META Is Unlikely To Leapfrog OpenAI, Anthropic, xAI and Google
I believe META ( META 0.00%↑ ) lacks the required culture to leapfrog its AI competition.
Zuckerberg blows with the wind (Libra; Metaverse; first open source AI, now proprietary models). To attract the best AI researchers, Meta needs to demonstrate a clear AI vision. I don’t have a sense for that. Zuckerberg is hiring AI leadership, but what do the people doing the work look like? Are they passionate? Are they curious? Are they allow to take risks and fail? (the way Zuckerberg is moving his AI staff around, I doubt there is much risk-taking).
Meanwhile Google, OpenAI, Anthropic, even xAI have a clear purpose, and each company has its roots in AI.
Google ( GOOGL 0.00%↑ ) has been running AI under the hood of its various software products and devices for years. Google Search has always had AI models powering the service. All of Google’s products are underpinned by AI.
The AI research that spawned OpenAI came out of Google DeepMind.
Anthropic founder Dario Amodei came out of Google before he went to OpenAI, and eventually formed Anthropic.
Google is the only company that runs a full AI stack with its in-house designed TPU chips, its cloud business (GCP) and its proprietary language models. Speaking of which, I can’t wait to try Gemini 3.0.
xAI - the new AI entrant - has a clear purpose around AI and has something to show for it. Grok is providing xAI a far better return than anything AI-related cooked up by Meta.
Open Source: Then of course there is the open source threat. If you are a hardcore programmer, I’m not sure why you would not use an open source model, local to your machine, that is highly customizable and private, rather than renting a model from OpenAI, Google or Anthropic.
Here is Mark Zuckerberg ( $META ) on CEORater:
