I’d just write the list then assign randomly. Or perhaps pseudorandomly like sort by hash and then split in two.
One problem is that it is hard to come up with 20 or more completely unrelated puzzles.
Although I don’t think we need a large number for statistical significance here, if it’s like 8/10 solved in the cheating set and 2/10 in the hold back set.
Yeah any time its regurgitating an IMO problem it’s a proof it’salmost superhuman, but any time it actually faces a puzzle with unknown answer, this is not what it is for.
Further support for the memorization claim: I posted examples of novel river crossing puzzles where LLMs completely fail (on this forum).
Note that Apple’s actors / agents river crossing is a well known “jealous husbands” variant, which you can ask a chatbot to explain to you. It gladly explains, even as it can’t follow its own explanation (since of course it isn’t its own explanation but a plagiarized one, even if changes words).
edit: https://awful.systems/post/4027490 and earlier https://awful.systems/post/1769506
I think what I need to do is to write up a bunch of puzzles, assign them randomly to 2 sets, and test & post one set, while holding back on the second set (not even testing it on any online chatbots). Then in a year or two see how much the set that’s public improves, vs the one that’s held back.
making LLMs not say racist shit
That is so 2024. The new big thing is making LLMs say racist shit.
Can’t be assed to read the bs but sometimes the use after free only happens in some rarely executed code path, or only when one branch is executed then later another branch. So you still may need fuzzing to trigger use after free for Valgrind to detect.
I swear I’m gonna plug an LLM into a rather traditional solver I’m writing. I may tuck deep into the paper a point how it’s quite slow to use an LLM to mutate solutions in a genetic algorithm or a swarm solver. And in any case non LLM would be default.
Normally I wouldn’t sink that low but I got mouths to feed, and frankly, fuck it, they can persist in this madness for much longer than I can stay solvent.
This is as if there was a mass delusion that a pseudorandom number generator can serve as an oracle, predicting the future. Doing any kind of Monte Carlo simulation of something like weather in that world would of course confirm all the dumb shit.
Yeah plenty of opportunities to just work it into the story.
I dunno what kind of local models you can use, though. If it is a 3D game then its fine to require a GPU, but you wouldn’t want to raise minimum requirements too high. And you wouldn’t want to use 12 gigs of vram for a gimmick, either.
I think it could work as a minor gimmick, like terminal hacking minigame in fallout. You have to convince the LLM to tell you the password, or you get to talk to a demented robot whose brain was fried by radiation exposure, or the like. Relatively inconsequential stuff like being able to talk your way through or just shoot your way through.
Unfortunately this shit is too slow and too huge to embed a local copy of, into a game. You need a lot of hardware compatibility. And running it in the cloud would cost too much.
I was trying out free github copilot to see what the buzz is all about:
It doesn’t even know its own settings. This one little useful thing that isn’t plagiarism, providing natural language interface to its own bloody settings, it couldn’t do.
All joking aside, there is something thoroughly fucked up about this.
What’s fucked up is that we let these rich fucks threaten us with extinction to boost their stock prices.
Imagine if some cold fusion scammer was permitted to gleefully boast that his experimental cold fusion plant in the middle of a major city could blow it up. Setting up little hydrogen explosions, setting up a neutron source just to make it spicier, etc.
When confronted with a problem like “your search engine imagined a case and cited it”, the next step is to wonder what else it might be making up, not to just quickly slap a bit of tape over the obvious immediate problem and declare everything to be great.
Exactly. Even if you ensure the cited cases or articles are real it will misrepresent what said articles say.
Fundamentally it is just blah blah blah ing until the point comes when a citation would be likely to appear, then it blah blah blahs the citation based on the preceding text that it just made up. It plain should not be producing real citations. That it can produce real citations is deeply at odds with it being able to pretend at reasoning, for example.
Ensuring the citation is real, RAG-ing the articles in there, having AI rewrite drafts, none of these hacks do anything to address any of the underlying problems.
Yeah. I’d love to see the prompt, gab’s nazi ai prompt was utterly pathetic and this one got to be pretty bad as well.
Actually, having read it carefully, it is interesting that they actually don’t claim it was hacked, they claim that the modification was unauthorized. They also don’t claim that they removed the access from that mysterious “employee” who modified it. I’m thinking they had some legal reason to technically not lie.
It re consumes its own bullshit, and the bullshit it does print is the bullshit it also fed itself, its not lying about that. Of course, it is also always re consuming the initial prompt too so the end bullshit isn’t necessarily quite as far removed from the question as the length would indicate.
Where it gets deceptive is when it knows an answer to the problem, but it constructs some bullshit for the purpose of making you believe that it solved the problem on its own. The only way to tell the difference is to ask it something simpler that it doesn’t know the answer to, and watch it bullshit in circles or to an incorrect answer.
I think they worked specifically on cheating the benchmarks, though. As well as popular puzzles like pre existing variants of the river crossing - it is a very large puzzle category, very popular, if the river crossing puzzle is not on the list I don’t know what would be.
Keep in mind that they are also true believers, too - they think that if they cram enough little pieces of logical reasoning, taken from puzzles, into the AI, then they will get robot god that will actually start coming up with new shit.
I very much doubt that there’s some general reasoning performance improvement that results in these older puzzle variants getting solved, while new ones that aren’t particularly more difficult, fail.
Did you use any of that kind of notation in the prompt? Or did some poor squadron of task workers write out a few thousand examples of this notation for river crossing problems in an attempt to give it an internal structure?
I didn’t use any notation in the prompt, but gemini 2.5 pro seem to always represent state of the problem after every step in some way. When asked if it does anything with it says it is “very important”, so it may be that there’s some huge invisible prompt that says its very important to do this.
It also mentioned N cannibals and M missionaries.
My theory is that they wrote a bunch of little scripts that generate puzzles and solutions in that format. Since river crossing is one of the top most popular puzzles, it would be on the list (and N cannibals M missionaries is easy to generate variants of), although their main focus would have been the puzzles in the benchmarks that they are trying to cheat.
edit: here’s one of the logs:
Basically it keeps on trying to brute force the problem. It gets first 2 moves correct, but in a stopped clock style manner - if there’s 2 people and 1 boat they both take the boat, if there’s 2 people and >=2 boats, then each of them takes a boat.
It keeps doing the same shit until eventually its state tracking fails, or its reading of the state fails, and then it outputs the failure as a solution. Sometimes it deems it impossible:
All tests done with gemini 2.5 pro, I can post links if you need them but links don’t include their “thinking” log and I also suspect that if >N people come through a link they just look at it. Nobody really shares botshit unless its funny or stupid. A lot of people independently asking the same problem, that would often happen if there’s a new homework question so they can’t use that as a signal so easily.
Yeah I think the best examples are everyday problems that people solve all the time but don’t explicitly write out solutions step by step for, or not in the puzzle-answer form.
It’s not even a novel problem at all, I’m sure there’s even a plenty of descriptions of solutions to it as part of stories and such. Just not as “logical puzzles” due to triviality.
What really annoys me is when they claim high performance on benchmarks consisting of fairly difficult problems. This is basically fraud, since they know full well it is still entirely “knowledge” reliant, and even take steps to augment it with generated problems and solutions.
I guess the big sell is that it could use bits and pieces of logic gleaned from other solutions to solve a “new” problem. Except it can not.
And it is Google we’re talking about, lol. If no one uses their AI shit they just replace something people use with it (also see search).
It’s google though, if nobody uses their shit they just put it inside their search.
It’s only gonna go away when they run out of cash.
edit: whoops replied to the wrong comment
It would have to be more than just river crossings, yeah.
Although I’m also dubious that their LLM is good enough for universal river crossing puzzle solving using a tool. It’s not that simple, the constraints have to be translated into the format that the tool understands, and the answer translated back. I got told that o3 solves my river crossing variant but the chat log they gave had incorrect code being run and then a correct answer magically appearing, so I think it wasn’t anything quite as general as that.