Isn’t it obvious to anybody paying attention what their actual strategy is? They know they probably won’t “win” a commercial AI race vs OpenAI or Google, and pr

Isn’t it obvious to anybody paying attention what their actual strategy is? They know they probably won’t “win” a commercial AI race vs OpenAI or Google, and presuming some other company does win the commercial AI market with closed source tech, they may be forced into a multitude of bad outcomes: 1. having to pay out the nose for licensing/SAAS of the winning tech vs shipping a noticeably worse second rate in-house product 2. Struggling to compete with the winning company for hiring, assuming the winning company makes a ton of money at high margins and can outbid Meta. 3. Having to continue building “applications” products on other companies’ platforms, something Zuck has repeatedly noted as a problem and is a major motivation for their investments in VE. Also, they’d have a hard time commercializing the current iteration of AI products given their current products and business relations anyway. Even if they “won” in terms of superior tech they’d be in one of the worst starting positions for actually making money off it, given their lack of enterprise SAAS business customers, the potential negative impact on existing product lines in consumers’ hands, (what I expect to be) advertisers’ relative indifference to genAI, and being generally considered less trustworthy than most other big tech companies and so less able to pivot to enterprise software. So for them the strategy that makes the most sense is to open source really good, but not cutting edge, AI for others to build on top of it. It probably costs a lot less to play permanent catch up because they can let others make the big investments in speculative research and don’t have to pay for the absolute top experts. Then by open sourcing it they make it harder for OpenAI and Google to commercialize their offerings at high costs, as many companies will build off Meta’s tech for more control, and what customers they do get will be constantly deciding if what they’re paying for is worth a slightly-worse but cheaper open source alternative. This also allows Meta to more easily influence the AI applications space despite having less capable models, and make it easier for 3P commercializations to take off rather than ceding everything to OpenAI/Google potentially going all in on 1P, which would be greatly beneficial if Meta ever were to own a platform that benefits from GenAI (cough cough the metaverse). It’s smart but transparently obvious because like, it’s Meta lol, they don’t spend this kind of money out of the goodness of their hearts.