Ive been watching “AI in context” for a few weeks (they make long form biopic content on current state of AI - really good stuff).
This dropped today; it’s about the wheeling and dealing behind closed doors at OpenAI re: Sam Altman’'s firing. It’s a lot more watchable than that sounds :)
https://www.youtube.com/watch?v=_eYTkvZqbnQ
The line that brought me to a stand still was “the future of AI depends on the moral compass of like 5 people”.
I know we’re all about local LLMs here…but it’s sad to see yet another “Don’t Be Evil” mission statement get swallowed up market forces. OpenAI was meant to bring balance to the Force, not leave it in ruins, and yet…
I think folks here (and Lemmy generally) are more savvy about AI then the gen pop…but even if you’re training a nanoGPT model from scratch on hardware you own …you’re still beholden to outside forces. Eg: the people’s champion - Qwen - seems to have split or gone closed weights for 3.7.
Things that make you go hmm.
Anyway…just signal boosting a cool video.
Why especially in this community assume that corporate closed AI are all that matters?
These five people are taking bad decisions regarding AI. Thankfully, there are dozens of others who work on open models and who want to keep things open.
Because it costs a lot of money to train models, whether open or closed, so the people with money are in charge. Yes, local AI is awesome but we’re not going to see newer versions without that money being spent by someone.
Actually, training a very big model on trillions of tokens using huge clusters is one way of doing that, but certainly not the only way.
Communities have been fine-tuning models on much smaller machines and on a budget of hundreds of bucks.
Right now, money is abundant and feeds even open on its models, so it would be stupid to not go in that direction. But when it dries up, it won’t be the end of models improving.
Network effects. Nothing happens in isolation. And (as is already being shown with Qwen) it’s not as if the incentive structure doesn’t apply to open weight labs. There are no saints in AI - same dynamics everywhere.
Right now, training a new near frontier llm from scratch is a multi million dollar proposition. Until training costs are 0, that $$$ needs to come from somewhere. Which incentivizes the same behaviours.
IOW, if you think GLM et al are the good guys to Anthropic and OAI’s bad…no…they’re just trying to undercut them for market capture.
That’s the cynical read, anyway. ICBW.
Then why think things depend on the moral compass on some people if you think there is an unescapable dynamics that’s at work?
I do think that open source community is winning that battle and it’s an important battle and the ego of this five psychopath is kind of obscuring that huge victory that’s won by hundreds, thousands of developers and researchers.
I don’t think GLM or Anthropic or OpenAI are “guys”, good or not. They are companies, you don’t anthropomorphize these, they are beings that only crave for profit.
Actual people with actual morality are the people who are deciding to work there or to quit there. There’s a reason why OpenAI is bleeding people. There’s a reason why people like Le Can accepted to work for a company like Meta but imposed that they continue to publish.
The five people that this article mentions are the trees that hide the forest.
The inescapable mechanic is that the most egocentric people capture all the spotlight, almost by definition, but they are not AI, they are not the developers, they are not the researchers, they are not the people who innovate there. They are the people who take working efforts and turning into a soulless profit machine that often drives moral people away.
I wish we were less blind to the actual dynamics at play and were spending less time on people trying to get artificial spotlight.
AI (frontier LLMs) are capital intensive. In that way, it’s unlike many other kinds of software. I don’t think near-frontier open source models will be a thing for very long. Even if you have the papers that describe everything, it will take many millions of dollars just in compute to train frontier models, and then a small army of people doing reinforcement tuning. The people who invest that capital will want to maximize returns and will be hostile to the idea of open source.
Another capital intensive industry that arguably does a lot of good, and probably has a lot of idealist researchers is pharma. It’s also one of the most predatory industries.