It's about proving you're doing it right.

Some technological wonder working in a way the observer doesn't understand may be perceived as 'magic'.

I think the same holds true for AI.

For problem fields seeming easy for humans, but hard for machines, the solution must be AI.

If a solution is found, the solution suddenly consists of procedures, if's, else's, and that dispells the magic.

So, in a way, AI may never exist.

No, this is the only way to create useful documentation (provided it's for "internal use only" and an unspecified readership). It's no use (re-) writing "network admin for dummies". There's only one thing worse than no documentation - outdated documentation.

Just write down, what you need in a form you can use. Chances of anybody else reading through these papers is close to zero: Your _potential_ successor has

• to know that documentation exists
• to be able to find it
• to think it contains anything valuable
• to have time to read it
• to habe the capability to understand it and make use of it

99,99% of all known possible successors will just hotfix problems as they arise and blame everything on the predecessor. So just write up the things you need and tend to forget in a way you can use...

You're not talking about the food, are you?

But doing opensource programming might be the road to survival for some. Even without job, becoming a major contributor of a suitable open source project just might

• keep your skills sharp
• look good on your resume
• get you noticed by interesting companies