frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Ask HN: Is "make me laugh" the next Turing Test?

2•ryandvm•2h ago
It seems that in the last few years the Turing Test has rapidly fallen off as a useful experiment for testing artificial intelligence. With the right guardrails, the latest frontier LLMs are more than capable of simulating perfectly believable human conversation partners.

Instead, perhaps the next best practical test for separating humans from AIs is essentially, "can it make you laugh?"

The idea here is that LLMs are largely trained to be able to produce statistically likely sentences based on a massive training corpus, and while this allows for incredibly impressive query responses, it is absolutely atrocious at statistically unlikely outcomes, that is, humor.

Humor is often defined by being mentally incongruous which would necessarily make it statistically unlikely - the very opposite of what LLMs are good at.

Thoughts?

Comments

subject4056•1h ago
I expect lots of humans to not be able to make you laugh on command, so no it's probably not a good test
PaulHoule•1h ago
Yeah, I was on the outside looking in when it came to sketch comedy until I developed a new character that can make people laugh just with hand gestures. It's really funny how you go nowhere telling other people's jokes and you really need to write your own material.
dhruv3006•1h ago
Well have you seen the hallucinations or should I say " confabulations ".
ksaj•21m ago
I've found AI, or at least ChatGPT, to be really good at making great jokes about very specific topics.

You always have to ask it for several, because a lot of them are rubbish. But the hit rate is reasonable, considering the difficulty in understanding what is funny or not to a human.

Also, if you like absurdist humour, then even the fails might work. If I could post a pic here, I would. I once asked ChatGPT for a strip comic about something. What it produced didn't make sense, but I laughed about it every time I read it. Uploaded it and got the same result from others. There was a really polite naivety in what it produced, but it was surprisingly potent.

I'm pretty sure the llm could not explain what about it was funny. Basically a vegetable accidentally killed another vegetable simply by touching it. But then it apologized. That apology was bloody hilarious.

The prompt was for a comic strip with a hook that as far as it knows, nobody has ever seen before. And the result was shockingly great.

And of course it was absolute luck, because the same prompt just spat out junk thereafter.

Glitches of Street Fighter II [video]

https://www.youtube.com/watch?v=HtYQqGPL_NM
1•CHB0403085482•1m ago•0 comments

There's Plenty of Room at the Bottom: Invitation to Enter a New Field of Physics [pdf]

https://www.cs.jhu.edu/~basu/Papers/Feynman59.pdf
1•helloplanets•2m ago•0 comments

Show HN: Idea Forge – Multi-model product validation(validated an OpenClaw idea)

https://ideas.sparkngine.com/
1•oliverchoy•3m ago•0 comments

Show HN: I taught GPT-OSS-120B to see using Google Lens and OpenCV

1•vkaufmann•6m ago•0 comments

What color are your bits? (2004)

https://ansuz.sooke.bc.ca/entry/23
1•tomodachi94•9m ago•0 comments

Simple email API for proton bridge

https://github.com/pgray/mayl
2•pgray•13m ago•0 comments

Stay Hungry, Stay Foolish (2005)

https://stevejobsarchive.com/stories/stay-hungry-stay-foolish
2•walterbell•14m ago•0 comments

High-deductible health plans associated with 46% worse mortality

https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2844474
4•brandonb•15m ago•1 comments

Logic That Patterns Find

https://udara.io/logic-and-noise/
1•udara•15m ago•0 comments

Show HN: I made myself a no-nonsense TODO app

https://todo-app-light.pages.dev/
1•anjandutta•16m ago•0 comments

Functional Programming in an LLM World

https://notes.druchan.com/fp-in-llm-world
1•druchan89•16m ago•0 comments

I completed the MV in one day. Below is a detailed breakdown of how it was done

https://www.youtube.com/watch?v=rxWNmzQpW2c
1•luokai•19m ago•1 comments

Show HN: "hard questions" as a shared language for cross-domain reasoning

https://github.com/onestardao/WFGY
2•wfgy-github•20m ago•0 comments

When people think about aliens, they make assumptions which are wrong

https://www.mindlessalgorithm.com/alien-assumptions/
2•helloplanets•21m ago•0 comments

Brandon Wint: How Do We Know When A Poem Is Finished?

https://www.youtube.com/watch?v=k9hPmp09ssc
1•marysminefnuf•25m ago•0 comments

GraphQLite, Graph Network Extension on Top of SQLite

https://github.com/colliery-io/graphqlite
2•ekianjo•28m ago•0 comments

Show HN: Clawhosting.io– Managed OpenClaw

https://clawhosting.io
1•rezaghp•30m ago•0 comments

Modular Buys BentoML

https://www.modular.com/blog/bentoml-joins-modular
1•carefree-bob•33m ago•0 comments

Show HN: Hosting dynamic webcal on GitHub pages

https://github.com/sfw185/BJJCal
1•faridw•33m ago•0 comments

Why capital is fleeing Tech for the Tangible Economy

https://wwai.substack.com/p/the-great-rotation-capital-flees
4•watchwiseai•45m ago•1 comments

Harmless reward hacks generalize to shutdown evasion and dictatorship in GPT-4.1

https://arxiv.org/abs/2508.17511
1•toliveistobuild•47m ago•1 comments

Exponential Code, Network Effects in AI, & the Return of Apprenticeships

https://www.implications.com/p/exponential-code-network-effects
1•swolpers•50m ago•0 comments

Jonathan Blow on Why Modern Software Is Bloated

https://www.youtube.com/watch?v=GOtWR_T2VOk
3•cable2600•56m ago•0 comments

Venezuelan crude oil exports to Israel resume after 6-year gap

https://www.bloomberg.com/news/articles/2026-02-10/venezuela-sending-first-crude-oil-cargo-to-isr...
2•OgsyedIE•59m ago•0 comments

Secretary of War Hegseth Announces End of Support for Harvard University [video]

https://www.youtube.com/watch?v=eh5duiL3MwQ
2•nomilk•1h ago•1 comments

Vibrant Frog Collab – AI-Powered Writing Assistant

https://frogteam.ai/VibrantFrog/default.html
1•am-piazza•1h ago•0 comments

Show HN: Track and analyze AI coding tool usage across your team

https://trackr-bay.vercel.app/welcome
1•usmansidd•1h ago•0 comments

(Rust) Tracking Issue for Generic Constant Arguments MVP

https://github.com/rust-lang/rust/issues/132980
1•anfilt•1h ago•1 comments

How many of the 3,191 billionaires can you name?

https://billionaires.linolevan.com/
1•linolevan•1h ago•1 comments

mrdoob Ported Quake to JavaScript/Three.js

https://mrdoob.github.io/three-quake/
2•davidbarker•1h ago•0 comments