It goes beyond upfront accreditation. Doctors and lawyers get accredited credentials, sure, but more importantly they're also subject to strict regulations around malpractice. You can BS your way through most degree programs, but you can't BS your way through a law or medical practice because sooner or later you'll wind up doing something illegal and be stripped of the right to practice.
In software, it would not be enough just to require a degree. We'd have to also have some auditing entity that ensured the software written in practice conformed with the standards of professional engineering. And as the industry stands today, that applies to 0% of software created in a professional setting, never mind that very few corporations would be cool with auditors poking around in their code, at least without strict NDAs.
Not that I'm arguing that we should go this way. I for one embrace the wild west era of software for as long as it lasts. But if we were to give accreditation any weight, it would have to come accompanied with a standardized method to verify the accreditation is actually meaningful.
I suppose it has more to do with developers dealing more with logical systems than with physical and legal systems that have more direct and immediate consequences. In the case with doctors and lawyers, a bad process will have a more visible impact. In software, there's lots of "bad code" that still operates with the expected outcomes, regardless of the processes used (or ignored) that took them there.
Auditing and forensics does seem to get applied more frequently in cryptocurrency projects, but as you said, that sort of accreditation still has to be meaningful.
In software, it would not be enough just to require a degree. We'd have to also have some auditing entity that ensured the software written in practice conformed with the standards of professional engineering. And as the industry stands today, that applies to 0% of software created in a professional setting, never mind that very few corporations would be cool with auditors poking around in their code, at least without strict NDAs.
Not that I'm arguing that we should go this way. I for one embrace the wild west era of software for as long as it lasts. But if we were to give accreditation any weight, it would have to come accompanied with a standardized method to verify the accreditation is actually meaningful.