You're not a founder, you're a gig worker
Let’s break it down:
• OpenAI researchers helped build the most valuable model IP on the planet. They got six figures and a mission patch. Microsoft got the exclusive license. Sam got a multi-billion-dollar for-profit arm and total control.
• Zuck shows up and offers those same researchers $100M+ to switch stories. No nonprofit halo, no saving humanity. Just compute, comp, and “build God with us.”
• Sam cries foul. Why? Because the myth breaks: 1 that you’re part of a meritocracy 2 that you’re “on the journey” 3 that it’s about ideals, not exits.
Meanwhile…
• Scale AI workers in the Philippines were getting ghosted on pay while tagging data for $2/hr. Alexandr Wang becomes a billionaire. Meta buys half the company. Lays off 200. Wang walks into a new gig.
• Anthropic and OpenAI staff who helped push frontier alignment? They watched their orgs raise at $15–80B valuations and get nothing unless they were on the secret early cap table.
• Google DeepMind researchers? They’ve shipped more papers than any other lab. Most of them will retire with zero startup upside, while Google repackages their work as “Gemini.”
⸻
What happened?
Tech work used to come with a story: “You’re an owner. You’re early. You’ll make it.”
But that’s done. Today’s tech worker is pre-union labor in a zero-equity gig economy with a gold badge.
You’re just in time labor for capital-scale compute. And you’re still being told you’re a chosen one.
⸻
Who wins?
• Founders who lock the cap table before Series B
• VCs with double-digit preferences
• Corporates like Microsoft and Meta, who now own the model stacks and the labor
Everyone else? You’re a smart mercenary with branding issues.
⸻
Zuck isn’t the villain here in fact he’s briefly rebalancing things.
Sam’s mad because the theater curtain fell—and half the cast walked off stage for a better
I suppose the Harley-Davidson salesman told you that the motorbike would get you all the women too.
Investment is always about risk, and being a early-stage employee of a startup was always a matter of buying your shares with labour rather than capital, but the payoff was always a risky proposition.
Start a business. Otherwise, acknowledge that you are trading economic upside for safety.
Zuck isn't paying $200m per employee he's spending $200m on marketing and the droves of talent that will come and work for far less on the promise of a big pay out and prestige...
The only thing that has changed is the phrasing of the headlines.
Nothing in your post is special about AI. The only difference is the out of pocket valuations that a company releases a product and immediately it’s 50B before demonstrating any ability to run or manage a business, generate profit, etc.
The view that “we must distribute Google+NVIDIA+OpenAI+Meta+Microsoft market cap to the AI researchers who paved the way for the GPT papers” is noble maybe, but doesn’t really make a whole lot of sense
I currently work with near cutting edge video models another major player.
I’d go ahead and say I’m pretty well established in the space. Not a hero of it by any means, but well established.
I had 6% in my first startup as a programmer, 15% at my second, 5% at the third and 90% of the current business.
The minority stakes gave me enough capital to start my last one and it is bootstrapped w/o any VCs this time.
From my point of view, you either take the risk and get a bigger slice of the pie or you're not and getting a smaller one. It's perfectly OK to work from 9-5 in a big company and spending your time outside work to try to start something. The system is not rigged, the more risk you take, the more control you can keep the better the outcome will be.
Yeah, this guys had worked their tails off, but they got paid for that and had a change to make it big if the company sold and they vested. However, they didn't work as a startup founder without any money for 2 years with maxed out CCs. This risk is not a healthy one, especially when you have a family.
Bootstrapping an AI company now is hard, because of the capital requirements, but building something with AI models (or creating specialized ones) is still doable.
This system is so much better than the non IT sector jobs, where you're limited to work in a big company. You have choice to earn something above the salary, most people haven't.
That’s not the game anymore.
Today, most of these companies lock the cap table before engineers show up. The infrastructure costs are massive—bootstrapping isn’t just hard, it’s often structurally impossible in AI unless you’re rich or already plugged in. The story everyone’s still selling—about meritocracy, about “just take the risk”—doesn’t match the way power and ownership actually work now.
The people building the models aren’t getting meaningful equity. The platforms are licensing them behind closed doors. The upside is flowing to founders, corporate partners, and late-stage investors who shape the game long before most people touch it. And yeah, a few still break out. But the vast majority? They’re just very clever labor with a better kitchen.
If you made it work before, that’s awesome. But let’s not pretend the system hasn’t shifted underneath the next generation.
For the rest of the industry, the capital (Meta's 5 GW data center etc.) and operational ($200/month subscriptions for developer tools) are super troubling; however I can't help but observe that the practical impact on my day-to-day so far (with LLMs for search and programming assistants like LLM autocomplete and Claude) has been majorly positive.
I do understand the frustration, especially if you have been affected personally, but it's just the way the world works. Starting a company isn't easy either, and as I'm finding out requires a completely different skillset.