Here's what's happening right now.
Companies are cutting people because AI can do their work. And the people getting hired to replace them? Their job is basically to train the AI. Label the data, write the prompts, build the workflows. Once that's done, they're out too. You're training your own replacement and calling it a job.
This is pushing a wave of people toward starting their own thing. Not because everyone suddenly wants to be a founder, but because building something of your own is becoming the safest bet when employment keeps pulling the rug.
The tools are ready. You can build an app in a weekend. You can launch a product with no team. The barrier to shipping has never been lower.
But the barrier to picking the right idea has never been higher.
When millions of new founders show up at the same time, they all grab from the same pool of obvious ideas. Another project management tool. Another AI writing assistant. Another marketplace. They build fast, launch into a crowded market, and fail. Not because they couldn't execute, but because the idea was dead before they started.
The other failure mode is worse. You build something genuinely new, but nobody will pay for it. No market. No urgency. You spent six months solving a problem that isn't painful enough.
Both failures come from the same blind spot: founders don't know what just became newly possible.
Every week, thousands of scientific papers get published. Inside them are breakthroughs that quietly remove blockers. A new material makes something cheaper. A new method makes something faster. A new model makes something accurate enough to actually sell. These breakthroughs open windows where a product that was impossible last year is suddenly buildable, and nobody has built it yet because the research is too new for anyone outside academia to have noticed.
That's what Fontlume finds.
We read millions of papers. We use AI to find connections across fields that no human would spot. Then we check commercial demand, map the competition, and identify what's still wide open. The result is a short list of real, specific, new opportunities every week.
Five opportunities, five categories:
Build This Weekend. The science is done. The product layer is missing. Ship it.
AI Makes This Possible. A recent AI capability just removed a blocker that used to require humans or expensive hardware.
Deep Tech Bet. The science crossed a threshold. Needs investment, but the window is open now.
Hidden in Plain Sight. Should already exist. The data is public, the market is obvious, nobody connected the dots.
Watch This Space. Not ready today, but the pieces are falling into place fast.
For each one: what changed, when, what you'd build, who's competing, and what's still wide open.
We want you to succeed. That's the whole point. Not to impress you with how many papers we scanned. Just to put the right idea in front of you before anyone else sees it.
First issue ships May 22nd.
Theis