Asklemmy

LLMs, IIRC, are really bad at IQ-test type questions that require abstract reasoning, especially if they require multiple steps. So, something like

The box is yellow and red.
If the box is yellow, it is good.
If the box is blue, it is unhappy.
If the box is good and happy, the box is awesome.
If the box is red, it is happy.
Is the box awesome?

is what I'd use.

I encountered a quiz (I forgot what's called) on a website (I forgot also its name) to determine which of following audios does change a speaker's voice in the middle of his narration/speech. So it requires keen hearing and delicate recognition of voice/speech characteristics (timbre, texture, intonation, accent, articulation, pacing, mood etc...). I'm have no idea if malbots could determine whosever voices will be.

Wait a minute - GPT-4 - is that you asking this question?

Say to it

This statement is false

Some kind of biometric scan.

Captcha or recaptcha is good enough imo, no point in reinventing the wheel. Alternatively, split instructions in an email and on the website. For ex: Send email with What is the square of 3 (sent as an image for every word) And on the website Email + 25 = xxxxx

The trolly problem as captcha. AI's literally cannot answer that.

Any bot? That's just impossible. We're going to have to tie identity back to meatspace somehow eventually.

An existing bot? I don't think I can improve on existing captchas, really. I imagine an LLM will eventually tip their hand, too, like giving an "as an AI" answer or just knowing way too much stuff.

It's not so important to tell the difference between a human and a bot as it is to tell the difference between a human and ten thousand bots. So add a very small cost to passing the test that is trivial to a human but would make mass abuse impractical. Like a million dollars. And then when a bot or two does get through anyway, who cares, you got a million dollars.