← Video Library · AI Strategy
Mastering AI: Boost Your Job Skills, Not Your Anxiety
Published September 8, 2024 · 16 views on YouTube
Key takeaways
- Automating one deliverable at a time with a tool like Claude or Copilot teaches you the technology as you go.
- Explaining and breaking down a task for AI forces you to document it, which solves the classic 'no documentation' problem when handing work off.
- AI can take over a deliverable, but it isn't replacing the person's job outright, according to this discussion.
- Final edits, quality judgment, and reading the emotional tone of a piece of work are described as tasks AI still can't do.
Does automating your job with AI mean you’ll lose it?
Not according to this discussion — the speakers argue that automating a single deliverable with AI tools like Claude or Copilot teaches you the technology, forces you to document your own work clearly, and makes it easier to hand tasks off. They say the technology isn’t yet capable of replacing the human judgment and final review that a job still requires.
What happens when you automate one deliverable at a time?
The speaker describes a process: someone takes the time to pick one deliverable from their job, learns a tool such as Claude or Copilot, and figures out how to automate that specific piece of work. Two things follow from this. First, the person teaches themselves the technology in the process. Second, to make the AI do the work correctly, they have to dig deep into the deliverable and be able to explain and break it down clearly — which produces documentation almost as a side effect.
Why does this solve a common corporate documentation problem?
The speakers point to a familiar situation in “Corporate America”: someone gets promoted into a new job but keeps getting pulled back into their old one because there was never enough documentation to hand it off to someone else. Explaining a deliverable well enough for an AI to execute it creates that missing documentation, making it easier to actually transition work to another person.
Is AI going to replace people’s jobs?
The speaker says no, not yet — “I don’t see the technology doing that now.” They argue that expertise and a human in the loop are still required for things like making final edits, judging whether output is good quality, and gauging the emotional read someone will have when they see a piece of work. Those are described as things AI is not able to do.
Full video transcript
And so if everybody took their time, found one deliverable in their job, figured out how to use whether it’s Claude, co-pilot, whatever, and figured out how to automate that, a couple things would happen. One is they would teach themselves the technology. Second thing is they’d have to really dig into that deliverable to be able to explain it and break it down for the AI. And then what that ends up doing is it ends up creating a way for them to hand that off to somebody else.
So now, instead of — because we’ve all seen it in Corporate America — you get promoted into a new job and you still have your old job, it follows you around forever, right? Because you have to train people and there’s never enough documentation. But if we thought about things that way, just like you’re saying, make your job better, easier to do — it’s not, they’re not going to get rid of you. Not yet. I don’t know, I don’t see the technology doing that now. You still need that expertise and that human loop to be able to, like you said, do those final edits or decide if this is good quality or not — you know, what’s the emotional take going to be on somebody when they read it and see it. Those are things that the AI is not going to be able to do for you. And so I think, I think that’s important.