Thread

ChatGPT now knows what you do on other websites via ad collector

764 points · 395 comments · lmbbuchodi

mavsman · 1789923911 · focus · HN ↗
To me, this quote just about sums it up:
> The mechanism is standard adtech. What has no precedent is running it on an AI chat product.
As someone who has been well aware of this mechanism for quite some time, I still feel icky anytime I re-read the details of it.
What a time to be alive.
1. jsrozner · 1789925147 · focus · HN ↗
  People think they're interacting with an "intelligence," when actually they're just getting a maximally optimized Weizenbaum feed. We're living through the sloppification of the human mind.
  See e.g., <a href="https://www.science.org/content/article/ai-chatbots-are-becoming-experts-changing-people-s-minds-what-s-their-secret" rel="nofollow">https://www.science.org/content/article/ai-chatbots-are-beco...
  1. brianleb · 1789932513 · focus · HN ↗
    Could you define (or link) to what "Weizenbaum feed" means/references? Too obscure for me but I'm interested. Thanks.
    fweimer · 1789932889 · focus · HN ↗
    Likely a reference to ELIZA: <a href="https://en.wikipedia.org/wiki/ELIZA" rel="nofollow">https://en.wikipedia.org/wiki/ELIZA
    handoflixue · 1789938211 · focus · HN ↗
    Are there really people that can't tell the difference between ELIZA and something that can solve open Millennium Prize Problems in math?
    MisterMunchkin · 1789942986 · focus · HN ↗
    They had to solve the car wash problem by hardcoding the solution in. The AI equivalent of adding another IF statement.
    They had to steal the work of researches solving these open problems and then rewrite their solution. The AI equivalent of fraud.
    nearbuy · 1789966467 · focus · HN ↗
    They did not hardcode that, because:
    a) There's zero evidence of them doing so
    b) Some models released before the car wash problem was discovered would consistently get it right
    c) Hardcoding it is pointless. No one is seriously asking that. It's just a trick question. Hardcoding one trick question won't fix its weakness at other simple trick questions.
    d) Since it went viral on the internet, the next time they updated the knowledge cutoff, the LLM would likely be aware of the trick. It will fix itself without the labs doing anything special, even assuming the new models weren't smart enough to naturally figure it out.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.