I was talking with someone VC-ish, about my frustration with all the endpoint hardware and communication software being hopelessly insecure for various real threat models.

But that, even if I somehow managed to pull off a successful superior solution, as a startup or an open source/hardware project, I didn't want to see all the worst criminals flock to my service.

Also, I didn't want to be in an adversarial relationship with my own government at times, nor to secretly compromise the solution.

(Probably the compromise-compromise I'd choose would be proactive: I'd have to backdoor for my government from the start, and publicly disclose that there's a backdoor, so I'm not misrepresenting to my users. Which would mean dramatically less adoption, a lot of privacy&secury people cursing it/me, and eventually the backdoor would also be exploited by parties other than the intended.)

And also, I don't have the stomach for adversaries that would include foreign state dirty-tricks agencies.

Most ostensible security solutions on the market are obviously weak, or just plain BS. The ones that might not be, I don't see how they don't run into the same barriers.