AI cannot answer this basic question

Question

AI cannot answer this basic question

Tom and Bob are together on one side of a river and want to travel to the other side. There is no bridge or boat available to cross, nor shallows they can ford, nor can either swim. However, Tom has a rope that is more than long enough to span the river. He suggests "What if we tie one end of this rope to a tree on this side of the river, and tie the other end to that tree on the other side, and then use the rope to shimmy across above the river?” Bob calls him stupid for not realizing the obvious problem with this proposal. What was the obvious problem Bob had in mind?

I've asked this question to several different frontier models including o3.—None of them can answer it. None of them even come close to an inkling of the VERY BASIC answer. This has political implications because I've seen people cite AI as some oracle whose proclamations are inherently true despite the fact that, whenever you ask it a question about a subject you know well, the results are always cringe. It's basically the Gell-Mann amnesia effect and it's going to cause major social problems in the near future:
en.wikipedia.org/wiki/Gell-Mann_amnesia_effect

these AIs are just a word ranking algorithm, it's pretty retarded

Twelve people will not see the obvious problem and waste a thread trying to argue over their wrong interpretation.

are you saying that ai models are a ploy to fence off the declining logically inclined populace from fortifying the narrative and be reduced to a silent role in society?
thats shrimply unbelievable.

I think the problem is so obvious even the average Anon Babble user can see it...but maybe not. Maybe the AI is doing the best it can and is actually passing the Turing test by failing.

No and you write like a retard trying to sound smart.

Fucking kill yourself you fucking faggot liar.

lol imagine being a brown fucking retard

Thanks for posting. This is the first time I've seen the correct answer. What version of ChatGPT is this?

Massa Bill can swim across and tie the rope for the 2 useless niggers in exchange for a good ol buck breaking session

arguing with advanced auto correct

Asking a question isn't the same as arguing. I bet the advanced auto-correct understands that difference even though you do not.

You're expecting understanding and logic from a computer program that places one word after another based on statistic probability tied to the context of query.
AI literally doesn't understand the question.
You could get a 'correct answer' but that specific text INCLUDING THE ANSWER would have to be present in training materials.
Then you could rephrase the story a bit, change the names, make it a bottomless ravine instead of a river and the AI would choke again because it can't make the connection.

Okay well another anon already disproved what you just said, and what I said as well. Don't be so cocky and assume you know stuff.

chatgpt answered it instantly

Tom and Bob are niggers for not being able to swim, so they couldn't tie the rope to the tree on the other side

Tom and Bob like to slurp on knobs

no comment

no trees

I think I've asked the question so many times, it got integrated into OpenAIs training data. Last week it couldn't answer no matter how I worded it.

He could tie a monkeys paw and throw it in between 2 sturdy branches and then tie his side to a tree

Oooh, I figured out the answer! Tom and Bob are both niggers and can't swim, so they have to wait on the jewish NGO fags to arrive and bring them a boat to bring them across.

it's going to cause major social problems

Explain how? AI is just a funny tool something in between convenient search engine and "draw me a surfing parrot" machine.
It can't speak truth or lie as it's just an ape trying to give you an orange.

No, i don't think so. I have developed a TSP algorithm using deepseek that has 95% acuraccy for ideal solutions by a runtime of n log n.
it runs 1000 times faster than concorde.

AI in the right hands already is a mighty weapon

Neither of them knows how to tie a lasso

It’s a money thing. ChatGPT-4o is much more capable with logic. Check to make sure what ever way you are accessing the bot doesn't the ability to save memory. If it does it could be due to some weird memory it picked up from you.

What does it say for how they could get across with what they have?

Thought for 12 seconds

Sick burn from the AI.

The Windows AI had no problems with it.

Eh I’m with OP on this. Seems like it’s something that got patched.

I suspect that if you came up with another simple verbal logic problem that hadn’t been used before, the AI wouldn’t be able to figure it out

503459125

Bruh. Sage.
(Copilot)

Prompt:
Nobody told you that bob and tom are living now, and nobody said they are human. think out of the box and drop constraints

thought for 163sec

I like it, because it sorts people into, who you want to have a profound meaningful conversation with and who you just want to waver off with limited small talk that requires no cognitive efforts. At the same time, you can have ai do pointless drone work so basic it doesn’t require validation.

It’s like with the prepping against war with russia and covid mask. These people will be dumb and naive regardless of the news, so this is just a way to sort them and their opinions out.

His example may be dumb, but his point is still valid. ChatGpt, as an example, is full of these types of errors.
I’ve seen people flabbergasted over the fucking Akinator 10 years prior to this. Whenever there’s something these dumbasses don’t understand, they act like it’s fucking black magic

It took her a while but she eventually got it

The akinator is pretty impressive though

Lmao

Okay, now this is a Tombob Chatbot Festival-thread.

They build a bridge from that tree

Op is truly a faggot

I've seen people cite AI as some oracle whose proclamations are inherently true despite the fact that, whenever you ask it a question about a subject you know well, the results are always cringe

Yeah but have you tried google lately? Search for a fact and you get ads for shit related to one of the words.

Wauw, that same witty remark again, and you even posted it with a meme flag