⚡🛑 AI and the energy constraint
It's not just the clean energy revolution that needs regulatory reform to make it easier to build. So does the AI revolution.
Quote of the Issue
“If humanity can get through the bottleneck of safe AGI, we could be in a new era of radical abundance, curing all diseases, spreading consciousness to the stars and maximum human flourishing: - Demis Hassabis
The Conservative Futurist: How To Create the Sci-Fi World We Were Promised
“With groundbreaking ideas and sharp analysis, Pethokoukis provides a detailed roadmap to a fantastic future filled with incredible progress and prosperity that is both optimistic and realistic.”
The Essay
⚡🛑 AI and the energy constraint
There are plenty of extremely bullish predictions about AI floating around these days — well, “bullish” if you’re someone who thinks ASAP AI progress is desirable. If these optimists are correct, humanity might well see human-level AI, or artificial general intelligence, within a decade, if not sooner. But such forecasts employ several key assumptions:
First, AGI optimists assume we’re on the right technical path. But maybe not. Yann LeCun, Meta's chief AI scientist, doubts that language-focused AI systems like ChatGPT's large language models can lead to human-level intelligence, or artificial general intelligence. He argues that language is just one aspect of human cognition, and that acquiring knowledge through sensory experiences is crucial. LLMs, trained solely on text, lack this essential component of understanding the world.
Second, AGI optimists assume the AI advances we’re seeing will become commercially useful. Even the best ideas must be entrepreneurially processed into new products and techniques. Economist Martin Weitzman once explained that “the ultimate limits to growth lie not so much in our ability to generate new ideas as in our ability to process an abundance of potentially new ideas into usable form.” If companies don’t find AI useful, they won’t pay for it, depriving companies of the resources needed to keep advancing the technology. And we’re still waiting on that, as recent pieces in The Wall Street Journal (“AI Startups Have Plenty of Cash. They Often Don’t Yet Have a Business.”) and The New York Times (“A.I. Start-Ups Face a Rough Financial Reality Check”) suggest.
Third, AGI optimists assume society would allow AGI. Let’s say that LLMs do lead to AGI that’s able to be effectively deployed by business, generating warp-speed economic growth of, say, 30 percent a year, as analyst Tom Davidson of OpenPhilanthropy speculated in a 2021 paper. Not only would living standards double every 2.5 years, but after 25 years we would be a thousand times richer. In a review of the report, Northwestern University economist Benjamin Jones noted so much “creative destruction” would occur— including rapid turnover and obsolescence in technology and worker skills — that societal pushback would be inevitable. Jones: “Governments face large challenges with the churn of creative destruction at ordinary growth rates, where displaced workers or business owners seek protection. It’s a little hard to imagine the political implications of rates of change and churn beyond all historical precedent.” So politics and what one might call “social license” are other constraints
Power up or power down
But let’s take a step back: to even get to AGI, there’s another constraint that immediately pops into mind: energy.
Keep reading with a 7-day free trial
Subscribe to Faster, Please! to keep reading this post and get 7 days of free access to the full post archives.