It's more than that. Finding a real-world, compelling, useful product based on a new technology is much harder than just developing the technology. This is, of

It's more than that. Finding a real-world, compelling, useful product based on a new technology is much harder than just developing the technology. This is, of course, obvious, because in both cases, you will have to develop the technology. For one, the project ends at that point; for the other, some of the most challenging work is just getting started. If you are responsible for developing a new technology, you can claim success if you release something that works as advertised, even if nobody actually uses it ("it was ahead of its time", "naysayers poisoned the market", etc.). Since there is no way to tell ahead of release whether it will succeed in the market, your project and, by extension, your position within the company creating it, are secure. Much, if not all, of the blame can be externalized. On the other hand, if your goal is to release a usable product, you've just massively increased your personal exposure. First, you have to become very opinionated because any product that tries to please everyone generally succeeds in pleasing nobody, so that means there will be people who hate it, often before it is even released. Second, it is usually possible to tell if a product is useful before it actually releases (if you aren't putting blinders on yourself), so you can find yourself in a position of failure before release, a failure that you cannot shift to anybody else. Part of the reason the iPhone succeeded was because the person who, ultimately, was leading the project was the CEO. He wasn't worried about what his boss thought, so the entire team could focus on the complete application. But that was only part. The other reason is that the product, the way the technology would be used, was part of a feedback loop that influenced the technology itself. The advantage of taking on all that risk of failure is that you end up with technology that is better, technology that is more suited for solving real problems.