Job or solo business or entrepreneur Daniel Isenberg wrote this test in an HBR blog post to help you decide - be honest.
Follow @Lucas1Mike for more insights
Your manager can generate an AI report in 30 seconds.
3 things to put in yours that AI can’t fake:
1. Insider knowledge nobody wrote down
AI knows your industry. It doesn’t know your company.
It doesn’t know the workaround everyone uses, the step where deals quietly stall, or the unwritten rule that blocks the obvious fix.
The people doing the work know. Ask them what they complain about at lunch.
2. A bold idea with your name on it
Small tweaks are a commodity. Nobody argues with them.
Take a position that makes people push back. If most readers agree, it’s too safe.
Aim for more disagreement than agreement.
AI can summarise. It can’t stick its neck out.
Bold, but backed by what you found.
3. A first step for Monday
Generic reports end with “explore AI opportunities.”
Yours should end with something a real person can do this week, given real constraints.
AI made generic knowledge free.
The gap now is knowing the problem well enough to see where AI fits.
Pick one problem. Research it for two weeks. Write what you think.
Then put it in front of someone who can act.
What’s the boldest idea you’d put your name on?
Lose-Lose.
The interesting part of Microsoft offering exits to a quarter of its staff isn’t “AI already did their jobs.”
It’s that salary savings are being redirected to fund the enormous AI infrastructure bet.
You lose.
The longer-term wager is still that AI becomes a substitute for a large fraction of today’s tech workforce.
You lose.
Banking is the perfect target for AI.
3 reasons banking is the perfect target for AI:
Banks aren’t adopting AI because it’s trendy. They’re doing it because of the maths.
1-A large people cost base. Salaries and benefits typically make up 45-60% of a bank’s noninterest expense.
2-Repeatable processes. Approvals, reviews, reports and exceptions follow patterns AI can learn.
3-A deep and growing IT foundation. The infrastructure to plug AI in is already there.
Job losses were always going to happen. Banks that don’t build on AI will watch startups eat their lunch and dinner.
Follow for AI + Job implications through an investor lens.
Source: FFIEC Call Report data via BankRegReports (link in comments)
Anthropic is reportedly preparing to tell IPO investors its total addressable market is worth over $30 trillion, based on the full scope of work AI models could complete. (Yahoo Finance)
Post-IPO, Anthropic's revenue = tokens bought.
Tokens bought = budget to be found.
Biggest budget = your monthly pay.
If mathematics is a craft.
Will AI release the bottleneck?
However highly we rate a skill, if the method can be described and the result checked, it can be industrialised.
Which part of your own job fits that description?
Your company calls the layoff AI. You lose the job.
Leadership usually keeps theirs.
That can be the plan.
A layoff labelled AI protects leaderships career twice.
It keeps the overhire off their record. And it puts an AI plan on it.
History shows us that when a new technology needs huge physical assets — rail, power plants, fibre, now GPUs and data centres — the technology usually wins, capacity gets overbuilt, and "early equity" often loses even if the industry later thrives. Someone still makes a killing. They buy the wreckage after the crash.
The broken thinking about charging user lienses is that most bottlenecks in companies are the users. Skill automation is historically what has produced wealth and net-new opportunity at the macro level. But if you were a weaver (Luddite), your prospects weren't augmented. They collapsed. Skill was the bottleneck to progress, and automation removed it.
Everyone says AI will augment your skills.
History says automation exists to remove them.
There's a popular argument going around: focus AI on augmenting humans, not automating them away.
Automating existing things doesn't create value — it just makes things worse.
Go too far and you get an "AI automation paradox."
The Industrial Revolution says otherwise.
Skill automation is what produced wealth and net-new opportunity at the macro level. But if you were a weaver, your prospects weren't augmented. They collapsed. Skill was the bottleneck to progress, and automation removed it.
That's really what automation is. A process of deskilling.
Not every case looks the same. Elevator operators lost their job to automation too, but the skill that got removed was the operator's, not the rider's. The rider's experience actually improved. It just wasn't the operator who benefited.
We talk a lot about what we want progress to look like. Progress doesn't run on what we want. It runs on removing whatever bottleneck is currently slowing things down.
Nobody wanted the weaver's skill to become worthless. It became worthless anyway. And the same people who don't want their own skill devalued still want the lower prices, the faster service, and everything progress produces on the other side.
You can't have one without the other.
That's the part business won’t say out loud.
Follow to see AI through an investor's lens. I watch as an investor. I write for the employee. The investor lens is just how I get to the truth before the PR does.