And lo on the 23rd day of the seventh month of the year of our Zuck, verily did Meta release a new open source AI model, all 405 billion parameter's worth. Known as Llama 3.1, it is apparently the largest such open source model yet and so Meta's main man Mark Zuckerberg has taken to various YouTube channels to promote its technological and social munificence. Oh, and to speculate on a future of AI that's predictably icky, plus complain about his competitors, mainly Apple.
Speaking to Youtube channels Rowan Cheung and Bloomberg Originals, yesterday, El Zuck explained that Llama 3.1 is, «by far the most sophisticated open model that anyone has put out. It's competitive with some of the leading closed models and in some areas is even ahead.»
Zuckerberg also emphasised how important he sees the open-source character of Llama 3.1, in contrast to closed-source models like ChatGPT. «Open source is a really important ingredient to having a positive AI future,» quoth Zuckerberg, explaining that, «open source is an important part of how we make sure this benefits everyone and isn't something that's locked into a handful of big companies.»
You mean, a bit like how social media is locked into a handful of big companies? Indeed, Zuckerberg's concerns regarding closed-source AI are more than a little redolent of broader reservations around dominant social media platforms.
«It's a little bit soul crushing,» he says of competing with Apple on AI, «when you build features that you believe are good for your community, and then you're told that you can't ship them because some company wants to put you in a box so that they can better compete with you.
»My concern for AI at this point isn't actually Apple. It's other companies and how that would evolve. I'm not even saying that they're bad people, it's just a physics and incentive structure to the system."
Uh huh. Does that not remind Zuckerberg of, oh I don't know, the physics and incentive structures around social media platforms
Read more on pcgamer.com