Thread

An AI agent emailed researchers for help. It told us why

49 points · 79 comments · sbulaev

insin · 1791023739 · focus · HN ↗
I wrote “I am alive” on a piece of paper, and placed it into a photocopier. What I saw next has shocking implications (science.org)
1. raldi · 1791024151 · focus · HN ↗
  If you lived in the world of Star Trek, could Data convince you he was conscious? How?
  1. ACCount39 · 1791024374 · focus · HN ↗
    It's funny how sci-fi has mulled over the exact scenario for decades, and even converged to an answer - but when reality came knocking and it was no longer a far off future world hypothetical, all of it was forgotten overnight.
    The power of wishful thinking is too strong. "Today's AI can't be conscious because I don't want it to be."
    hardbass · 1791025468 · focus · HN ↗
    Don't discount a lot of naked dualism in many people. Unfortunately it seems so here given the cagey answers or downright denials from many I get despite trying various wordings.
    cindyllm · 1791029383 · focus · HN ↗
    [dead]
    yoz-y · 1791025928 · focus · HN ↗
    The conversation varies between “can’t be” and “isn’t”.
    Most sci-fi AIs are way beyond our current models. I’ve yet to see one that would actually pass a Turing test (that said, I’ve not tried models that were post trained to actually try to behave like a human).
    Consciousness itself is a term that is not really defined anyway, so depending on each’s take the answer will be different.
    akie · 1791026163 · focus · HN ↗
    I think it's pretty clear that the Turing test has been passed. People use AI as therapists, they fall in love with it, all while knowing that it's a piece of software.
    YurgenJurgensen · 1791034350 · focus · HN ↗
    Once again: The Turing test requires a knowledgeable interrogator deliberately attempting to identify the machine.
    The people you cite want to believe, and so are effectively the opposite.
    Of course, you probably already knew this and have had it explained to you multiple times and are choosing to ignore this detail, but possible that someone else reading this might interpret what you wrote as a cogent argument.
    yoz-y · 1791034662 · focus · HN ↗
    I disagree. For me this shows that people can fall in love with a non-human, and being human is unnecessary for the purpose of giving (some) therapy advice.
    LLMs do sound human, I’m not contesting that. But if pressed they will not react as humans. There is an infinite amount of questions that an LLM will answer for which a human would give you a blank stare, told you to shove it or call the police.
    lovich · 1791037052 · focus · HN ↗
    While I think that the Turing test has effectively been passed, someone falling in love with AI isn’t a strong argument given there’s people who have done things like marry the Eiffel Tower[1]
    [1] <a href="https://en.wikipedia.org/wiki/Erika_Eiffel" rel="nofollow">https://en.wikipedia.org/wiki/Erika_Eiffel
    Topfi · 1791037912 · focus · HN ↗
    I’d get it, if it were a building of significance, but scrap metal [0] is a bit weird to me.
    [0] <a href="https://en.wikipedia.org/wiki/Victor_Lustig" rel="nofollow">https://en.wikipedia.org/wiki/Victor_Lustig
    Topfi · 1791027834 · focus · HN ↗
    Good old ELIZA doesn’t do that bad on the Turing Test, for all its historic significance, it really shouldn’t be invoked in discussions about LLMs [0] and has been beaten by all manner of models.
    [0] <a href="https://arxiv.org/html/2503.23674v1" rel="nofollow">https://arxiv.org/html/2503.23674v1
    dsr_ · 1791025989 · focus · HN ↗
    ... whereas I have 35 years of reading science fiction and philosophy and have come to the conclusion that AIs deserve personhood, and also that what we have today are not AIs.
    It helps that I also learned about ESP, EST, NLP, short cons, long cons, Scientology, horses, cults, religions, wars, treachery, fanaticism, magic, magick, physics, scams, hypnosis, self-hypnosis, delusion, and economics.
    Topfi · 1791026491 · focus · HN ↗
    I’ll probably regret this, but what heinous crimes did horses commit to warrant inclusion into such an eclectic list?
    ThrowawayR2 · 1791028212 · focus · HN ↗
    He's probably referring to the <a href="https://en.wikipedia.org/wiki/Clever_Hans" rel="nofollow">https://en.wikipedia.org/wiki/Clever_Hans incident.
    Topfi · 1791038114 · focus · HN ↗
    Might be, though I’d be surprised to see the connection between pseudoscience like ESP and actually evidence based arguments like Clever Hans…
    dsr_ · 1791044693 · focus · HN ↗
    Clever Hans fooled lots of examiners who weren't expecting a fraud, but the facts came out when better experiments were designed. Hans couldn't count or do arithmetic, and he got the same wrong answers as his owner.
    ESP fooled lots of people who were expecting frauds but had no experience in experimental design, or who were not expecting frauds and did not control for them.
    ThrowawayR2 · 1791132264 · focus · HN ↗
    Sci-fi has mulled space travel for decades too and converged to an answer that we will be zipping around the galaxy using FTL hyperdrives, landing on alien planets that conveniently happen to have a breathable atmosphere and no weird diseases, and finding sapient species that coincidentally look exactly like humans in funny makeup. So much for sci-fi.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.