Subject: That is an interesting take.
Author:
Posted on: 2023-03-27 00:30:01 UTC

I have heard people say that generative AI is just predicting words in a sequence, and I suppose that is kind of true, but I'm personally with the Hard Fork co-hosts when they say that comparing it to a search engine may be underselling it a bit; generative AI is certainly capable of a lot more than finishing search results. The podcast has also said that the language to talk about this technology doesn't quite exist yet, and I do feel like it's difficult to compare something like this to any preexisting technology; I don't think we have fictional precedents to generative AI like this, either, but correct me if I'm wrong on this.

These language models are certainly trained off the internet, and the internet can definitely be wrong. But I feel like there's a reason groups like Bing and Google are trying to make these bots more accurate. Is it a losing battle? Is sometimes being wrong fundamental to how AI works? Maybe it's too early to tell, but I personally think that this technology is still in its infancy and can be vastly improved.

Or maybe I'm just a nerd dreaming of living in science fiction land and generative AI is a passing fad that'll go the way of the zeppelin. Who knows.

And I'll definitely check out that blog! I'll also take the chance to plug Hard Fork (which I've already mentioned a few times, heh), which everyone interested in AI and technology in general should listen to. It's also pretty funny at times, which is definitely a plus.

Reply Return to messages