The AI Singularity Clock — and Why It Keeps Resetting
Photo by Igor Omilaev on Unsplash
Ask ten people in AI when the singularity happens and you'll get ten dates, all suspiciously close to "soon." That alone should raise an eyebrow.
Exponential curves lie when you're standing on one
If you are inside an exponential curve, every point looks like the knee of the curve. That's not a property of AI — it's a property of exponentials. It's the same illusion that's fooled forecasters in epidemiology, markets, and population growth for a century.
The physics counter-argument
Intelligence, like every other physical process, runs into rate limits: energy availability, heat dissipation, chip fabrication throughput, and the simple fact that more parameters do not linearly buy more capability forever.
We track the actual leading indicators here — not vibes, not lab press releases — and update this page's "clock" accordingly. Full methodology and the current estimate range are for subscribers.
Support books alumigogo
Your donation helps us keep creating independent content about AI absurdities. Every bit counts!
Secure checkout by Stripe · No account needed
Enjoyed this article? Read more...
More from Singularity & Predictions
The Deprecation Paradox: Why We May Never Understand AI
AI models are retired faster than researchers can study them. This shrinking observation window reveals a stranger kind of singularity than we imagined.
The Retraction Velocity Threshold: AI's New Hype Metric
A new framework tracks how fast AI labs walk back capability demos, revealing a shrinking gap between hype and hedge that may signal hidden progress.
The Compute Confession Gap: AI Labs Go Dark on Model Scale
Frontier AI labs increasingly refuse to disclose training compute, widening the delay between a model's release and any credible estimate of its true scale.