


Meta's Superintelligence Lab Considers Shift To Closed AI Model (yahoo.com) 12
An anonymous reader quotes a report from Investing.com: Meta's newly formed superintelligence lab is discussing potential changes to the company's artificial intelligence strategy that could represent a major shift for the social media giant. A small group of top members of the lab, including 28-year-old Alexandr Wang, Meta's new chief A.I. officer, talked last week about abandoning the company's most powerful open source A.I. model, called Behemoth, in favor of developing a closed model, according to a report in the New York Times, citing people familiar with the matter.
Meta has traditionally open sourced its A.I. models, making the computer code public for other developers to build upon, and any shift toward a closed A.I. model would mark a significant philosophical change for Meta. Meta had completed training its Behemoth model by feeding in data to improve it, but delayed its release due to poor internal performance. After the company announced the formation of the superintelligence lab last month, teams working on the Behemoth model, which is considered a "frontier" model, stopped conducting new tests on it. The discussions within the superintelligence lab remain preliminary, and no decisions have been finalized. Any potential changes would require approval from Meta CEO Mark Zuckerberg.
Meta has traditionally open sourced its A.I. models, making the computer code public for other developers to build upon, and any shift toward a closed A.I. model would mark a significant philosophical change for Meta. Meta had completed training its Behemoth model by feeding in data to improve it, but delayed its release due to poor internal performance. After the company announced the formation of the superintelligence lab last month, teams working on the Behemoth model, which is considered a "frontier" model, stopped conducting new tests on it. The discussions within the superintelligence lab remain preliminary, and no decisions have been finalized. Any potential changes would require approval from Meta CEO Mark Zuckerberg.
Yikes (Score:2)
Re:Yikes (Score:5, Insightful)
In this case, the aim of Meta is to corner the AI market, knowing that the scientists and students who could invent revolutionary AI models are bad at programming, so you give them free tools that don't require programming, just a bit of python glue. This makes the competitors dependent on Meta, and their businesses controllable. Google did the same thing with its TensorFlow framework. And the LLM companies are doing it with "foundation models". Microsoft did it by turning a blind eye on Windows piracy in the early days.
One reason to close off AI models is so that the investors can be more easily told excuses, sight unseen, when they demand progress.
Re:Yikes (Score:5, Insightful)
How many companies used llama instead of paying OpenAI for API access?
If Meta saw OpenAI as a market competitor then denying them revenue is a cunning move.
Especially if llama isn't Meta's core competency but ChatGPT is OpenAI's.
Maybe even legal in that aspect, though they both seem super guilty of not paying for copyrighted material.
Does Not Compute (Score:2, Informative)
Associating the term "Superintelligence" with Meta and Zuckerberg does not compute. It's like associating Calculus with the tic-tac-toe playing chicken in the mall.
Who doesn't wish they were Apple? (Score:1)
But are your products that good?
That people will pay and love your device ?
I think all the eye toys so far don't meet that standard and the deep surveillance factor and extreme commercialism potential
Don't do it, Zuck (Score:1)
Re: (Score:2)
Speak for yourself. I pointed out the amazing contributions Meta was making to both VR and AI but nobody has wanted to listen. In the meantime I made a killing on their stock.
Open it, get the knowledge, close it (Score:2)
You can be better than that.
Meta's newly formed superintelligence....that's (Score:1)
poor performance (Score:2)
"delayed its release due to poor internal performance"
Meta recently hired top talent from OpenAI. Those guys probably reviewed Meta's codebase and saw a lot of fundamental flaws, they will want to either rewrite it or start from scratch.
Privatization is inevitable as Meta is greedy (Score:2)
Typical classist move is to try to take private ownership of public property, we've seen this time and time again, it's business as usual, steal from and cheat those who create good things so entitled people can continue their free ride at our expense. None of these massive transnational corporations are ethical, there desperately needs to be legal reforms so the people in charge are held accountable for their decisions and an end to limited liability for shareholders. Otherwise this systemic corruption wil