My understanding is that the main reason splitting up work is effective is context management.
For instance, if an agent only has to be concerned with one task, its context can be massively reduced. Further, the next agent can just be told the outcome, it also has reduced context load, because it doesn't need to do the inner workings, just know what the result is.
For instance, a security testing agent just needs to review code against a set of security rules, and then list the problems. The next agent then just gets a list of problems to fix, without needing a full history of working it out.
How much do you babysit claude, and how much do you just "let it do its thing"?
I haven't had anything as severe as OP, but I have had minor issues. For instance, claude dropped a "production" database (it was a demo for the hackerspace, I had previously told claude the project was "in development" because it was worried too much about backwards compatibility, so it assumed it could just drop the db). Sometimes a file is dropped, sometimes a git commit is made and pushed without checking etc despite instructions.
I'm building a personal repo with best practices and scripts for running claude safely etc, so I'm always curious about usage patterns.
Even in small companies, its important to discuss what the expectations around AI are. In the absence of any further requirements (i.e. assuming privacy is not a major issue, regulatory issues etc), it can be as simple as clearly saying: "You can use AI, but you are ultimately responsible for what you deliver. It is expected you verify the data, test the code, and otherwise validate the responses."
Something as simple as that gives an expectation, without being overbearing to start with.
This is a really important point - my file manager just says "Name" with sorting. So while its not perfectly defined, it doesn't make the promise of saying its alphabetical.
I came from an academic background, and didn't have anyone else in the team with programming experience when I was researching. I found my time as a contributor for the scikit-learn project to be invaluable in learning about the requirements to build not just working code, but more robust reliable code that others can depend on. Having my work reviewed by those with more experience, making recommendations etc, was fantastic.
So in short, see if you can contribute to a well-run open source project with a good community.
I think one method here is to incorporate your own site into the content as much as possible. For example, if you are a creator, get people to sign up to a newsletter to get the source files. Get people onto your platform/forum/whatever as well as watching through YouTube. Easier said than done, but better than not doing anything.
From there, you also ensure that you have a backup of all your videos. I've talked to people that only had their stuff on YouTube/Facebook/whatever. It is super risky. If you have a backup, and YouTube bans you, you can rehost elsewhere, it won't be as big, but you might still have a business afterwards.
I'm a big fan of Grammarly and have been using it, and paying for it, for years.
The advantage is not spell checking. It is grammar and style improvements. It tells you things like "this language is informal", or "this is a better word for that".
The problem is the rules aren't specified, if you fall afoul of a rule you do not get told which rule it is, and the rules are AI-based decisions that are hard to even understand. If the process was transparent, it would be a lot easier to deal with.
This happens regularly (perhaps not frequently) to people with their Gmail account. Google's AI says no, and there are very few people reporting successfully navigating the system to get their account back. This is one of the anti-google arguments, the other main one being they kill big projects all the time.
You are right though, it can be crippling, and this thinking part of your personal (and business) risk profile. Ideally your email is at least tied to a domain name you separately own. That way, if your Google/MSFT/Apple/whatever account gets blocked, you can switch over the email to another provider and still get access.
* edit to add: I don't know if this is more a Google thing than other companies, I was just heavily invested in the Google ecosystem and have researched these issues specifically with Google, leading to me ensuring I'm not as reliant on them as I used to be. I figure now I have about a full days worth of hassle if all my google stuff gets blocked, but otherwise I'm recoverable.
The implication here is that this is a bad thing (although OP doesn't say that). I'm not sure I think the same. If AI can be used to automate the more tedious parts of writing, that's fine with me.
For me, as long as the paper has been confirmed by the authors as being correct, what is the matter house those words were formed?
Aussie here, we try to get both the .com and the .com.au (and now also the .au), because people will just use one or the other, regardless of what the link says.
I am fine with "extra scene to setup the next season cliffhanger", but sometimes the "mainline story" isn't even finished. The season just stops at some point before the payoff, then gets cancelled in the off-season.
I'm not sure about the USA, but government jobs in Australia also are very stable with strong contracts and benefits (in general, every department is different).
For instance, if an agent only has to be concerned with one task, its context can be massively reduced. Further, the next agent can just be told the outcome, it also has reduced context load, because it doesn't need to do the inner workings, just know what the result is.
For instance, a security testing agent just needs to review code against a set of security rules, and then list the problems. The next agent then just gets a list of problems to fix, without needing a full history of working it out.