Sell $META
This might not be “the big one” for the AI trade, a bear market like the dot-bomb of 2000. But why take a chance? I’m backing out of the weakest link, which is $META.
The Generative AI trade is on its last legs. Millions of baby boomers are anxious not to be bag holders when it ends. It makes sense to pare back on the weak links in the value chain.
Why sell $META?
- Meta has suddenly stopped hiring expensive AI talent.
- Meta’s LLama models are trailing in the AI race, and they’re not open source.
- As with its “metaverse” efforts a few years ago, Meta has a habit of getting out ahead of the pack, then pulling back.
Like Google $GOOG, Meta is an advertising company disguised as a tech outfit. It’s the Fox in the tech henhouse, copying the innovations of others with “good enough” solutions, then using the power of its cloud network and financial base to destroy the competition. The difference is that it’s not a hosting company. As with Apple, Meta’s cloud is only for Meta.
But when CEO Mark Zuckerberg tries to innovate, he often fails. Meta Quest headsets have been losing money for years. His metaverse never took off.
LLama is a cross between these two approaches. It seeks to copy other Generative AI models and beat them with semi open source. But Meta insists on controlling improvements. That means LLama isn’t open source. Developers have read the fine print and backed away, in favor of models from China with real open source licenses.
Now, with even OpenAI CEO Sam Altman admitting AI is a bubble, Meta is backing away, renaming its AI group “TBD.”
The Bottom Line
Meta still dominates emerging markets. Its services represent what’s left of the “free Internet” there. Facebook is still popular among older Americans, and it has taken advantage of X’s freefall with Threads. The question is whether this moves the needle, whether it will deliver outsized gains, or “alpha,” at a price-earnings ratio of nearly 27.
I’m not interested in that risk, and I guess many older investors feel the same way.
