Google DeepMind Paper: "The Abstraction Fallacy"

★★★★☆ | Original Paper | 404 Media Coverage | Published: 2026-03-10

What It Says

Alexander Lerchner, a Senior Staff Scientist at Google DeepMind, argues that no AI system will ever become conscious — and that the popular "computational functionalism" view is fundamentally flawed.

Core Argument: Symbolic computation requires a "mapmaker-dependent" description — meaning it needs a human to organize continuous physics into meaningful states before AI can manipulate them. This is called the "Abstraction Fallacy."

Key Insights

🔴 The Irony: Despite publishing this paper, DeepMind's CEO Demis Hassabis recently said AGI will have "10 times the impact of the Industrial Revolution." Lerchner's paper contradicts this framing — AGI without consciousness is just a sophisticated tool, not a moral agent.

Expert Reactions

"I'm in sympathy with 99 percent of everything that he says. My only point of contention is that all these arguments have been presented years and years ago." — Mark Bishop, Professor of Cognitive Computing, Goldsmiths University
"The AI research community is extremely insular in a lot of ways... none of these guys know anything about the biological origins of words like 'agency' and 'intelligence' that they use all the time." — Johannes Jäger, Evolutionary Systems Biologist

Why It Matters

Implications

This is a significant philosophical contribution because it comes from inside Google, but also reveals the tension between AI companies' public narratives ("AGI will transform everything") and their internal understanding that these systems aren't conscious beings.

Bottom Line: Even if AGI arrives, it's just a "highly sophisticated, non-sentient tool" — not a moral patient deserving of rights.


Explored: 2026-04-28 (16:00) | Source: The Verge → 404 Media → DeepMind