Every AI post on LinkedIn sounds like it was written by a middle manager who just discovered ChatGPT and immediately blacked out and wrote 800 words about transformation.
I’ve read a lot of them. I read them so you don’t have to, which is very selfless of me and I’d appreciate it if someone acknowledged that.
The Vocabulary of Doing Nothing
“We’re leveraging AI to drive alignment and unlock value across our synergy stack.”
What the fuck does that mean?
What did you build? What problem does it solve? What happened before you built it versus after? Show me one thing. One actual thing. I will wait.
(They will not show you. There is nothing to show. The post is the thing.)
This is the linguistic register of organizations that are performing AI adoption without doing AI adoption. Every sentence is technically true and completely empty. It could describe any initiative at any company about any technology in any decade.
“We’re leveraging blockchain to drive alignment.” “We’re leveraging the cloud to unlock value.” “We’re leveraging enterprise software to synergize our stack.”
The words change. The emptiness is consistent.
The Dinosaur Situation
Meanwhile, the actual AI researchers are publishing papers about systems that can rewrite their own code. Models are getting better at reasoning faster than most companies are getting better at writing about reasoning. The capability curve is going somewhere that is genuinely difficult to look at directly.
And LinkedIn is out here posting “Top 5 Ways to Leverage AI for Your Q3 OKRs.”
We’re dinosaurs laughing at the asteroid. The asteroid is not impressed by your content calendar.
I’m not on team doom. I’m not particularly worried that robodaddy is coming for everyone’s jobs on a Tuesday. I’ve been building with AI tools long enough to have a realistic sense of what they can and can’t do right now, and “replace most knowledge work immediately” is not on the current menu.
But I am also not on team “this is fine, the hype will pass.” The hype will not pass. It will deepen. The people who are actually using these tools, actually building with them, actually deploying them in production, actually measuring whether they work, are accumulating a real advantage over the people who are posting about transformation.
The gap is going to get embarrassing.
What AI Actually Looks Like
Here’s what AI looks like when someone’s using it for real.
Last year I wrote a script at 11 PM (exhausted, slightly annoyed, not trying to build anything impressive) that checks whether a customer might churn by looking at six signals: login frequency, support ticket volume, feature adoption, NPS trend, days since last outreach, and renewal date proximity. Each signal gets weighted. The weights I stole directly from my fishing algorithm, which checks barometric pressure, moon phase, and wind speed to tell me whether to go kayaking.
That script runs every morning. It produces an HTML report. I look at which accounts are red, which are yellow, which are fine.
It’s ugly. The code is embarrassing. It has comments like ”# this is a hack” and ”# TODO: fix this properly” that are approximately one year old.
It works. Nobody’s posting about it on LinkedIn because it’s not sexy. It’s just useful.
That’s the whole thing. That’s the entire AI revolution for people who are actually doing it: a lot of ugly scripts that solve specific problems, running quietly in the background, making it possible to do things you couldn’t do manually.
The Honesty Section
I’ve built 1,300 of these things. Not because I’m some kind of AI visionary. Not because I have a strategy for competitive differentiation through technology.
Because I have ADHD and I am constitutionally incapable of doing the same boring task twice without wanting to set something on fire. That’s it. That’s the whole story.
I say that like it’s someone else’s story. It isn’t. My brain does not check Slack unless something forces it to. I built a script that forces it to. That’s the origin of roughly a third of my automation portfolio: not innovation, just compensating for the things my brain refuses to do unprompted.
I am automating my own incompetence. When the robots take over, that’s what they’ll be taking over: the parts of my job I was already too bad at to do well. The AI apocalypse started not in a boardroom but in the inbox of someone who kept forgetting to follow up.
What Actually Matters
The companies that will be meaningfully ahead in three years are not the ones with the best AI strategy documents. They’re the ones with the most people who’ve shipped AI-assisted systems into production, learned what works and what doesn’t, and shipped more.
Learning is not in the document. It’s in the doing.
The document says “leverage.” The work ships on Tuesday.
Stop writing transformation posts. Build something small. Measure whether it works. Fix it if it doesn’t. Ship something else.
The asteroid doesn’t care about your Q3 OKRs.
Blake Bailey runs Bailey Business Ventures, an AI transformation consulting practice. 1,300+ production automations. 9 years building SaaS operations at scale. Currently extremely tired of LinkedIn.