As @vv points out, a lot of people think I am talking about tweaking LLMs to make this happen. I'm not, though they might or might not be *part* of the answer here. Think "AI" in much broader terms in this thought experiment.
0 Boosts
0 Likes
But I don't think it needs to *remain* impossible. I actually do think there is a lot of other space for exploration. Most of the work I have done throughout my career has been in the space of "but that's not how it works", and developing into the spaces of research where actually, there is another possibility.
My question by putting it into the abstract wasn't because I wanted to trick people into "ha ha, guess what, it's LLMs", because it's actually *propagator networks*, which are a dramatically different (and in many ways, diametrically opposed, but can be used on their own or in conjunction with machine learning techniques) branch of "AI".
But again, I didn't want to get into the specifics of the technique. But in some ways it doesn't matter, because LLMs have won the entire brainspace of what people *think* is possible right now.
If you want to know *WHY* I wasn't specific, it's pretty simple: I didn't want to debate whether or not the specific branch of AI I think is promising (which again, is not LLMs) can or can't be successful. I wasn't trying to trick people into "ha ha I'm just talking about LLMs". But I think given the social context we're in, it's understandable that was peoples' impressions.
Well the poll has ended, and holy moly, there was a lot of response here, both to the poll and also in the comments. It was a dead heat for a while there, and still came pretty close to it.
Within the last hour, I kind of skewed the results by throwing this out there, that I *do* have thoughts on how to build a system that isn't based on LLMmaxxing which fulfills these properties, and that seemed to bias a lot of the votes coming in within the last hour towards yes https://social.coop/@cwebber/116969383379862650
But still, a lot of people are clearly disturbed, that any kind of improvement to automation of intellectual labor, even with all of the *main* things that we tend to complain about with *present-day* (and LLM-based) AI removed, making those "better" is unclear if a win given the lack of a change in socioeconomic context.
So... what now?
One thing is clear from reading many of the replies here: LLM companies (whom I am *explicitly against*), trying to brand their tools *as* AI, clearly won: both pro and anti LLM people are jumping into the replies to tell me that the only thing that can be called "AI" is LLMs, and conceiving of anything else or which has any other properties is impossible.