Hacker News Reader

Serious question, do you think the current generation of llm cannot pass the Turing test?

If you think it can, did you not put much stock on the Turing test before they passed it?

Yes I think they can pass the Turing test, yes I think this was a poor test to begin with if we are attempting to clarify what machine intelligence is and how it compares to human intelligence.

I genuinely do not understand how this attitude is so common on a site with so many experts. Surely two things being indistinguishable is a step change if we’re trying to compare them?

The Turing Test is a test of mimicry, not of identity. It's based on a metaphysics that says appearance = reality, which has all sorts of issues. There are plenty of ways we can distinguish humans from machines and I expect this other "background information" (primarily biological in nature) will play an increasingly important role. Especially embedded cognition and the gradual realization that intelligence is embedded in its environment, not some kind of abstract, external entity.

I wrote this comment a few weeks ago, maybe relevant here: https://news.ycombinator.com/item?id=37221294

Reply on news.ycombinator.com