The Coming Age of AI-Generated Knowledge

Imagine opening tomorrow's newspaper to find nearly every article written not by journalists, but by artificial intelligence. That future might be closer than we think, according to NVIDIA CEO Jensen Huang.

The 90% Prediction Speaking recently on a tech program, Huang made waves with his forecast: "Within two to three years, about 90% of new knowledge worldwide will be synthesized by AI." For context, that's roughly the timeframe between now and the next presidential election.

"People hear this and immediately panic," Huang acknowledged. "But is it really so different from learning from textbooks written by authors we'll never meet?" He argues the critical issue isn't the source of information - human or machine - but its reliability.

ChatGPT's Unique Teaching Ability Huang spotlighted ChatGPT's groundbreaking feature: users can ask the chatbot directly how to use it. "No tool in history has had this capability," he noted. Beyond multilingual communication, ChatGPT can rapidly learn new languages mid-conversation to better serve users.

This advancement promises to democratize technology access dramatically. Soon, mastering complex programming languages like Python might become optional rather than essential for interacting with computers.

The Verification Imperative As AI-generated content proliferates, Huang stresses verification remains paramount: "We'll still need human judgment to separate fact from fiction - just as we do today with traditional media."

The implications ripple across education, journalism and research fields already grappling with AI disruption. Will future historians mark this decade as when machine-generated knowledge surpassed human output?

Key Points:

• AI knowledge dominance: Upcoming years may see artificial intelligence producing nearly all new global information
• Reliability over source: Verifying facts remains crucial regardless of whether content originates from humans or machines
• Natural language revolution: Future computer interactions may require simple conversation rather than coding expertise