20
submitted 1 month ago by [email protected] to c/[email protected]

They train on sneer-problems now:

Here’s the “ferry‑shuttle” strategy, exactly analogous to the classic two‑ferryman/many‑boats puzzle, but with planes and pilots

And lo and behold, singularity - it can solve variants that no human can solve:

https://chatgpt.com/share/68813f81-1e6c-8004-ab95-5bafc531a969

Two ferrymen and three boats are on the left bank of a river. Each boat holds exactly one man. How can they get both men and all three boats to the right bank?

top 24 comments
sorted by: hot top new old
[-] [email protected] 20 points 1 month ago

If there’s only one ferryman, the puzzle becomes trivial (he just shuttles each boat one by one:

Does he now?

[-] [email protected] 18 points 1 month ago

The rules did not mention the ferryman couldn't use the stargate on each bank. The rules exclude teleporting, vut a stargate works via a different method.

[-] [email protected] 7 points 1 month ago

no-one said he couldn’t be yoda

[-] [email protected] 14 points 1 month ago

Two ferrymen and three boats are on the left bank of a river. Each boat holds exactly one man. How can they get both men and all three boats to the right bank?

Officially, you can't. Unofficially, just have one of the ferrymen tow a boat.

[-] [email protected] 15 points 1 month ago* (last edited 1 month ago)

Yeah that's the version of the problem that chatgpt itself produced, with no towing etc.

I just find it funny that they would train on some sneer problem like this, to the point of making their chatbot look even more stupid. A "300 billion dollar" business, reacting to being made fun of by a very small number of people.

[-] [email protected] 12 points 1 month ago* (last edited 1 month ago)

Some insufferable dipshit that gets paid $10,000 bucks a week spent one of those weeks working on this.

[-] [email protected] 8 points 1 month ago

they had a meeting about it, the meeting lasted an hour and had 8 people, and they collectively earned $1 million in that time

[-] [email protected] 8 points 1 month ago

Gonna be really fucked if we finally discover that the people doing the actual work were true believers who didnt even get paid well. And all the money went to altman/infrastructure costs. (I could look in financial statements to figure stuff like this out but erug effort).

[-] [email protected] 13 points 1 month ago

There's an "I am no man" joke in here somewhere that I am too tired to figure out.

[-] [email protected] 14 points 1 month ago* (last edited 1 month ago)

Ferryman 1 calls to Gwaihir, the Lord of Eagles for aid, and The Windlord answers to fly him back across.

[-] [email protected] 4 points 1 month ago

ferrymen ⊄ men?

[-] [email protected] 9 points 1 month ago

Officially, you can’t. Unofficially, just have one of the ferrymen tow a boat.

Or swim back. However, the bot itself appears to have ruled out all of these options.

[-] [email protected] 13 points 1 month ago

At first glance it seems impossible once N≥2, because as soon as you bring a boat across to the right bank, one of you must pilot a boat back—leaving a boat behind on the wrong side.

In this sentence, the bot appears to sort of "get" it (not entirely, though, the wording is weird). However, from there, it definitely goes downhill...

[-] [email protected] 13 points 1 month ago

The downhill is honestly glorious because it seems so proud of itself when the real magic is that the boatmen can magically teleport back to the right bank under certain arcane circumstances.

[-] [email protected] 12 points 1 month ago

Somehow, the "smug" tone really rubs me the wrong way. It is of great comedic value here, but it always reminds me of that one person who is consistently wrong yet is somehow the boss's or the teacher's favorite.

[-] [email protected] 10 points 1 month ago

This is obviously a math olympiad gold medal performance, Fields medal worthy even!

[-] [email protected] 8 points 1 month ago* (last edited 1 month ago)

Gonna take a wild guess that enough people have tried this joke to the point where they've added custom code or lines in the system prompt about it:

prompt and outpoo

[-] [email protected] 10 points 1 month ago

Now we need to make a logic puzzle involving two people and one cup. Perhaps they are trying to share a drink equitably. Each time they drink one third of remaining cup’s volume.

[-] [email protected] 5 points 1 month ago

Now we need to make a logic puzzle involving two people and one cup. Perhaps they are trying to share a drink equitably. Each time they drink one third of remaining cup’s volume.

Step one: Drink two-thirds of the cup's volume

Step two: Piss one sixth of the cup's volume

Problem solved

[-] [email protected] 5 points 1 month ago* (last edited 1 month ago)

Hmm, maybe too premature - chatgpt has history on by default now, so maybe that's where it got the idea it was a classic puzzle?

With history off, it still sounds like it has the problem in the training dataset, but it is much more bizarre:

https://markdownpastebin.com/?id=68b58bd1c4154789a493df964b3618f1

Could also be randomness.

Select snippet:

Example 1: N = 2 boats

Both ferrymen row their two boats across (time = D/v = 1/3 h). One ferryman (say A) swims back alone to the west bank (time = D/u = 1 h). That same ferryman (A) now rows the second boat back across (time = 1/3 h). Meanwhile, the other ferryman (B) has just been waiting on the east bank—but now both are on the east side, and both boats are there.

Total time

$$ T_2 ;=; \frac{1}{3} ;+; 1 ;+; \frac{1}{3} ;=; \frac{5}{3}\ \mathrm{hours} \approx 1,\mathrm{h},40,\mathrm{min}. $$

I have to say with history off it sounds like an even more ambitious moron. I think their history thing may be sort of freezing bot behavior in time, because the bot sees a lot of past outputs by itself, and in the past it was a lot less into shitting LaTeX all over the place when doing a puzzle.

[-] [email protected] 4 points 1 month ago

Can someone explain the origin of this problem for someone ootl?

[-] [email protected] 9 points 1 month ago

River crossing puzzles are a genre of logic problems that go back to the olden days. AI slop bots can act like they can solve them, because many solutions appear in their training data. But push the bot a little harder, and funny things happen.

[-] [email protected] 5 points 1 month ago

That makes sense. So it's just an obviously bullshit river crossing puzzle that the chatbot is calling a classic problem.

[-] [email protected] 5 points 1 month ago

Pretty much. Our friend up top (diz/OP) has made a slight hobby of poking the latest and greatest LLM releases with variants of these puzzles to try and explore the limitations of LLM "cognition".

this post was submitted on 23 Jul 2025
20 points (100.0% liked)

SneerClub

1190 readers
132 users here now

Hurling ordure at the TREACLES, especially those closely related to LessWrong.

AI-Industrial-Complex grift is fine as long as it sufficiently relates to the AI doom from the TREACLES. (Though TechTakes may be more suitable.)

This is sneer club, not debate club. Unless it's amusing debate.

[Especially don't debate the race scientists, if any sneak in - we ban and delete them as unsuitable for the server.]

See our twin at Reddit

founded 2 years ago
MODERATORS