Like the before times, eh? Gonna need a pack of cheap earbuds, some silly bands, and a lack of care for my future hearing capacity.

There was a subreddit for everything. “The front page of the internet.” You could find subs for birds shitting on things, with buff guy arms photoshopped onto them.

That was what attracted me to Reddit. I first subscribed to things like AskReddit, NoStupidQuestions, TodayILearned, AmITheAsshole, MildlyInteresting, WhatIsThisThing, … stuff like that.

Latter I dove more deeply into niches like AskPhilosophy, LearnPython, JavaScript, LinuxQuestions, HomeAutomation, Consciousness, PostgreSQL, …

Then I started also using it for news…

It had everything. Every little division had its own subculture. If I didn’t like one, I could find alternatives — plenty of them.

It also became somewhat of an archive too, given that you’d use Google to search Reddit. You could find anything on there.

I mostly learned of Lemmy because I began to despite Reddit. That’s not really an effective marketing strategy. Lemmy needs something more than “Fediverse” when at least half of the population isn’t willing to understand what that is or outright accept it as a necessary solution. Lemmy needs to compete, giving more people a reason to participate. If it can’t incentivize more people to participate, then more people aren’t going to participate.

I don’t know what that looks like for Lemmy. It’s not whatever Lemmy does right now, though.

There’s a fundamental difference between AI hallucinations and humans believing something which is incorrect.

With an AI, it’s a problem of statistics. The context window biases the model in just the right way to produce an answer which is incorrect. But, to the model, the answer was correct. That’s to say, it didn’t just fuck up its calculations… no, it arrived at the correct, albeit incorrect, solution.

That’s not the same as what a human experiences in its “hallucinations.” Not even close. Often, a human can sense that something is off (to say the least).

A human can work out their own error. An LLM can not. That’s not because an LLM isn’t good enough to do so. It’s because the LLM is missing something important. Once you add that thing into the architecture, I’d argue it’s de-facto no longer an LLM. Because, by that point, they’re leveraging new science for new behaviors in a much more complex system which overcomes the universal scale limitations of LLMs.

That’s my logic in a nutshell. LLMs might “lead” to AGI simply for the fact that it makes obvious we still haven’t achieved AGI, and so people would ideally start looking toward other approaches.

To me, an LLM is perhaps one subsystem in what should be a pretty large ecosystem of tooling. AGI would likely be the whole system, not just the LLM and really the LLM doesn’t deserve the credit (imo).

Because an LLM doesn’t scale to generalized intelligence. That’s obvious when you look at the scale issues, such as hallucination.

Whatever technology would have potential to produce AGI, it might be similar to LLMs but procedurally it will be a different thing.

IMO, the closest we might get is akin to comparing a porta-potty versus a house. A porta-potty can never become a house. You need more plumbing, a kitchen, …