Hacker News Reader

Hacker News

Comment by Animats | Hacker News Reader

Full thread

Animats

Cyc is going great, according to the web site. "The Next Generation of Enterprise AI"[1]

Lenat himself died in 2023. Despite this, he is listed as the only member of the "leadership team".[2]

[1] https://cyc.com/

[2] https://cyc.com/leadership-team/

View on HN

A sad epitaph

I’m very glad he tried, even if he only proved that it was impossible

He didn't prove any such thing.

Especially if you consider using a SOTA LLM to recreate and expand ResearchCyc (for fun and not profit)

Maybe Cyc was a success and Lenat lives on as it's consciousness?

Dead is dead. And even if Cyc had consciousness--which it doesn't--it certainly wouldn't have his consciousness.

It's an interesting thought experiment / philosophy / sci fi story premise though; if he spent all those years encoding his own thought processes and decision making into a program, would he have been able to create a convincing facsimile of himself and his "consciousness"? A turing test with a chatbot based on decades of self-reflection.

The 2014 film Transcendence touches upon this, but I can't recommend watching it. It's surface level and the rest of the film is mediocre.

Stephenson's "Fall; or, Dodge in Hell" deals with the hypothetical digitization of human minds. It has some interesting ideas like some of the synthetic minds deciding to wilfully become a hivemind

You're completely moving the goalposts. And I don't find it interesting at all (I mean, I find the general subject interesting and have delved into it for decades but I don't find this sort of casual question based on no such research, trying to connect it to the wrong thing [Cyc] at all interesting) ... would he have been able to create a convincing facsimile? If and only if the encoding method were effective--that's a tautology. Was Lenat's methodology effective to that end? No, of course not, and that wasn't its intent.

"based on decades of self-reflection"

Daniel Dennett--sadly lost to us--explained in detail why "self-reflection" is not even remotely effective to this end ... our internal processes are almost entirely inaccessible to us.

Reply on news.ycombinator.com