- cross-posted to:
- technology@lemmy.ml
- cross-posted to:
- technology@lemmy.ml
You must log in or register to comment.
Capitalists the moment the free market™️ no longer works for them: “I love state intervention!”
Looks like they are panicking
“Daddy! China hurtin’ my feewings! Hewp daddy! China bein’ a big meanie!!!”
The USA wants a world where AI wants is given permission to consume all copyrighted content for free, but we are charged for access to scholarly papers.
Publicly funded scholarly papers, aren’t they?
Good point! Obviously the solution here is to stop funding the science!
(/s)
there’s that free market we keep hearing about.
AI models trained on public information should be open sourced and publicly available.
Billionaire’s should not own this behind closed doors.
Silicon Valley is just a scam. They’ve made billions off selling your data, and selling targeted ads with your own data to be shoved back into your face. Social media has ruined society. Every “innovation” has been some way of stealing from you, whether it’s your data, your attention, or now, the entirety of documented humanity.
/end yelling at clouds
Good luck with that. DeepSeek has already been reverse engineered.
Isn’t DeepSeek open source? Is there a need to reverse engineer it?
“Open source” in ML is a really bad description for what it is. “Free binary with a bit of metadata” would be more accurate. The code used to create deepseek is not open source, nor is the training datasets. 99% of “open source” models are this way. The only interesting part of the open sourcing is the architecture used to run the models, as it lends a lot of insight into the training process, and allows for derivatives via post-training
Deepseek actually released a bunch of their infrastructure code, including the infamous tricks for making training and interference more efficient, a couple of weeks ago.
It certainly is a lot more open source than OpenAI, that’s for sure.
Yes, and no. Yes in that they’ve released the research papers, pretrained parameters and weights of the model itself. Which is more than I can say for “OpenAI.” But no in that it doesn’t include training data or other critical components. Luckily, they’ve shown how they did it which makes it easy for anyone else to reverse engineer the process. That’s what Altman is afraid of.
They released the major components of their training and interference infrastructure code a couple weeks ago.
They should probably just focus on their products. They introduced so much ‘guardrailing’ into their product its practically useless. Beyond that, there is a crap ton that can be done with the current crop. There are no guarantee’s of a technical moat and we don’t know where the next advance will come from.
I mean hell, google slept on transformers after creating them in the first place and ended up scooped by OpenAI’s team. So who knows.
But good product, good UI, no BS, no gimmicks, that sells. If OpenAI is that company, I’ll bite. If they arent, I shop with my feet.
One example of a product I would buy right now. I give the agent/ system my shitty sloppy demo code in a ipython notebook. Its shitty, but it works. Maybe I have to give it an example of what “correct” output needs to look like.
And then… I walk the fuck away. And in an hour (or two, I don’t give a shit), and my demo research code has been committed to a production ready git.
Instead they are doing, whatever the fuck they think they are doing with “Deep Research”, which, as of every use I’ve tried to make of it. Its completely worthless.
