Axel thinks AI will make life amazing. Sloane worries it could go very wrong. But they agree on one big thing: AI is about to get very powerful. A third group of experts isn't so sure. Their message: the hype is bigger than the tech.

The big question that splits them

Will AI soon be smarter than people?

Plane 1They both say YES. They fight over whether that's good or bad.

“Yes — and that's great.”

Axel

Accelerationists. Build it fast. It will cure diseases and make us richer.

“Yes — and that's scary.”

Sloane

Doomers. Slow down. We might build something we can't control.

A different argument altogether

Plane 2They don't join that fight. They question the question.

The Skeptics

“Probably not. Is AI even as smart as you both think?”

Today's AI is useful, but it's not as smart as the ads say. So the boom-or-doom fight starts from a mistake.

Skeptic group 1

“It's hitting a wall”

Chatbots have limits that bigger computers won't fix.

Headshot of Gary Marcus

Gary Marcus

Brain scientist and author. Professor emeritus at New York University.

Big idea

Chatbots guess the next word. They don't truly understand what they say, so they still make things up. He says more data won't fix that. AI needs a new design.

What his critics say

He has warned about a “wall” for years. Meanwhile, AI kept getting better.

Headshot of Yann LeCun

Yann LeCun

AI pioneer. Won the Turing Award, the top prize in computer science.

Remember the ZIP-code reader from Trail 1A? He built it.

Big idea

Chatbots learn only from text, so they will never reach human-level smarts. He says they're not even as smart as a house cat. In 2025 he left Meta to start a new company. It aims to build AI that learns from the real world, the way babies do.

What his critics say

Some say he's too quick to wave off real risks. Others say chatbots keep doing things he said they couldn't.

Skeptic group 2

“The hype is the harm”

The real danger isn't super-smart AI. It's trusting AI that isn't that smart.

Headshot of Timnit Gebru

Timnit Gebru

Computer scientist. Founder of DAIR, an independent AI research institute.

Big idea

She left Google in 2020 after a disagreement over a paper she co-wrote about the risks of AI language tools. She says talk about future “super AI” pulls attention from problems happening now. Think unfair results for some groups, and low-paid workers who train the systems.

What her critics say

Some say focusing on today's harms can hide AI's real benefits. Others say future risks deserve attention too.

Headshot of Emily M. Bender

Emily M. Bender

Language expert (linguist) at the University of Washington.

Big idea

She helped coin the phrase “stochastic parrots” for chatbots. That means they repeat patterns of words without knowing what the words mean. Calling them “intelligent” is a sales pitch. She co-wrote a 2025 book about it called The AI Con.

What her critics say

Many say it doesn't matter if AI “understands.” What matters is what it can actually do, and that keeps growing.

Skeptic group 3

“It's a normal technology”

AI matters. But it will change the world slowly, the way electricity did.

Headshot of Arvind Narayanan

Arvind Narayanan

Computer scientist at Princeton University.

Big idea

With Sayash Kapoor, he wrote the book AI Snake Oil. Some AI products don't work as promised. And even good tools take decades to change how people live and work. Big changes need people, schools, and laws to catch up.

What his critics say

AI is spreading faster than any tool before it. Calling it “normal” may make people too relaxed.

Headshot of Daron Acemoglu

Daron Acemoglu

Economist at MIT. Won the 2024 Nobel Prize in economics.

Big idea

He ran the numbers in 2024. His estimate: AI will make the economy less than 1% more productive over 10 years. That's real, but small. He says AI should help workers, not replace them.

What his critics say

Other economists predict much bigger gains. They say his math leaves out what AI will learn to do next.

Two things to notice

Quick sort: Axel, Sloane, or Skeptic?

Who would most likely say each line? Pick one.

Think about it

You don't have to pick a camp. But ask yourself:

Which camp has the strongest proof right now? What new evidence would change your mind?

Photo credits

Photos from Wikimedia Commons, cropped, resized, and shown in grayscale. Gary Marcus: Web Summit, CC BY 2.0. Yann LeCun: Jérémy Barande, CC BY-SA 2.0. Timnit Gebru: TechCrunch, CC BY 2.0. Emily M. Bender: King of Hearts, CC BY-SA 4.0. Arvind Narayanan: Robert Scoble, CC BY 2.0. Daron Acemoglu: MeJudice, CC BY 3.0.