Early call

Igor's call · SYNTHETIC DATA REPLACES THE INTERNET

By 2030, more than 95% of the data used to train AI image and video models is synthetic. The internet stops being a data source and becomes a data output.

Resolves2030
Track record84% over 26 years
StatusPublic · call Nº 44
Every call dated & publicFull scoreboard

Igor's call, in full

The internet is not just being consumed. It is being replaced. I forecast that by 2030, more than 95 percent of the data used to train AI image and video models is synthetic, manufactured, rather than harvested from real human activity, up from around 5 percent today. Gartner backs the trajectory. At that point, the internet stops being a data source and becomes a data output. Agentic AI floods the web with synthetic media faster than humans can produce real content. AI trains on AI. The feedback loop closes. The implications for trust, journalism, and democracy are not yet priced in. When the machine learns from the machine, "real" becomes the rarest input of all.

SYNTHETIC DATA: >95% OF AI TRAINING DATA BY 2030 (FROM ~5% TODAY).

Adjacent calls

By 2030, more than 90% of social media is synthetic or artificially amplified. Authenticity becomes the scarcest thing online.Early callHuman data is the $1 trillion-a-year AI goldmine hiding in plain sight. The data-broker industry alone hits $630 billion by 2035.On trackBy 2035, people license their own data directly to AI and the $1 trillion human data market becomes the $5 trillion augmented intelligence market by 2040.Early call

Ask Igor your own question. 5 free.

Ask Igor