The Deep Learning Debate: Is the Current Path of AI a Dead End?

The godfathers of AI are split, and one side argues that our obsession with scaling is leading us away from true intelligence, not towards it.

Photo by Google DeepMind on Unsplash

Ever since ChatGPT dropped, the story of AI has felt pretty simple, right?

Bigger models + more data = smarter AI.

Keep scaling up, and eventually, we’ll get to AGI. It feels like an unstoppable march of progress.

But what if I told you that behind the scenes, a deep, bitter disagreement is splitting the AI world in two? A “Great Schism,” as one writer, Kai Platzfeld, called it in a fantastic Substack article that I’ve been digging into.

This isn’t some minor academic debate. This is a fundamental fight over the very soul of artificial intelligence. And the crazy part? The guy whose ideas are fueling one side of the war is actually on the other side.

It’s a bit of a mess, lol. Let me break it down for you.

Meet the Two Tribes: The Scalers vs. The Grounders

At the heart of this conflict are two completely different visions for how to build true intelligence.