Context:
I can’t help but get déjà vu watching OpenAI’s “code red” response to Google Gemini. Fourteen years ago, I was in the only Facebook engineering building when Google+ launched. It truly felt like a sneak attack from a massive, well-resourced competitor. Google+ was polished, fast, and—at least on the surface—scary. Facebook’s response was a full-on lockdown: Mark rallied the troops, people pulled all-nighters, and there was this intoxicating sense of urgency, like we were at war. It was a blast. But here’s the thing: despite all the noise, Google+ turned out to be a nothing burger. Google copied the features but missed the magic of making social actually work. The only lasting impact? “Login with Google” gave them a real wedge, but otherwise, Facebook won.
Market Signal:
Fast forward to now—2025. A lot of us who lived through those Facebook/Google+ wars are now in leadership at OpenAI, itching for another round of that adrenaline-fueled hypergrowth. I get it. Once you’ve felt that energy, you want it again. But here’s where I think the analogy breaks down: OpenAI’s “code red” isn’t Facebook’s Google+ lockdown. This is a much more existential moment for OpenAI.
Takeaways:
The crucial difference? Google isn’t entering unfamiliar territory this time. With Gemini, they’re reclaiming ground they arguably owned all along. Google has the distribution, technical talent, data, brand, ad platform, and even hardware. This isn’t a moonshot—it’s a homecoming. Sam (Altman) is a visionary and a world-class marketer, but with the hand he’s holding, I don’t see how a “code red” for ChatGPT changes the game. Even if OpenAI ships a better product for a minute, Google can and will catch up. Unless there’s some wild, game-changing trick (which I can’t even imagine right now), incremental features or harder/faster shipping just don’t matter in this landscape.
And, ironically, that’s the story of AI itself: it reduces the leverage of human capital. In the age of steam shovels, it doesn’t matter if you have a digger who can dig a little faster. In the age of AI, having a team of brilliant engineers who can “code harder” in a crisis just isn’t the difference-maker it used to be.
Asks:
So, for founders, operators, and investors:
Don’t get caught up in nostalgia for the old playbook. The terrain has changed.
Ask yourself: What actually moves the needle in an AI-native world?
Where are the new defensible moats?
And if you’re at OpenAI or a similar company, what’s your real edge when your opponent owns the distribution, the data, and the platform?