It is a fine line. From practical experience: there is a class of inventors who have genuinely convinced themselves that their broken tech will work if only they get the bugs worked out, and this is always just over the horizon. They probably see themselves as doing you a favor by lying to you because you will reap the benefits after all, once the pay-day arrives.

I've seen lots of misery because of this, both with the inventors themselves, their families and employees. In one case a suicide.

It's very sad and one of the downsides of being in the business I'm in because I have come to recognize the signs of this particular ailment fairly quickly and it is a huge balancing act in both convincing my client not to invest while at the same time not pulling the rug out from under a quite possibly already unstable person.

Fraud can be intentional, or it can be the consequence of someone genuinely believing their own misguided version of reality, usually powered by some form of 'wouldn't it be great if' and associated lines of thinking.

Even so, Theranos - to me at least - seems to have crossed over into outright fraud from very early on in their lifetime. There were significant questions about the tech early on and without a good answer to those the project should have never moved as far as it did without some serious caveats about what they believed to be possible and total transparency about what had actually been achieved.

For a nice example of how gullible people are check out the Ilium thread on the homepage right now, it's a great example of how wishful thinking can cause investors to part with their money:

https://news.ycombinator.com/item?id=14167870

uBeam is another, so far they produce roughly two ecstatic press releases per year but there is no product in sight.