[ai_what]

I like toying around with local AI models. I do so a lot actually. But the AI gold rush is starting to leave a sour taste in my mouth.

It's astonishing how quickly this technology went from "new" to being used by profiteers trying to milk it for every last dollar. I can't even remember how many new projects/websites I come across on a daily basis, which are mostly just a wrapper for ChatGPT with custom prompt. Not to mention what the bigger companies are doing, as stated in OP's article.

The irony is that even the "Wild West" of cryptocurrency (I know) had a brief, innocent era where enthusiasts drove innovation for the sake of innovation, not just profit. That guy that bought pizza with bitcoin for example, it was fun and kinda cool.

AI however, went from something most people had never heard of to full-speed exploitation. It's as if greed just keeps getting worse in this world. Everything is a potential source of income now. Apologies for the rant...

[Decabytes]

This is true for everything now a days. People collect limited edition things speculating they will be collectibles in 20 years, they jump on new programming languages and try to develop the first libraries so their name can be attached to the de facto CSV parser for the language, they try to get on new social media platforms early to become the next influencer etc.

The world is a lot less naive then it used to be

[DavidPiper]

> The world is a lot less naive then it used to be

I have a slightly different perspective on this - I think people have become more naive than they used to be.

"Speculating on future value" in the ways you've described - either investing money in collectibles or investing time in writing fresh libraries, investing time and money in attempting to become an influencer, etc - is basically gambling.

Your odds of success are either extremely low, competition is high, or investment takes a long time to bear fruit with high risk (collectibles becoming worthless, new programming languages dying, social media following never taking off, etc).

Spending 5 years chasing trends or buying up collectibles - basically trying to get lucky - looks "less naive" when it works, but it's basically survivorship bias (or first-mover and other strategic advantages, etc). If you fail, you've sunk a large amount of money and/or time in your life into a bunch of "worthless" pursuits.

(And to counter the "but it's not worthless if you learn something" argument, that's true, but it's not the point being made here. And using that as post-fact justification for losing all your time/money is a self-sustaining problem if you don't consciously evaluate your opportunity cost in advance.)

[coldtea]

>The world is a lot less naive then it used to be

Or rather more hopeless. You could get by in the past without hustling all the time like that.

[talldayo]

> AI however, went from something most people had never heard of to full-speed exploitation.

To be fair, the timeline goes a bit further back. Markov chains had their moment in group chats a decade ago, it's hard to find a communal IRC room without someone's Markov bot lurking around. Then came BERT and GPT-2, both of which had their innocent days of getting finetuned into somewhat-usable form. After GPT-3 was the "oh we can sell this" moment, and all the cutesy, semi-efficient research that was being done got sidelined so people could make their own ChatGPT-killer. That is the disturbing part to me; so many people (OpenAI included) abandoned efficiency for marginal gains. This seems to be the trend going forward, likely exacerbated with GPT-5; we're going to need bigger boats (servers).

[kjkjadksj]

Look how crypto started: people interested in it mining on their own when that was still possible. Meanwhile for most things AI its not some upstart hacking at home but due to the demand for compute from the getgo, someone who needs funding and therefore has to play the game as its being played.

[FuckButtons]

I think the difference is that crypto requires a certain leap of faith to imagine its utility. But a language model doesn’t. So if you’re someone who doesn’t really care too much about the technology for its own sake, there’s still a descent chance you can imagine a use case for one.

[mglz]

> being used by profiteers trying to milk it for every last dollar

Not just this, but also enshittify service, spam content, increase spying and in general disempower users.