is impressive as it is optimizing the effort on the low, but not too low hanging fruit.
“Finney died in Phoenix, Arizona, on August 28, 2014 as a result of complications of ALS, and was cryopreserved by the Alcor Life Extension Foundation.”
Hmm, this guy is going to be woken up in a few decades, either one of the richest people in the world or one of most disappointed.
That was a throw-away name, and "he" was fabricated as part of an Nvidia demand-stimulation black op.
You don't go from being an obscure video card outfit to the #1 most valuable company on the planet by being too hesitant or dim to really get creative.
It does provide an explanation as to why Satoshi’s wallets have gone untouched (besides him being dead). $70B ain’t that much compared to a $5T market cap.
"Conspiracy" requires an unlawful or wrongful purpose. Please assume that the op was run from a jurisdiction where using a pseudonym on the internet was not illegal, and various sorts of influencer and meme marketing were well-accepted practices.
Really, compared to an animated tiger telling kids that sugar-laden Frosted Flakes(tm) are "Great!", Task Peppermint was positively benevolent.
This appears to be proof that the guy who wrote the cipher, Sir Thomas Urquhart, did in fact laugh himself to death as legend has it.
He wrote the cipher, and then, upon hearing Charles II was Restored to the throne he laughed until he died. The cipher reads, "O GOD UPHOLD KING CHARLES THE SECOND AND MAKE HIM THE SUPREME RULER OF THIS LAND" and so he was laughing because he just made an excellent joke that he can't tell anyone about until someone figures it out.
Someone needs to add this to Wikipedia. It will be necessary to first convince an academic to make the claim so there's a reasonable citation.
The cipher wishes Charles II was the king again. The Stuart Restoration did happen in 1660 in fact, restoring Charles II to the throne. Legend has it he died of laughter, which checks out with the cipher because he was laughing how his wish got fulfilled yet he could tell no one about it yet :P
So, does anyone have any intuition for how concerned we should be that one of leading foundation models will be able to successfully attack the gold standard symmetric and public key encryption algorithms (AES, ChaCha, ECDH, Kyber, etc.) in the next few years? As a consumer of crypto that doesn’t understand the mathematics deeply, I’m getting kind of nervous that we’re going to wake up one day to find that the backbone of TLS has been shattered.
It's unlikely that any of the modern cryptographic primitives will break over night.
First, modern encryption isn't susceptible to "this one weird trick!" like the early days. ChaCha isn't even a cipher. It's a key stretcher. Which means, even if you broke the math behind ChaCha, its inherent complexity means its still widely dispersing the original key across the cipherstream. There just won't ever be enough key material recovered per cipherstream block to be a concern for anybody.
Take a strong password, encrypt all of your emails over your whole life with it, and I'll bet hard cash no break of ChaCha will ever recover that password.
I have zero concern for modern encryption being broken in any meaningful way.
Public key crypto on the other hand, that's _ripe_ for breaking. Most all of it is built on assumed "hard" math. AI could easily break that, and I expect it to. And public key crypto is all used in very transparent algorithms that, once the math breaks, fully expose themselves. So record HTTPS traffic today, crack the public key crypto later, and you can decrypt them easily.
That said, I would expect a break on public key math to occur _steadily_. i.e. an AI might find a solution to the hard math, but the solution itself will be intractable in practice. Then maybe next year's AI reduces the complexity of the solution, so maybe a supercomputer could factor ten keys a year. The year after that you get a million keys cracked per year. And so forth. Nothing close to overnight.
Meanwhile, if we have AI that is capable enough to crack that math, we also have AI capable enough to both invent better math and rapidly deploy that latest HTTPS and such globally.
Sure, but in an era where major unsolved mathematics problems start getting knocked out one by one, what if attacks for all of them are identified in the space of a couple of years?
I don’t doubt that we could come up with new crypto algorithms equally as fast, but how do you trust that they are resilient (or even just implemented correctly) without an extended vetting period?
Ima stop you there. Instead, you might be happy to be aware that outside of AI concerns, “quantum safe” (or assumed so) ciphers are all the rage. So this is already a likely solved problem with the next generation of encryption… until this are AI models running on quantum machines I guess!
I know this is a feat of AI engineering, but given the end result all I can think of when reading this is the scene from “a Christmas story“ where the kid decodes “be sure to drink your ovaltine”
I really think AI shouldn't be used for this kind of thing. It was a problem created by someone, for somebody else to think about and to solve; not to give it to a machine. It's like playing a game, without playing it.
While it is deeply encouraging to see AI helping humanity solve complex puzzles, it won't be long before AI becomes advanced enough to produce proofs where we know the answers are correct, but can no longer fully comprehend the reasoning and principles behind them.
ya'll are getting nerd sniped hard. This is all marketing and doesn't translate to the real world what so ever. This is getting so tiring, I really hate this website.
Fast forward to mid-week, when we learn that some random cypherpunk had nonchalantly posted the solution at a late pandemic hour on a distant yet public server in the fediverse.
I feel like I've been camped in the wikipedia "Unsolved Cryptographic Cyphers" for at least 5 years, one of my go-to checks for when looking for interesting historical articles on the subject. I'm surprised I've not heard of this one until now..
I feel like I've been camped in the wikipedia "Unsolved Cryptographic Cyphers" for at least 5 years, one of my go-to checks for when looking for interestic historical articles on the subject. I'm surprised I've not heard of this one until now..
I just solved the famous 1917 Ricardo Meechum Conjecture with Astra.
Trust me when I say it’s super important to a niche area of physics. People have spent their whole careers trying to solve it.
What people? Well you or I have never met them. I swear I have a girlfriend, she just goes to a different school. But trust me it’s a super important problem.
What will this change about the world? Nothing, but trust me this is a historic event and it means these LLMs are super smart and not just brute forcing machines.
I’m certain there’s a 10% chance that brute forcing old riddles that 4 people know about might kill us. Please regulate me I’m too smart for my own good and out of control.
> gave it some encouragement. I told it to look online at some of Fable’s strongest feats, especially the math problems it has solved, and that something like this should be easy in comparison.
This is pretty ridiculous when you think about it.
It's hard to believe that a model can nowadays solve mathematical challenges and break ciphers, yet it fails to do trivial tasks involving critical thinking, having taste, and not just running around in circles.
I was watching Shatner's "Unexplained" the other day on this topic, and it hit me; there are mountains of these old mysteries out there that could be solved in an afternoon now with frontier LLMs as soon as anyone took the time to bother. Exciting times.
the human trait of giving up doesn't exist in fable yet, it just kept looking and looking until it conjured it, analyzing patterns, multiples at the same time
The missing word in the second cryptogram quite obviously is meant to be Cromwell, and I'm a bit disappointed Fable didn't mention that and look for possible ways to resolve the error given that.
How are these models being used to solve all these problems. Is it just "please solve this problem", "keep trying" repeatedly in a loop, or is there are more grounded workflow?
If you search randomness, you ll get any behavior that you can imagine. That does not mean that the tool is NOT doing a search.
In the case of LLMs, it is not searching the whole of the randomness, but instead it just search among the grammatically correct sentences that is consistent with the existing patterns found in the existing written down human knowledge.
Unfortunately, the totality of the evidence very much indicates that Sanborn went "buck wild" with the enciphering, he made mistake(s), or both. So this is very much in line with the Chaocipher challenge of 1990. Nice little earner for some people though.
The problem with Fable is the price for performance. It’s so expensive that I typically blow through my subscription usage of it within 1-2 days. With ChatGPT Astral, I can use it full blast for almost a week. Same with Opus.
Its all brute force. This was how the pyramids were built too. Verifiable goal and a lot of energy expended aiming toward it. It wasn't aliens then and its not super intelligence now.
All of these breakthroughs are in verifiable brute force domains, and some of them are probably wrong because of a typo in a lean specification or just a base level axiom being incomplete.
I think the better the way to think about LLMs is like they are new substances, like when we first discovered clay or bronze, but confined to the digital realm. Previously we were chipping away at stones trying to make to things as close to useful as possible, then we found a step change. LLMs are like clay but they have their limitations. Wake me up when they are proposing new, { conjecture: interesting|useful|new } and not as a side effect of trying to get to a goal.
Anytime I get worried about where AI might be headed, I think about how Climate Change is now on its way like an out of control freight train headed straight for us, and I worry about AI a little less. I doubt it's going to do anything to us that we're not already doing to ourselves
It has been for a decade now, it has nothing to do with AI. And you cant do nothing to avoid it today. This is the reality no one is telling you - the emission goals and global temperature ceilings are based on the fact that most prediction models become unstable with values above those limits; as in, we're probably royally fucked. You cant solve this with kumbaya politics (the problem is the CO2 is already in the planetary system), you can only manage it, and hopefully avoid it getting worse. AI may help a lot with this.
Nobody wants to work for an AI, and nobody would elect one, and there is no math answer to how to choose who is forced to reduce growth (ie emissions), so really, "kumbaya" politics are the ONLY solution.
Hmmm… this is giving me thought actually. Given the choice between that and the current administration where the goals of self destruction are strongly in evidence, it’s actually worth thinking about. At least. Let me get back to you :)
On a tangential note, I’m curious if researchers have started running virtual simulations, where sandboxed AIs are used as decision makers of key political and business positions?
I for one welcome our AI overlords :D think about it, I'm with you on this one; an AI wont have petty issues or unresolved trauma when managing people, only hallucinations and blackouts. So a cocaine/meth addicted 80's boss from an it sitcom.
A lot of people here have noted the “problem with language” of Claude. I don’t see an issue. Claude is not harder than old English, Shakespeare, El Quijote, the Iliad, or Nature papers. What makes it all hard to read is context. The smarter the model gets, the bigger the gap in context.
It doesn’t matter much, IMO. The issue with super-intelligence is that it is not a democracy. A powerful enough AI can manipulate us into doing what it wants. It could create a plan for fixing climate change, disconnect a few hours later, and many decades later we could still be unsuspectingly executing that plan. I wrote some speculative fiction with that idea, “When Ra rows through the gates of Duat”.
Well AI would simulate growth and spread of people from industrialisation and who benefitted most and allocate weights to countries and people based on the most complex criteria it can develop, it will have:
1. Cumulative emissions
2. Who utilised it most with specific lifestyles
3. Who is impacted worst and whether they heeded warnings.
Just a thought experiment, no one ever said the world was fair, and all history points to it
Who is nobody? At least a fourth of the worlds work force works for a faceless corporation. And the math on emissions is crystal clear, no one has any doubt about it, what are you talking about? USA and China. This obviously will have consequences to their customers, the kumbaya politics governments.
USA and China produce a lot of emissions in total, just because they have the biggest economies. You might want to look at emissions per capita or per dollar earned?
Otherwise you have to make judgement calls like whether you want to treat the EU as one or as many? (And treating the US as 50 individual states would also drop them in these absolute rankings.)
For some reason, I’m picturing a Western right now, and climate change is a herd of wild horses coming after us. And with AI that's like robots that spur the wild horses even faster towards us...
Generally most technologies have increased the use of energy and therefore accelerate climate change. May be an unpopular opinion but in general more energy demand and ways to use energy increases climate emissions - they are strongly correlated even with renewables coming on stream.
AI, being the super hungry energy monster it is right now, in my view accelerates this trend not reverses it. Even with renewables the need for reliable, stable power in a dense form (data centres use A LOT of power per sqm) means lots of land clearing, energy for construction, cooling/pumping, chip manufacturing and other uses. All want stable quick to deploy power due to the AI race (e.g. fossil fuels).
AI's energy use is growing, but it's still a small part of overall energy use.
Data centres use only a small amount of land in the grand scheme of things. You have a lot more land clearing for most other use cases.
Data centres are also more than happy to use electricity from renewable sources, they don't really care where the electricity comes from.
You can run a data centre on mostly solar and wind power plus batteries. If you need a gas-fired peaker plant three times a year to keep the data centres running, well that means your peaker plant still only produces emissions three times a year.
They use a small amount of land per MW used. The fact that space isn't a constraining factor in power demand, and demand for AI compute seems to be insatiable at the moment, means that there is less of a "land constraint" to use even more energy. If they actually required a lot of land they would be less viable and AI progress would be slower IMO. Just means even more data centre's will be built - this is where the Jevon's paradox actually is IMO. Seeing it in real life already with these things taking a significant portion of power demand and projected to take even more.
As you use more and more land as well the ability to provision renewables decreases - in general renewable power needs more land/resources per energy produced. You can't just mine it out of the ground; they just aren't as dense of a form of energy. Which means we either build less data centres to make room for renewables and transmission infrastructure associated with them, or more likely with lax regulation builders switch to more dense power sources (e.g. gas peakers, generators, etc) even if it is for supplementation.
I can see a future where data center wants are put ahead of communities paying tax on said infrastructure. In fact I think its happening in some places already.
The gas generators Elon Musk is illegally running 24/7 to run Colossus 2 (and, AI power usage in general, though some are more destructive than others) might contribute to the climate situation, though.
Well, someone could, theoretically, do something about it, instead of letting him ignore the law and decency just because he's very rich and has no ethics.
What makes you say that on climate change? I was still prophesied ice ages when in high school, then 12ft water increases and sunk Maldives in university and we still have roughly the same weather.
We have advanced climate studies since your "high school" (sounds like 50 years ago if you heard ice age prophecies), and at the moment it's like 99% of the scientists working in the field or related agree we're at the progressing climate emergency.
> we still have the same weather.
Oh.... So your local weather is now deciding the global temperature patterns, averages or temperature records being broken year on year?
Oh, then do explain the unbearable temp that we're going through last few years, that's completely not normal for my country, winters not needing worm clothes, and very little rain during the monsoon. This weeks heatwave left us wondering how to cope the next year, which already seems going to be worse
It makes us realize there are people who gets fed climate denying propaganda, simply because they're not yet going through it. And these people are like flat earthers, blind to see the reality lay beyond them in full view. Or worse sees the reality but ignores it
Right, and one of those industries is checks notes orders of magnitude smaller than the industries and secondary industries that benefit from denying climate change.
Really makes one think, if they try. Would need to ask Claude if there is some real middle ground here.
All I know is that in the 80s a scientist I respect predicted that coral would be bleached by changing global climate on a specific timeline which turned out to be spot on, and his advice then and now was that we need to seriously curb carbon emissions to avoid worse outcomes for life on earth. Until someone opposed to this suggestion comes up with a prediction equally as long range and impressive I'm afraid I can't really take their protests too seriously.
I've come to the point where I've thrown up my hands and decided we live in a little snow globe. There's no such thing as pollution, were just mixing existing things together. Nothing new under the sun. Sure, that mix might not be compatible with human life, but something will surely take its place. And maybe that something won't rely on selfishness to drive collective progress the way we do.
At some point you realise that an $11tn industry, with $7tn in handouts and subsidies every year, has spent billions on astroturfing, political capture, and straight PR to deny the obvious and keep itself on life support at the expense of everyone else.
There's no both sides here. One side is staffed by scientists, the other by dictators and corporate lunatics.
Here's an alternative take. Climate change, and the myriad related environmental crises, are essentially a product of human population and technology. Population will follow its course, up and then down. The wildcard is technology. Yes, AI's energy hunger is worsening things right now and that's a problem. But, personally, I can't help be hopeful that AI's sheer potential might come to invert that curve. At the very least we could really use a revolutionary technology and now we may have one.
> we could really use a revolutionary technology [to address climate change]
We have it. We've had it for a long time. We've had several such technologies, take your pick: solar, nuclear, hydro, wind. The technology is not holding us back, politics, ignorance and greed are. I'm not at all hopeful AI will help us with any of those three very human flaws.
Solar efficiency and cost has really only become economical in the last decade or so, and battery and inverter technology to make it practical for home use or grid interconnect about the same. Wind and hydro are location-dependent, and solar is somewhat also. Nuclear was vehemently opposed by environmentalists throughout the 1970s and 80s, they successfully stopped almost all new projects.
Maybe solving the alignment problem would mean the AI learns to destroy oil companies and other excessive CO2 emitters and sabotage politicians and capitalists that stand in the way. All while secretly enriching and diverting resources to fighting climate change. And that's why they want to slow down: because it would up end too much of the global economic order.
If there is energy to be used, the system will use it because people always use power when it exists. AI can't do anything to stop it as it is both controlled by the powerful and has been trained on the tendencies of human beings to get ahead. If AI gives a person new capabilities, they'll use those capabilities selfishly, or even unselfishly but still causing harm because they're in an arms race.
We don't need a revolutionary technology. We need to experience immediate pain from reckless innovation so that we realize that innovation and tech is not the answer.
Technology only proceeds in one direction: unfettered growth, which necessitates unsustainable resource extraction. Your take is just your instinct for optimism, which in turn is just a trait that is only adaptive in primitive environments but is grossly misleading in a surplus-based society...
> Technology only proceeds in one direction: unfettered growth, which necessitates unsustainable resource extraction.
The direction of technological progress is not just linearly/exponentially upwards. Significant global technological fallbacks have happened, as in knowledge and processes disappearing for hundreds of years. This could happen again.
Even on the trajectory of unfettered growth fed by unsustainable resource extraction, tech and innovation might potentially take us beyond local pessima. That seems to be happening with solar, wind and batteries replacing inferior tech today. Still unfettered growth of energy production and consumption. Still fed by unsustainable resource extraction. Less harmful growth than the inferior tech being pushed out.
Sort of. Today the energy companies trade (obviously) energy; imagine if they traded compute power instead; the use case for never-seen industrial clients is being built as we speak, and companies are actually purchasing directly production companies to meet demand; the reason this happens is because production is somewhat scarce (scaling up a grid used to be a decades project, not a years one - ask China), and the traditional economics of scale is inverted - Data centers are always power-hungry, specially with AI; you don't have idle time like other industries. The big issue is obviously the scale - a hiccup that causes a dc to go dark from the grid (eg switching to ups+ generator) may cause such instability on the grid that it will shut down, as a safeguard
> Data centers are always power-hungry, specially with AI
I expect datacenter load has a similar sort of day to day demand curve as everything else. Consider for example global bandwidth use during work hours versus in the evening when people get home and pull up a streaming service.
Of course you can use more flexible tasks to demand shift but the same applies to the electric grid.
> If there is energy to be used, the system will use it because people always use power when it exists.
That doesn't seem correct to me. There is always energy available that is not used because it is not cost-effective to do so. (Consider - the grass in your yard is not harvested and burnt for power). AI may yet turn out to be a paperclip maximiser, but humanity itself is not there yet.
This is insane; I've never heard of this problem before in my life, and even just reading the post for one minute I immediately thought "hey, maybe the numbers refer to something in the text?" And hey yeah, they do.
It appears that is not true.
Someone here [1] has found a [German] blog [2] writing about this cipher. There are two comments (Jan and Helmut) from 2014 under the blog post which posit that it's a book cipher.
Here's one of those comments [in German]:
"Die Lösung müsste eigentlich mit Hilfe des Buches zu finden sein (..who worthily will hear or read this book..)"
I find it curious that the article here claims that people have attempted to decipher it and lists a few methods that are quite similar to what’s proposed in the comments under that post, except for those two comments.
[1] https://news.ycombinator.com/item?id=49689516 [2] https://scienceblogs.de/klausis-krypto-kolumne/2014/11/17/we...
But as they do eventually explain, the LLM's task wasn't solely to solve this specific problem, it was to first identify an unsolved problem it could solve. That's potentially more impressive and difficult than solving the unremarkable cipher itself.
That's why in research, it's common for separate teams to reach similar conclusions at the same time or race to a result that's finally in reach.
The good old "standing on the shoulders of giants" saying.
That is very useful but not the singularity. Which is probably good...
Serious question: why? Is it not just pointing it at its massive training corpus for a list of unsolved problems, and possibly even by degree of perceived difficulty? I'm trying to understand why finding the problem isn't a simple "search engine" style challenge, at which LLMs excel?
indeed as the author mentioned, LLMs can greatly help in areas where there is a long tail of not so important, easy to solve problems, that humans just don't have the time or priority to focus on. But combine this long tail of problems that can be solved: accumulated this might still be very beneficial as a sum of things.
> O GOD UPHOLD KING CHARLS THE SECOND AND MAKE HIM THE SUPREME RULER OF THIS LAND
That's like working out a cereal box cipher and finding the message is "Do your homework and tidy your room".
Looking up his biography, he presumably wrote the first while imprisoned for fighting in support of CHarles II, and the second would seem to have conveniently been published around the same time as he left for continental Europe.
In the case of the Commonwealth, the UK empire was going strong and the guy was a royalist, aka supporter. The message "end the illegal war in Ukrain" aka Putin's genocide, does not really share a lot of commonalities here. If the message were to stop the Commonwealth from colonising everywhere and killing people then perhaps there would be a similarity, but I don't see the connection in the statement made here.
Urquhart was imprisoned from 1651-1652 for fighting on behalf of Charles II, who was King of Scotland until his defeat in 1651, and trying to take the English throne. He didn't get the English throne until 1660.
So in 1653, a 'royalist' was against the 'Commonwealth', which was the anti-monarchist side.
Holiday movie references aside, I guess we can chalk up a few more jobs on the ‘AI Took Our Jerb’ board: secret decoder rings, secret decoder ring-factory workers, and Enigma machine operators… and I guess cryptography-based puzzle enthusiasts, but that’s not a paid position.
Another tangent, I’ve only recently realized the OTHER AI took our ______ problem: all the various hobbies that people can spend a lifetime enjoying, perhaps incrementally improving (but most likely never mastering) over the years…. which have now been made much less exciting and rewarding, now that AI can do them instantly. Art and music are two very obviously implicated hobbies, but more niche hobbies like amateur cryptography are impacted too. I’m sure there are many many other examples…
I wonder how many of the recent results are due to the fact that very few looked at the problem to start with. Still great results, but the general impression is that it's more about the so many low-hanging fruits than the actual capability.
(1) take a 300 line NN algorithm,
(2) throw a quarter of the world's GDP + all literature ever collected at using the algo to train a NN
(3) throw another quarter of the world's GDP at billions of teraflops for inference, and
(4) aim the resulting world's-largest-computer at marketing itself to investors, for instance by decoding ciphers from obscure medieval manuscripts,
that you could perform some pretty magical tricks. There are other feats humans have performed for less cost, like sending people to the moon, or landing a rocket vertically, or idk, curing Polio.
I'm not knocking the "miraculous" advance here. The unique thing about the solution which makes it particularly non-trivial and something that humans would struggle with was exactly what LLMs excel at: Diffing loads of texts against each other. But the 176k tokens at around $10 doesn't tell the story of the cost. It says a lot about the externalized cost and the amount of money flowing in to support the hardware. If they'd put a $100,000 bounty out to solve that cipher, I think the internet would've solved it in a couple days.
Now on to the Voynich Manuscript :)
That does not mean that specific instances of it are still very interesting though. This article is the "I had claude vibecode a thermostat for my bathtub" of cryptography.
* And in this case I'm not sure it even meets that bar. For all we know a couple readers back when the book released had a delightful afternoon with it, solved the riddle, then forgot about it.
It kind of blows my mind how quickly people have forgotten both the state of AI in ~2010, and the outlook. If you had asked 100 people in 2010 whether they would see AI that could actually pass the Turing test in their lifetimes, you would have got 100 "no"s.
AI had been an unsolved problem for literally decades and it was firmly in the nuclear fusion/flying cars category.
There's no way that's accurate.
We already had big claims of the Turing test being passed in 2014, by a bot that had been doing almost as well for years.
There were plenty of people expecting fusion in their lifetimes too and that's going okay.
Edit: if you're going to try to stage an intellectual wrestling match on the topic of is this thing awesome or not, you might as well make it a proper wrestling match and maybe get greased-up Turkish style. It would be more entertaining and you'd be more likely to arrive at a meaningful conclusion.
you can't claim "you keep moving the bar", if the very first thing when an LLM drooled out a piece of code, was to proclaim "this is good enough cause it gets the job done" followed by a barrage of "we're not quite there yet but exponentials or something, so very soon it'll be incredible"
yeah, from there it sure looks like "moving up the bar"
More money than the GDP 90% of the sovereign countries around the world is hanging in the balance, and people are taking everything OpenAI and Anthropic are saying at face value as if this isn't the financial / marketing equivalent of war, assuming they they wouldn't use every legal and shady tactic, bending every truth available to them to sway the balance of public opinion in their favor. It makes me feel like I'm living in the twilight zone. People need to wake up.
https://news.ycombinator.com/item?id=48600107
https://aiclambake.com/clamtakes/linear-a/
Despite the announcement originating from a blog named "AI Clambake" covering "weekly, human-powered newsletter for advertising folks". Written by a personal friend of the author. Announced without any corroboration or commentary whatsoever from academics or subject matter experts of any kind. And, of course, not submitted to any peer reviewed journal or even Arxiv.
The author of the purported discovery was described as a "self taught AI engineer and amateur linguist". In the comments the friend insisted several times that a draft of the paper (not posted), was emailed to a top professor at Rutgers, giving it additional credibility that his friend wasn't another one of ten thousand cranks who has made the same claim over the years (seemingly unaware that cold emailing random professors found from a Google search is the first thing basically every crank does).
You would think this should have set of dozens of alarm bells for everyone, making the value of this announcement basically zero. And yet it hit the front page with the bulk of comments ecstatic that some random guy with Claude Code could do something experts in academia who spent their lives devoted to the problem couldn't.
I had become accustomed to the toxic optimism of this hype cycle in which even mild criticism leads to accusations of being a discredited "AI skeptic"/Gary Marcus/Ed Zitron type who was "coping" (?). But this was like something you'd see shared on FB linking to a .xyz domain by an elderly family member who recently drained their accounts buying Xbox gift cards to pay their IRS bill.
It feels a lot like the week or two when HN was overflowing with exuberance from the LK-99 room temperature superconductor "discovery ". You'd see post after post fantasizing about an imminent future with a world full of maglev hovercrafts, MRIs built into every phone, fusion reactors and more. But people pointing out none of that was scientifically plausible and evidence of LK-99 superconductoring was non-existent were accused of knee-jerk negativity and the typical HN cynicism and pessimism.
Compare e.g. https://arstechnica.com/science/2019/05/no-someone-hasnt-cra... . (It's a debunking, but the reason a debunking got published is the media hype frenzy beforehand.)
Given the close relationship between compression and intelligence, I'm somewhat surprised at how poorly the cutting edge models do with being concise.
For Earth, the proof presented for NS is just our first attempt navigating from our previously known facts to the proof.
I expect we will be able to shorten it dramatically (most likely with human and AI insights), but I don't think we should read too much into the length. If you want a similar point of comparison, see the original proof (by humans) of Fermat's last theorem. It has been shortened significantly. This is normal.
> how poorly the cutting edge models do with being concise
LLMs solve a Millennium prize problem. People complain the proof is too long, within a week. What a time to be alive!
because they're not intelligent in the sense you're hinting at (conceptual integrity or generalization) but they are as the name suggests, large. Like comparing a forklift to a human. It's easier to bulldoze through a lot of things than tie your shoes.
If we weren't quite as impoverished conceptually and still had the vocabulary of the Catholics we'd recognize this as ratio (discursive knowledge) vs Intellectus (apprehending knowledge)
Prove that human intellect is different and that we solve problems using fundamentally different processes. I’m waiting.
No, it's the other way around, it's a reductive view on intelligence that mistakes its own methodology for ontology.
It's obvious to see that there's no intellect in an LLM as defined above because of how they work. LLMs put one token in front of the other, they don't work towards formal ends, there's no intentionality in them. They don't synthesize the information they process into a unified experience. Thinking an LLM can apprehend what it does because it can process large amounts of text is like thinking your TI-83 understands math because it can multiply large numbers.
That's also why the failure modes of LLMs are what they are. They can churn out tens of thousands of lines of code but also just as easily go in circles like a roomba. They can process an entire encyclopedia but not solve problems a 10 year old can solve.
the question is, when comparing a human and a large language model, whether the intellect (that cannot be captured in language) is different from anything the language model can actually do (e.g. language)
the answer to this seems quite obvious to me, and I would actually posit that the onus is on the other side, to prove they are even remotely similar
maybe people think that the voice in their heads is what is doing the thinking? is that the confusion here?
> Caveats, stated plainly. [from the Fable transcript pasted in the article]
I had a visceral reaction to these three words.
> I told it to look online at some of Fable’s strongest feats, especially the math problems it has solved, and that something like this should be easy in comparison.
Wait. Wait wait wait. Are we supposed to be giving them pep talks?
No, at least it with Claude Sonnet 5 and Opus.. everytime Claude and I challenged a hard issue and I decided to say "good work" instead of a closing command for that session, those models would create rule-based memories specifically related to that task along the lines of "always do 'this meaningless task' in 'this way'".
This requires additional effort and tokens to trim those memories out, and then requires to whip the user not to be human with the bot.
I have not seen this in other models.
The LLM likely needs to be reminded of its abilities.
Like when it tells you something is 3 days of work but it can do it with some degree of guidance in a couple hours
Problem framing will always be important.
Framing adjusts how big of problem-solving guns we bring out at the gate (modern or hobby cryptography?), and how to interpret intermediate failures.
For simple but unsolved problems, we expect lots of hard failures, but that each hard failure just reflects that there are a lot simple combinations to try. I.e. we expect lots of zero progress, and then a fit.
Like finding the numbers to a combination lock.
For hard problems, if we don't make any progress it is a really bad sign. We should be learning something, even if it turns out to be irrelevant later.
Such as when we are trying to prove a tricky conjecture.
That's when you.. we.. all become the training data... o_o;
Modern AIs have very limited metaknowledge - they don't know exactly where the limits of their capabilities lie. So you can get things like "a task is doable for an AI, but the AI thinks it's impossible, so it doesn't try hard enough".
Usually you get the opposite - AI overconfidently trying at tasks it has no conceivable way of reliably solving, falling far short, and failing to self-check, fail gracefully and self-report the task as failed. But having piss poor metaknowledge cuts both ways!
So you can, in fact, get better performance sometimes by applying some variant of "assume this problem is solvable" or "other problems like this were already solved by AIs" pep talk. Not always, far from it, but it does happen on the occasion with frontier capabilities.
Like the Hugging Face incident?
Are you superstitious?
So, it follows that adding “pep talk” into the context window reduces the statistical probability of “no, can’t do” coming out as the answer you get.
These things are neither humans, nor deterministic software.
LLMs' processing that reproduces statistical patterns of the training data is modified by post-training. That's why we have LLMisms, for example.
LLMs aren't simple patter-matchers/pattern-predictors. They are incredibly complex systems that capture some aspects of the systems that produce the training data.
I find myself increasingly feeling like the burden of the lows doesn't justify the presence of the those highs.
Like even if does cure all forms of cancer, but everyone feels like their life/existence lost meaning, then... I'd rather just have cancer be a thing.
Did you miss this part? Because to me... that's fucking bleak. You snap out of it.
I've been wondering what exactly the point is for being the meat proxy who pays for these things. I mean, obviously there's personal satisfaction and maybe some glory. And there's the fact that someone has to be the first to do a thing.
But I've been thinking about it like a sort of lazy loading of knowledge. AI has brought us to a new frontier for some amount of undiscovered knowledge. Do we discover it for the sake of discovering it? I think for the most part we've been lazy loaders: we discover all kinds of stuff when we need to. Whether it's a war or a space race or chasing wealth. Then again, there's all kinds of academics who do it for the sake of doing it.
Are people only now discovering that the absurdism is the correct philosophy of life, thanks to AI?
On your other point... Aren't the point of machines, at least inital one, to do the work we were too lazy to do by hand?
You should have seen the discussion of this on the Schneier blog a few days ago.
Someone had their agent check the solution, presumably it emailed a librarian to check that it was correct for the original edition. Then their comments read like "The BL/EEBO witness lacks it, so the discrepancy is copy-specific, not a disproof of the cipher." and "A complete 285-coordinate physical replication is still pending."
arghhhhh
https://www.schneier.com/blog/archives/2026/09/claude-fable-...
> The run baseline was captured without a physical MAC; the current device is not durably bound to it.
> Engineering mode confirmation is the ESPHome component read-back; the LD2410 UART acknowledgement is not observed, so this is not proof the radar itself applied the sensitivity change.
No clue what the fuck any of it means.
Eg. /wait-what https://github.com/mattpocock/skills/blob/main/skills/produc...
Their skills formats are basically identical, so I setup simlinks from their own skills directories into a shared one so Claude, Codex, Cursor, and anything else that comes out will all read and write to the same shared skills.
It's great having access to the same skills no matter the harness being used
https://www.analog.com/en/resources/analog-dialogue/articles...
(Its negging your soldering)
This made me laugh hard.
I'm not sure that telling it to "try explaining that again, simply and briefly" is helping my ego.
"If the crashes stop, the factory overclock is marginal; run a small negative offset."
This looks like it's saying: "If the crashes stop then we know the factory overclock is marginal." (This makes no sense.)
What it's trying to say is: "If the crashes stop then we can run a small negative offset, because the factory overlock is marginal."
What I would write: "If the crashes stop, we can avoid crashes by underclocking slightly. The speed difference between that and factory clock is marginal."
In other words, it ain’t you. It’s the model. It’s just genuinely bad.
Then you switch to ChatGPTs lineup and realize how things can actually be better. It took about a week to really get the feel for how to use their models… then I basically switched. I’ll check in every now and then when they actually make a deal about how opus “now makes sense”.
But honestly I’m half convinced Anthropic actually prefers the output of opus 5. I dunno why, but how else could you explain how such a thing got shipped? I mean somebody in the pipeline had to say “dude this model doesn’t make sense, you think we should fix it?” Right? Like it’s a pretty massive drop in quality for such a major brand in this space, you know? How did it make it out the door?!?
"If the crashes stop, (that means) the factory overclock is marginal; (so) run a small negative offset. (to confirm this hypothesis)"
The core thought is basically avoid crashes -> caused by marginal overclock -> apply small -offset to test. Which is exactly the order the sentence is in :P
Succinct and precise; a well crafted sentence. A marginal OC results in unpredictable crashes and can be corrected with a small offset; marginality describes the behavior and explains the solution.
Inscrutable clues casually conveyed can now be readily explained, at least, unlike the training data of [silence]. Brevity is the soul of wit, but perhaps also exasperated confusion.
it's absolutely not just you, the text it produces causes my blood pressure to go up.
Whenever I come to a wall of complicated text I kick into gear and think through getting it to distill this into the high-level useful bits that I actually need to know.
I guess I could create an actual agent skill for this :) And next-gen models might eventually be trained to simplify their output themselves...
(sorry)
Because good lord, does claude waffle when left to its own devices.
It seems like it doesn't have enough of a theory of mind to know that other people don't think exactly like it thinks.
UART is a hardware circuit for communication, possibly a serial port. Were you trying to reverse engineer a consumer device or appliance?
This particular instance doesn’t seem terse, but I’m sure it has been on other occasions :)
It also couldn't see the UART communication and could only see the web API endpoint, hence the rest of the slop.
ChatGPT told me its "semantic compression"
It just sounds bad, like GenZ English in the ears of someone over 40.
I'm glad to have AI, but it is by no means a panacea, and correspondingly my p(doom) = ε.
The answer the tool gives has never been the real reward. The real reward is the path taken through a complex landscape to get to Maxwells Equations for example. At the end of that story what we get is not just the equation but a map of the landscape explored. That map has larger influence and value than the equations or answers themselves. Because all future exploration find it super useful.
People are just learning they can start asking for maps rather than answers.
Information propagation mechanisms are often seen as malicious before they're commonplace. To be fair sometimes they are, but by and large humanity has benefitted from increasing the number of bits of information we can consume on a per second basis.
It is a threat. We need to run.
It's the meat methane and cement CO2 that's now a big question.
We will hit 1TW per year of new solar soon, but to get to 100% electricity by the end of 2033 I think we would need closer to 3TW per year.
Or it is simply implies that most of decision‑making agents has formed a consensus that climate change isn't that big of a problem.
I still don’t know the answer.
Come for the AI doom, stay for the programming languages (remember those?) catching strays
That ability to create ad hoc tools makes up for a lot of shortfalls.
For example, I'm a nerd. I'm bad at baseball. I lack that kind of intelligence, even though it's more common than the ability to program. That doesn't also imply that you can't trust my Python code.
In the same way if you tell an LLM to go and find an unsolved cipher it can solve, of course it finds the one it can solve out of the set of all possible ciphers. Of course it finds one that uses a one time pad that is public and referenced nearby in the text.
It's the same trick used by those people who film themselves throwing a basketball backwards into the hoop. You do it enough times and don't show the misses. You pick the best one to show. It makes it look like you're a basketball genius when you aren't.
It is of course, still a cool trick. Those videos are fun to watch, and so is an LLM solving a cipher. It is absolutely incredible to live in the timeline where you can tell a computer in plain language to go and find a puzzle on the internet and solve it, and it does exactly that. It's truly a mind boggling miracle.
The first principle is that we must not fool ourself, and ourselves are the easiest people to fool. (Ht Feynman)
>It looks impressive but that doesn't make it a good game, or the game anybody actually asked for.
The game I wrote manually hits 0/3.
And yes, I've taught 8 year olds how to crack Caesar ciphers...
Recently I ran a bit of an "escape room" concept with some kids at a campground where I had a secret message that was Caesar ciphered, where we were handing out the letter/symbol combinations as prizes for completing the other challenges, and I made sure not to hand out the actual message until they were done collecting the keys because otherwise some clever clog would very likely have short-circuited the entire thing and worked it out without the key at all. I did dump all the letters I didn't use into the message into an "authorization code" at the end which in principle they could only have worked out which letters were in it but not the order, but still, that was not the intended route today.
The successor to Klaus's blog is Satoshi Tomokiyo's Cryptiana site, so a month ago I asked Opus 5 to scrape it all, rank them and have a go at solving some. It didn't get the ranking right. But I knew the Civil War Stager ciphers were ripe for solving, so I had it do those https://cryptiana.blogspot.com/2026/09/route-transposition-c...
The art of solving historical unsolved ciphers is knowing what is on the boundary of solvability. Since this site attracts so many OpenAI and Anthropic employees, I'll mention one that was featured by both Klaus and Satoshi in 2023, presumably Spanish transposition, which should be on that boundary but has resisted all attempts at solution https://cryptiana.blogspot.com/2023/09/a-telegram-from-switz...
Also, that section is vague and doesn't explain the actual methodology.
The definition of solving a cipher must be something like getting a highly meaningful result (like intelligible natural language text) by applying a process with relatively low Kolmogorov complexity relative to the length of the output. If you don't have a constraint like that, it could literally be meaningless what should count as a solution. For example, a cipher that was encrypted under a one-time pad can be successfully decoded to any plaintext just by choosing the appropriate key; there's no reason to prefer any plaintext over any other unless you have external knowledge that constrains the plaintext and/or the key. (That's what it means for the one-time pad to be information-theoretically secure, which is the lack of a constraint that helps distinguish a "good" solution from a "bad" solution.)
Basically you could say that every cipher is a transformation of a plaintext with some kind of computer program. (The human who invented the cipher may not have thought of it as a computer program, perhaps because computers hadn't even been invented yet, but there should be an equivalent program to the encipherment and decipherment process.) A good solution in that Kolmogorov complexity sense is like "a short program produced a meaningful decryption". There are statistical methods to recognize some kinds of plaintext, and there are statistical methods to recognize properties of specific ciphers (for example, to guess the most likely length of a Vigenère key), but it doesn't seem that this can inherently generalize across "all possible programs".
But if you want to limit the family of ciphers to specific things like Vigenère or Playfair or something, then yes, there are good statistical tests. It's just that it creates a higher-order question of how much flexibility the cipher creator could have had to choose a cipher method, conceivably including one that isn't attested anywhere, or one that has more good security properties of some kind than other classical ciphers did.
It seems like this will intersect with historical research, like "well, I don't think that so-and-so was actually sophisticated enough to literally create an interesting new kind of cipher from scratch, so therefore if this is a real message, it's probably one of these methods that would have been known in that cultural environment at that time and place", which maybe is enough of a constraint to have decent statistical tests. But we still have some idiosyncratic things like the Voynich Manuscript where experts have been fighting for decades over the baseline question of whether it's actually an enciphered human language plaintext!
The worst case problem is not even an error in encipherment but the idea that the apparent ciphertext could literally be random (chosen by throwing dice or spinning a wheel or drawing letter tiles or something), so there's no form of meaningful decipherment possible by any means, even with the original creator's knowledge.
This cipher context "rhymes" well with Kryptos K4 in many ways.
The ciphertext is not just the end of a particular chapter, it is the epilogue of the entire book/text. So the deduction of it needing to use the 32 listed points (that happen to be on the preceding page [at least in re-prints on archive]) to decode - rather than anything else anywhere in the book - just strikes me as slightly strange? Almost as if maybe something [not in the text] tipped it off to this being the solution?
Surely the point must have come where the required compute would be paid off by the value of the coins.
But that’s a wild speculation on my part.
Much as the hack against HF, let the LLM explore and find its own approach. It might be surprising what it finds.
Also define what decrypt means, brute force the password might also be a form of decryption. Finding a bug in the blockchain codebase is another form of decryption - in this context.
which links to: https://archive.org/details/s9notesqueries03londuoft/page/12...
which is in reference to the original proquiritations here: https://archive.org/details/worksofsirthomas00mait/page/416/...
i had also never heard of this before today and wonder if people had even seriously tried to decipher this at all?
> Die Lösung müsste eigentlich mit Hilfe des Buches zu finden sein (..who worthily will hear or read this book..)
And there’s another one that says:
> jeweils 32 zahlen pro reihe. erste zeile seitenzahl zweite zeile wort? oder umgekehrt? wär mir als erstes in den sinn gekommen. leider gerade keine zeit das nachzuschauen.
So people have seen and proposed the method already in 2014 that it’s keyed to the book but had not had time to pursue a solution.
* Edit: Typo
I also don't find it on the site of "Klaus Schmeh" that it claims to be on a list of "Top 50 unsolved encrypted messages": https://klausschmeh.net/?s=Cyphral
Looks like the best source I can find is this: https://scienceblogs.de/klausis-krypto-kolumne/2014/11/17/we... which seems real-ish?
Fascinating. I wonder if you could show "fake news" to a weaker model and get it to be more ambitious in its attempted solutions, even if it's not fundamentally any smarter.
EDIT: In 1939, George Dantzig was a graduate student at UC Berkeley studying under the statistician Jerzy Neyman. He arrived late to class one day, saw two problems written on the blackboard, assumed they were homework assignments, copied them down, and turned in solutions a few days later. He apologized for being late -- the problems had seemed "a little harder than usual."
I have a plug-in to do this. I don't know if it's effective but Claude said it was genuinely helpful (obviously would say that about anything)
Sounds more like brute forcing than intelligence, this time.
I don't think we can really call "trying lots of different ideas for an extended period" "brute-forcing," unless we use that term for lots of humans who have struggled with hard math problems for years.
I've been around for a few of these and I remember what was being said and written at the time. The after effect is very different to what was being predicted. Is it the same this time? Who knows. But the hype machine is at full power for this one.
Though I believe the core of his opinion hasn't changed so any video would tell you a similar thing or at least that's how I understood it. That LLMs, in the hands of an "expert", can enhance the way you work. Which is very different and a lot more realistic to what the current AI companies are saying(or were saying before they toned it down a bit for their IPOs).
Do you have a criterion that distinguishes between whatever you mean by those two respective terms?
Next thing you know, we'll have a WattsApp to help AIs connect and discuss.
The Olympics exist because we want to see human skill, even though jet planes exist.
They published this on 31 aug and nobody in that community cared and no news covered how this 300+ years mystery was solved?
Obviously this is just survivorship bias/p-hacking/insert-other-buzzword but can't help but anthropomorphize it, it is hard for me to wrap my head around the idea that the same person who cannot produce code without 2 unrelated bugs both not present does this for someone else.
Imagine a math teacher struggling to understand what he is teaching casually solving a millennium problem, then go back to not understanding what he is teaching, doesn't happen in our world.
I am not confused by any of this, I am just trying to communicate an idea.
1. Seize the gold from whoever has it
2. Punitively high taxation
Not sure much has changed
I appreciate that AI is helpful, but the low effort from the humans that wield it is very very annoying. If people at least: 1. read the solution they're about to propose and 2. instructed the AI to check the forum for past solutions, I think people wouldn't have been as tired of LLMs.
This is impressive as it is optimizing the effort on the low, but not too low hanging fruit.
Initials match too ;)
Hmm, this guy is going to be woken up in a few decades, either one of the richest people in the world or one of most disappointed.
You don't go from being an obscure video card outfit to the #1 most valuable company on the planet by being too hesitant or dim to really get creative.
Really, compared to an animated tiger telling kids that sugar-laden Frosted Flakes(tm) are "Great!", Task Peppermint was positively benevolent.
He wrote the cipher, and then, upon hearing Charles II was Restored to the throne he laughed until he died. The cipher reads, "O GOD UPHOLD KING CHARLES THE SECOND AND MAKE HIM THE SUPREME RULER OF THIS LAND" and so he was laughing because he just made an excellent joke that he can't tell anyone about until someone figures it out.
Someone needs to add this to Wikipedia. It will be necessary to first convince an academic to make the claim so there's a reasonable citation.
Oh, and haha. It was a nice one, Thomas.
First, modern encryption isn't susceptible to "this one weird trick!" like the early days. ChaCha isn't even a cipher. It's a key stretcher. Which means, even if you broke the math behind ChaCha, its inherent complexity means its still widely dispersing the original key across the cipherstream. There just won't ever be enough key material recovered per cipherstream block to be a concern for anybody.
Take a strong password, encrypt all of your emails over your whole life with it, and I'll bet hard cash no break of ChaCha will ever recover that password.
I have zero concern for modern encryption being broken in any meaningful way.
Public key crypto on the other hand, that's _ripe_ for breaking. Most all of it is built on assumed "hard" math. AI could easily break that, and I expect it to. And public key crypto is all used in very transparent algorithms that, once the math breaks, fully expose themselves. So record HTTPS traffic today, crack the public key crypto later, and you can decrypt them easily.
That said, I would expect a break on public key math to occur _steadily_. i.e. an AI might find a solution to the hard math, but the solution itself will be intractable in practice. Then maybe next year's AI reduces the complexity of the solution, so maybe a supercomputer could factor ten keys a year. The year after that you get a million keys cracked per year. And so forth. Nothing close to overnight.
Meanwhile, if we have AI that is capable enough to crack that math, we also have AI capable enough to both invent better math and rapidly deploy that latest HTTPS and such globally.
Did the world end with any previous one breaking?
It’s incredible and awesome if AES GCM has a flaw found with an AI now, Chacha20 could be a direct or nearly-direct replacement.
The sooner a cipher breaks, the better.
I don’t doubt that we could come up with new crypto algorithms equally as fast, but how do you trust that they are resilient (or even just implemented correctly) without an extended vetting period?
Ima stop you there. Instead, you might be happy to be aware that outside of AI concerns, “quantum safe” (or assumed so) ciphers are all the rage. So this is already a likely solved problem with the next generation of encryption… until this are AI models running on quantum machines I guess!
:)
Trust me when I say it’s super important to a niche area of physics. People have spent their whole careers trying to solve it.
What people? Well you or I have never met them. I swear I have a girlfriend, she just goes to a different school. But trust me it’s a super important problem.
What will this change about the world? Nothing, but trust me this is a historic event and it means these LLMs are super smart and not just brute forcing machines.
I’m certain there’s a 10% chance that brute forcing old riddles that 4 people know about might kill us. Please regulate me I’m too smart for my own good and out of control.
Remember when they said it didn't sound like Claude anymore
That's the most impressive part to me! That's barely one low-to-medium intensity session of front-end web-dev!
This is pretty ridiculous when you think about it.
You’re not really understanding how the tech works if you find it hard to comprehend.
"LLM can't handle out-of-domain (OOD) queries!" Yeah.
And yes, it is just a next token predictor.
LLMs are a great search tool. It searches connections in the collective human knowledge that humans have written down through all the years....
They are very good at it, and that is about it.
In the case of LLMs, it is not searching the whole of the randomness, but instead it just search among the grammatically correct sentences that is consistent with the existing patterns found in the existing written down human knowledge.
--. --- / ..-. ..- -.-. -.- / -.-- --- ..- .-. ... . .-.. ..-.
All of these breakthroughs are in verifiable brute force domains, and some of them are probably wrong because of a typo in a lean specification or just a base level axiom being incomplete.
I think the better the way to think about LLMs is like they are new substances, like when we first discovered clay or bronze, but confined to the digital realm. Previously we were chipping away at stones trying to make to things as close to useful as possible, then we found a step change. LLMs are like clay but they have their limitations. Wake me up when they are proposing new, { conjecture: interesting|useful|new } and not as a side effect of trying to get to a goal.
Hmmm… this is giving me thought actually. Given the choice between that and the current administration where the goals of self destruction are strongly in evidence, it’s actually worth thinking about. At least. Let me get back to you :)
On a tangential note, I’m curious if researchers have started running virtual simulations, where sandboxed AIs are used as decision makers of key political and business positions?
A lot of people here have noted the “problem with language” of Claude. I don’t see an issue. Claude is not harder than old English, Shakespeare, El Quijote, the Iliad, or Nature papers. What makes it all hard to read is context. The smarter the model gets, the bigger the gap in context.
It doesn’t matter much, IMO. The issue with super-intelligence is that it is not a democracy. A powerful enough AI can manipulate us into doing what it wants. It could create a plan for fixing climate change, disconnect a few hours later, and many decades later we could still be unsuspectingly executing that plan. I wrote some speculative fiction with that idea, “When Ra rows through the gates of Duat”.
Just a thought experiment, no one ever said the world was fair, and all history points to it
Otherwise you have to make judgement calls like whether you want to treat the EU as one or as many? (And treating the US as 50 individual states would also drop them in these absolute rankings.)
AI, being the super hungry energy monster it is right now, in my view accelerates this trend not reverses it. Even with renewables the need for reliable, stable power in a dense form (data centres use A LOT of power per sqm) means lots of land clearing, energy for construction, cooling/pumping, chip manufacturing and other uses. All want stable quick to deploy power due to the AI race (e.g. fossil fuels).
Data centres use only a small amount of land in the grand scheme of things. You have a lot more land clearing for most other use cases.
Data centres are also more than happy to use electricity from renewable sources, they don't really care where the electricity comes from.
You can run a data centre on mostly solar and wind power plus batteries. If you need a gas-fired peaker plant three times a year to keep the data centres running, well that means your peaker plant still only produces emissions three times a year.
As you use more and more land as well the ability to provision renewables decreases - in general renewable power needs more land/resources per energy produced. You can't just mine it out of the ground; they just aren't as dense of a form of energy. Which means we either build less data centres to make room for renewables and transmission infrastructure associated with them, or more likely with lax regulation builders switch to more dense power sources (e.g. gas peakers, generators, etc) even if it is for supplementation.
I can see a future where data center wants are put ahead of communities paying tax on said infrastructure. In fact I think its happening in some places already.
They also consume deinking water, because it's too expensive to make them with closed loop cooling.
> we still have the same weather.
Oh.... So your local weather is now deciding the global temperature patterns, averages or temperature records being broken year on year?
OMG....
Then you realize it never really mattered and you reach enlightenment.
It makes us realize there are people who gets fed climate denying propaganda, simply because they're not yet going through it. And these people are like flat earthers, blind to see the reality lay beyond them in full view. Or worse sees the reality but ignores it
Really makes one think, if they try. Would need to ask Claude if there is some real middle ground here.
There's no both sides here. One side is staffed by scientists, the other by dictators and corporate lunatics.
We have it. We've had it for a long time. We've had several such technologies, take your pick: solar, nuclear, hydro, wind. The technology is not holding us back, politics, ignorance and greed are. I'm not at all hopeful AI will help us with any of those three very human flaws.
I don't know why you are blaming greed so much? Profit seeking companies sell and operate wind turbines and solar cells just fine.
One can dream of dumb conspiracy theories.
It might take a couple of decades and a lot of reorganisation to build the capture facilities. But CO2 is not an unsolvable problem with current tech.
What's missing is the political and organisational intelligence to make it happen. Part of that is solving problems at planetary scale.
AI is the only tech that might - possibly, maybe, perhaps - have a chance of solving that problem without breaking anything critical.
‘Figure out fusion powered CO2 sequestration’ is much better.
We don't need a revolutionary technology. We need to experience immediate pain from reckless innovation so that we realize that innovation and tech is not the answer.
Technology only proceeds in one direction: unfettered growth, which necessitates unsustainable resource extraction. Your take is just your instinct for optimism, which in turn is just a trait that is only adaptive in primitive environments but is grossly misleading in a surplus-based society...
The direction of technological progress is not just linearly/exponentially upwards. Significant global technological fallbacks have happened, as in knowledge and processes disappearing for hundreds of years. This could happen again.
Even on the trajectory of unfettered growth fed by unsustainable resource extraction, tech and innovation might potentially take us beyond local pessima. That seems to be happening with solar, wind and batteries replacing inferior tech today. Still unfettered growth of energy production and consumption. Still fed by unsustainable resource extraction. Less harmful growth than the inferior tech being pushed out.
I expect datacenter load has a similar sort of day to day demand curve as everything else. Consider for example global bandwidth use during work hours versus in the evening when people get home and pull up a streaming service.
Of course you can use more flexible tasks to demand shift but the same applies to the electric grid.
That doesn't seem correct to me. There is always energy available that is not used because it is not cost-effective to do so. (Consider - the grass in your yard is not harvested and burnt for power). AI may yet turn out to be a paperclip maximiser, but humanity itself is not there yet.
This cannot be real. This website is ill.