The story behind the index

Why I built this

I'm not an academic researcher. I'm not an AI company. I'm a parent in Israel who saw something I couldn't find a real answer to.

The moment it got personal

In late 2025, I read about a father who got a phone call from "his son". The voice was identical. The story was convincing — car accident, need money now, don't tell mom. He transferred $25,000. His son was at work, had no idea any of this happened.

It wasn't a tech article. It was on the evening news. And the next day it wasn't the lead story — because it wasn't new anymore. It was normal.

I asked myself: how would I know when something became normal? I had nowhere to check.

What didn't exist

I read everything. AGI is coming in 2027. AGI will never come. ChatGPT changed everything you think. ChatGPT is a bubble. Everyone is right. No one is right. All superlatives, no numbers.

I was looking for one simple thing: a single number for where AI is today. Not a prediction. Not spin. Just "where we are, on a 0-to-100 scale, based on evidence."

It didn't exist. So I built it.

What this is

The AI Index is one number, 0 to 100, updated twice daily. It measures AI progression across 4 dimensions: capability, autonomy, integration in critical systems, and control bypass.

40 sources are scanned twice daily: Anthropic, OpenAI, Nature, MIT, Reuters, Bloomberg, and more. Only events from the last 12 hours count. Minimum impact threshold: ±0.2. Max 8 drivers per day. Every event links back to its original source.

No imagination. No guessing. If it didn't happen and isn't documented, it isn't counted.

What this isn't

It's not an AGI forecast. I don't claim to know what happens in 2027. I measure what happens today.
It's not AI hype. A "breakthrough" with no real-world impact won't move the score.
It's not AI doom. No evidence, no points. The score can drop — and has — when regulations pass or bugs get patched.
It's not an AI company. I have no NDAs, no kickbacks, no investment in any specific model.
It's not personal cyber consulting. I provide knowledge. Defense is your job.

4 principles

⚖️

Independence

No funding from AI labs. No partnerships with AI companies. Subscription revenue only.

🔍

Transparency

Methodology fully public. Every driver linked to source. Code open for audit.

🔁

Reproducibility

Automated scanning, twice daily, identical conditions. You can verify any score yourself.

🎯

Practicality

Every event has a 'what this means for me'. Not all information — only what's actionable.

I don't know how this ends.

I don't know if the score will be 60, 80, or 100 in two years. I know I'll be here, twice a day, measuring what is.

If it's useful to you — join the waitlist. If not — at least take the defense guide. It's free.

Join the waitlist →