The Singularity
We are not waiting for it. We are living through it. Why adoption alone changes everything, what the AI layoffs do and do not show, and how companies get through the transition.
October 31, 20269 min read9 chaptersBy Chris Johnston
01 · THE EDGE
Past the event horizon, no light comes back.
In physics, a singularity is where the models stop working. Nothing past the horizon can be seen from here.
02 · THE MERGE
We cannot see past this one either.
The singularity is not a machine waking up. It is people and software merging into one way of working. It already started.
1965
“The first ultraintelligent machine is the last invention that man need ever make.”
I. J. Good
1993
“Within thirty years, we will have the technological means to create superhuman intelligence.”
Vernor Vinge
2024
The Singularity Is Nearer: When We Merge with AI.
Ray Kurzweil
2026
You are here.
Inside it.
03 · ADOPTION
Even if the models never improved again.
Adoption alone changes everything. About one U.S. firm in five uses AI. The rest of the curve has not happened yet.
YOU ARE HERE
19.8% of U.S. firms, May 2026
Curve shape is illustrative. The marker is measured.
04 · AND THEY WILL IMPROVE
They will. The curve is measured.
4-5x
a year. Growth in the compute used to train frontier models, 2010 to 2024.
05 · THEN IT BUILDS ITSELF
Recursive self-improvement.
AI does the research that builds the next AI, which does that research faster. Good called it an intelligence explosion in 1965. Labs now build toward it on purpose.
Source: Sakana AI, Sept 2026 ↗
Not here yet. Today's still lack the judgment that open-ended research takes. It is where this is heading, not where it is.
06 · THE CUTS
Companies have started saying it out loud.
U.S. announced job cuts where the employer named AI as the reason.
4,2472023
12,742*2024
54,8362025
120,1362026Jan to Sep
* Derived: Challenger's 71,825 AI-cited cuts since 2023, less the published 2023 and 2025 totals.
“The adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce.”
“I think most companies are late.”
07 · NOT EVERY CUT IS AI
Some of it was the hangover.
Tech overhired after COVID and rates rose. The 2023 wave was mostly that. The 2025 and 2026 waves are where AI shows up in the filings. Both are true.
57%
Alphabet headcount growth, Sept 2019 to Sept 2022. “We hired for a different economic reality.”
293,753
Cuts in 2025 tied to DOGE, the year’s top reason. More than five times AI.
0
New York WARN filers who checked the new AI box in its first year. Over 160 filed.
08 · THE GAP
The gains are real. Society is not keeping up.
+14%
Issues resolved per hour by support agents using AI. +34% for the newest agents.
19%
Employment gap for workers 22 to 25 in the most AI-exposed jobs, against trend. Mostly less hiring, not more firing.
0
Federal laws that require a company to disclose AI’s role in a layoff.
The tools work at the task level. The work around them has not been redesigned, the people have not been retrained, and the rules have not been written. That gap is where the damage happens. It is also where the upside is.
09 · THE WAY THROUGH
Adoption is the part you control.
You cannot slow the curve. You can decide which side of the gap your company lands on. The ones that come out ahead do three things: put on real work, redesign the work around them, and train the people who run it.
That is the job we do. We build lobstersagents that run inside your business, give your team one place to run them, and stay on after launch. Every works from an instruction file your team can read and change.
You can already see adoption on the open web. Most new sites share the same machine-made look. We wrote a series on it, starting with The Hero Formula and The Little Mono Labels.
