Then came AI. Suddenly he’s building in minutes what used to take days or weeks. Economically, that’s massive progress. For him personally, it still feels like a loss, because the part of the job he loved most has suddenly lost its weight.
When AI hits identity
On the surface, the video is about software. Underneath, it’s about a much bigger question: What happens to us when the work we've built our identity around suddenly becomes automatable?
Someone who has defined themselves for years through writing good copy is noticing that AI does it pretty well too. The person who took pride in building beautiful slides is watching drafts appear in seconds that don't look much worse.
SEO analyses, translations, briefings, campaign concepts — a lot of this can be at least partially automated today. The results aren’t always brilliant, but they're good enough to shake your sense of self. That's what came up in most of the conversations after my talk.
Many companies are jumping to a quick conclusion here.
“If AI handles execution, we just need more generalists. Everyone should think broader. A bit of strategy, a bit of data, a bit of everything.”
There’s something true in that logic, but as a message it’s dangerous. To a lot of specialists, “become a generalist” sounds like “nobody here cares about your depth anymore.” That’s simply wrong. We actually need more depth than before, just in a different form.
Depth stays. The reach gets bigger.
This is the real point for me. A good specialist doesn’t get flatter in the AI world. They get broader in their impact. They stay deep in their craft and can move much more of the surrounding work themselves, because AI hands them tools they used to need other roles for.
- A good product manager can pull together research, a first data analysis, a prototype, initial copy, and a go-to-market sketch much faster on her own.
- A good marketer can build a complete loop from an idea in days, from campaign logic to landing page to reporting, where it used to take weeks and three different tools.
- A good developer doesn't just write code. They test architectures faster, generate documentation, and try several solution paths in parallel.
This has little to do with generalism. This is specialist work with much wider reach.
I think this effect is underrated. Most projects in companies don't fail because the individual specialists are weak. They fail because of the friction between them: the briefings, handovers, misunderstandings, waiting times, another round of alignment, yet another meeting.
That’s the invisible tax organizations pay, and AI lowers it noticeably. Not everywhere, not completely, but enough to change how far one person with good context can get before they need anyone else.
And that’s exactly why judgment matters more.
This sounds liberating at first, and it is. But it comes with a hard prerequisite. If I’m suddenly moving more of the whole, I also need to judge better what’s actually good. Otherwise I’m just producing average faster.
In Silicon Valley they call it “taste.” In plainer terms, I'd call it judgment.
The ability to tell whether a result is genuinely good or only polished.
Whether a piece of writing carries weight or only sounds professional.
Whether a strategy makes sense or only reads as plausible.
Whether an argument is strong or only well packaged.
It sounds soft. It’s actually brutal, because that’s the one thing AI can't do reliably. The more it produces, the more it matters who decides what’s worth keeping.
Which leads me to maybe the most important sentence I’d give my team right now:
.png?upscale=true&width=1116&upscale=true&name=Issue%20%236%20Newsletter%20NA%20TWR%20Header%20(1).png)
.png?upscale=true&width=1036&upscale=true&name=2026.06%20The%20Workplace%20Rewrite%20Issue%20%23006%20Imagery%20(1).png)
