I'm curious: is there a bias in perception of hardware projects testing new behavior compared to software projects? I don't have enough data points, so wanted to see what you think.

By contrast to software startups, it seems, hardware project that are based on a behavioral hypothesis are considered to be R&D to be conducted in academia. Not startup material. Even though the budgets for running these early experiments now are at par with software projects (also <<$150K).

For software projects, it's quite common to run an experiment and see if a certain new behavior/environment would stick and have traction.

Do you happen to know successful hardware startups/projects that were built based on an unproven behavioral hypothesis?

Thanks!