I've worked for a number of software startups - and declined to work for some others. It works like this.

Somebody has a great idea (maybe a new tech that they think everyone will want once it's released, or maybe a common problem that can now be solved, or even trying to build something that one of 'the giants' will want - and so buy you out). Write that all up, turn it into an 'elevator speech' - and contact some VCs. If one bites, you start. If one has a different-but-similar idea they'd like to fund, you probably jump (they've got better market data than you)

Then the tech folks (like me) try to turn that funded idea/solution/something into a real working product. It's an iterative process with a lot of time pressure. The first versions only kinda work, or work but in limited situations. Done right, it gets better.

No dishonesty involved - it just "works this way" (or at least it did). The VCs know there's maybe a hundred to one that everything takes off - but they stand to make a thousand (or more) if it does. Decent odds.

Here's the thing: that's software. Theranos was trying to follow this same playbook in the medical field, you know, where an incorrect diagnosis can play hell with somebody's life.