> No, Theranos was outright, egregious fraud, and most startups that "collapse under the weight of unrealized promises" are not committing fraud.

Most startups don't get the traction and publicity that Theranos got, however. I wouldn't be surprised if a lot of founders would go to similar lengths given the chance; they just happen to have someone apply some reality check on them before things get out of hand.

Look, there is still a big jump from the usual founder sharing partial vague optimistic statments (e.g., "our algorithm is great"), to the not so common direct lie (e.g., "we tested it with statistical significance, it works, here you can see these results on these real world inputs", while completely faking the results)

There`s a difference. I worked in a few startups. We make promisses that we can expect to accomplish in the future, we where clear that whatever we where working was still a prototipe or a promissing idea, but it was not ready.

She was going to great lenghts, saying her machine was already ready to production, and able to do acurate blood tests, when she knew it was impossible, and was using regular blood tests to fake results.

I would be fairly surprised. If only because, again, this is fraud and lands you in jail.

It happens more than you think, in many cases it’s easier to just sweep under the rug and move on. Few VCs want to admit they were swindled, and even fewer want the industry perception that they get founders investigated for fraud when things go sideways.

Theranos was unique in both the scale of the fraud and how non-investors were potentially harmed. Most of these are just founders who knowingly or unknowingly get in over their head and then basically walk away after having lied about the state of things.

I’ve heard a bunch of stories from VC friends, but the one I was the most involved with directly was probably 10 years ago. My mom mentioned to me that she had been chatting with some senior local government official who mentioned they had a startup datacenter provider taking over a nearby abandoned office complex and converting it to a green datacenter, bringing hundreds of jobs to the area - had I heard of them?

I had not, and after some digging determined that the company had claimed to have a bunch of unique technology, and leveraged that to get preferential terms from the local government - free rent, no property or payroll taxes, and other cash- and non-cash concessions. They had also raised millions from private investors by saying they had active datacenters with customers and they needed money to expand.

Except none of it was true. They had no customers, no datacenters, no employees other than the three founders and one of their daughters. They had made commitments in three different locations over the past three years to build a datacenter while simultaneously raising money by saying they ALREADY had active datacenters and customers.

I shared this with my mom, who shared it with local government, who hand waved it away as nonsense. But nothing ever happened at any of these sites, and after they ran through the money they filed for bankruptcy with revenue of $0 and assets of ~$12k. Their biggest investor took control of their land lease from the local government. Some investors complained online about being lied to, and the local government quietly swept it under the rug. I spoke with a local business reporter who basically shrugged and said, “yup, you’re right, but no one wants to talk about it”.

> Few VCs [...] want the industry perception that they get founders investigated for fraud when things go sideways.

That would be refreshing! I might pitch to them.

That fruit juicer company over promised and under delivered, wasn't fraud tho.

Actually, Juicero over promised AND over delivered, just in the wrong way.

I saw tons of teardowns of their machine saying how over-engineered it was for something that squeezes a bag of precut fruits and veggies.

> a simpler machine could have been built

That's the definition of over-engineering: https://en.m.wikipedia.org/wiki/Overengineering

You’re right about the colloquial definition but I think the OP is more on point about the crux of the issue. Most physical world engineering applications are about the skill of stripping away as much as possible and still getting the job done. It’s an exercise of working within constraints. “Over-engineering” is ignoring those constraints.

"Perfection is achieved not when there is nothing more to add, but when there is nothing left to take away. Now please take away this overpriced juicer."

Definition written by management who can't tell good from bad. Ya see too many engineers got involved I reckon, should done what I told em, keep it simple fellas.

Lying about the level of development of your social app is also quite different than this

Wasn't her and the companies case that they knew what they wanted to achieve, wearing the bandaid that communicates with the phone to the server was impossible and that its various simpler versions of devices did produce results that did knowingly mislead people about their health. That's is potentially life threatening and different from an i.e. chat AI company that blatantly lies, this may lead to some peoples misery as well, but Theranos had way less network hops leading to health threatening outcomes.

Absolutely. A bit more fair might be that startups often "collapse under the weight of unrealized DREAMS" :)

Agreed, but there's a non zero number of startups I had first hand experience with that I would define fraudulent without hesitation. They just didn't become as big or raised enough money to matter.