User Avatar

👋I'm Hannah! I help innovative parents + their teens understand and gain the critical skills they'll need in a future where AI powers the world.

ChatGPT scored in the 90th percentile on the SAT: Is it time to rethink how we measure human intelligence?
Hannah Grady Williams

It was February 14, 2011.

A date long-anticipated by millions of eager Jeopardy enthusiasts. The day that Jeopardy legends, Ken Jennings and Brad Rutter would face off with a robot, Watson.

"We've got this in the bag..." Jennings remembers thinking. "I've trained all my life for this... there's no way a robot can beat me."

But we all know how this story ends...Watson absolutely crushes Jennings and Rutter. $77,144 in winnings compared to Jenning's measly $24,000 2nd place.

How did this happen?

How did a machine replace the best Jeopardy players in the world? The documentary Most Likely to Succeed asked this fundamental question in the aftermath of Watson's victory:

"In the AI era, is fact-based recall, memorization, and drilling still a good way to measure student intelligence and learning?"

Their conclusion? No! Of course not! In a world where Watsons exist, what AI can do perfectly and for free should NOT be the way we train humans any longer.

That was 13 years ago. Yet we're still teaching kids obsolete things.

It's 2024.

ChatGPT4 just scored in the 90th percentile on the SAT. It's expected to 8X in potential each year. If humans want to remain relevant, we must stop teaching obsolete things to our kids. And start teaching different things.

How else could we measure human intelligence and capabilities if tests aren't the answer? Here's a starting list...

  1. Portfolios. Evidence showing IMPACT, not knowledge.

  2. Public Exhibitions.

  3. Peer Reviews + Collaborative Assessments

What else would you add to this list? Let me know in the comments 👇👇