Hacker News Reader

Comment by qingcharles | Hacker News Reader

I wrote my first ELIZA as a 9-year-old in Spectrum BASIC in 1986. All I wanted was my own KITT from Knight Rider.

Who knew it would take another 35 years before we got close?

We've hit the fun bit of the hockey stick curve now though. It scares me to imagine what is going to happen with AI over the next few years.

View on HN

I've worked with the various LLMS quite a lot at this point (for someone who doesn't do it professionally ... well, I guess I sort of do ... but not for any of the companies who make them.) Anyway, it's amazing how close these are to what our old (i.e., GOFAI) Spreading Activation models form the 1970s would do. You can wikipedia it FMI, but in practice the SA models would be loaded up with a context, and then the SA would keep it one track for a while, but as you got farther from the priming context it was very very hard to get it not to eventually spin off into some quasi-relevant confabulation. I think that the embeddings is just a modern way of creating a non-discrete semantic net, and the GPT is just implementing the moral equivalent of spreading activation.

(trying to understand your comment)

This FMI? (https://fmi-standard.org/) And what do you mean by "moral equivalent"?

From the context of the sentence, my most likely guess is "for more information", but yes, that comment is a prime example for "how not to write if you want to be understood".

https://simple.m.wikipedia.org/wiki/Moral_equivalence

Sorry. The couple of typos don’t help. (I tried to edit them but it was too late.)

Yes, per @fhars, “fmi” is pretty common text-speak for “for more info” (which google will tell you if you search for it).

“Morally equivalent” is a term used in the classical hacker community to mean something like “essentially the same as, although not technically Computationally equivalent”. (And, as I noted below, it is also a standard, if a bit obscure, English phrase: https://simple.m.wikipedia.org/wiki/Moral_equivalence)

np. Thanks for the clarification. The discreet bit was clear but TIL re classic hacker speak.

Considering how many fell for the original ELIZA and took it seriously, despite being told what it is, I'm scared shitless. People's willingness to impute actual intelligence to a dumb machine is concerning.

People have, and will continue to concede mental authority to rule based systems. "I'm sorry but the rules say" is one of the very common statements you here, but when you look at the situation one would think it would be easy to recognise the rule in question does not apply. The rules in cases like that could be considered a 'dumb machine'.

Now imagine what happens when the machine becomes semi-smart.

During COVID we had a school administrator proclaim "We must OBEY the COVID!!". Since then it has become my catch phrase for humans unquestioningly following orders.

Got close to a bunch of special cases glued together?

Reply on news.ycombinator.com