I love this piece and agree with basically everything in it, but I think there’s a bit of gap which probably seems out of scope but I think is actually very much in scope for any discussion of how organizations adopt AI.

And that is: what is people’s place in an AI world?

When you talk about shipping code without reviewing it, on the one hand we can have a measured discussion of the technical and business rewards and risks, but on the other hand someone who’s life is writing code now feels like they’re working at a Rust Belt steel mill in the 1970s. “I feel like I have a place in the world where I’m valued” is part of “what it would take to make me comfortable pushing code without reviewing it”. And that’s not a conversation about continuous integration and test coverage. (A flippant retort: Hey boss, what would it take for you to get comfortable doing a fundraising round where AI has all the conversations with no review? Should we think about rightsizing the C-suite?)

And so the selling point of AI adoption is what you call “existential threat” — that if a company doesn’t pivot to AI, it will cease to exist — and that if you as a worker don’t pivot to AI, you be unemployed and die face down in a gutter, unloved, with rats gnawing on your flesh. That’s scary! Rats have sharp teeth! But even if you buy that first threat “let’s push crayon code and iterate as we go” doesn’t exactly help people feel comfortably distant from the gutter rats.

It’s also not lost on people that some leaders seem to be excited about AI not because it lets them accomplish more but because it lets them get rid of all those pesky needy humans. (Why else would you be having massive layoffs when faced with massive opportunity?) This widens the chasm between the beach and the ice floe (amiobsoleteyet.com).

You said you’re not writing for “teams with low engineering discipline and trust” but I’m not sure there are that many teams with high levels of trust on this front. You’re not writing for “behemoths who are on the verge of busting through the red tape to finally get a Claude license”, but are you writing for teams at Meta, at Alphabet, at Microsoft?

Mind you, I have no answers to any of this, and I’m not trying to write an anti-AI polemic, I just think the shape of the world as it exists, our collective vision of the future, and our place in it need to be in the “shared reality” you’re talking about to make any progress at all…