Does Gemini Mean Trouble for Google?
Hello free tech newsletter subscribers and welcome!
I’ve got a writeup of whether Google’s Gemini launch saved their bacon, or left them starving. Or something. Was it good or not? That’s what I’ll write about.
Paid subscribers get more stories about Meta AI, Messenger getting encryption and who’s using AMD’s Nvidia AI chip competitor.
Happy Thursday y’all!
Tom
Big Story
"Early impressions of Google’s Gemini aren’t great | TechCrunch"
"Google stock up after company announces Gemini AI model"
"Alphabet Rallies as Gemini Release Eases Fears Over AI Position"
"McDonald’s says it will use Google AI to make sure your fries are fresh - The Verge"
"Google’s Gemini AI won’t be available in Europe — for now"
Depending on who you ask, Google's release of Gemini was a. Great move or a botched release.
Investors seemed pleased. GOOG stock had its best day since August 29th. Wells Fargo's trading desk said, "I’d summarize GOOG as proving that they still have some bite.” Bank of America analysts wrote "data suggesting that Google has best in class, proprietary, AI capabilities can be positive for the shares in 1H′24." The only frequent negatives were questions over how Google will monetize Gemini. McDonald's even announced it will be using Google's Large Language Models to help it optimize store operations in 20-24. Essentially, companies seem positive on it, users, not so much.
Google only released Gemini Pro, with Gemini Nano for on device models coming this month and Gemini Ultra, the most capable model delayed until next year. Gemini Pro is not doing well in actual use with many anecdotal mistakes posted online like getting the 2023 Oscar winners wrong, and even getting it wrong in different ways for different users. There are also some examples of bad translations. And the bot has a prohibition about commenting on some current events. It seems to do better at coding, but there were a few examples of mistakes there too.
Meanwhile Europeans can't get it. Google did not launch Gemini in the EU or UK because of regulatory hurdles.
Today on DTNS my colleague Justin Robert Young is going to make the case why this is a bad move for Google and will lead to consequences for its leadership. I don't think he's unreasonable. But I always hedge my own personal feelings against the market mood. The market is far from infallible, but when it seems to be positive about something and I'm not I know I'm missing something. IN this case I think it probably comes down to the market believing that OpenAI will not progress as fast in 2024 as it did in 2023, giving Google time to catch up even if it's still stumbling. Google has enough size and momentum to make that work the way any big company can. But that hinges on OpenAI stalling which isn't a guarantee, and flat wrong if you talk to people involved with OpenAI.
I do think Google is playing the part of IBM in the 1980s right now. OpenAI would be Apple in this metaphor and Gemini the IBM PC. IBM ended up losing that market to Microsoft. Despite Microsoft actually backing OpenAI, there is no Microsoft in this analogy. You could also flip it and say Microsoft is IBM and OpenAI is a young Microsoft and Google is HP. Which would have Google eventually adopting OpenAI platforms. I don't think either of metaphors fits perfectly but we're definitely in a world where a young upstart is challenging the slow incumbent and we've seen plenty of other examples of that in tech through the years. Which is why perhaps the best thing to do is keep an eye on Anthropic. Maybe they're Apple after all.
Keep reading with a 7-day free trial
Subscribe to Tom Merritt Tech Newsletter to keep reading this post and get 7 days of free access to the full post archives.