Is AGI Really Just Around the Corner?

Ah, Artificial General Intelligence (AGI), the tech world's favorite game of 'Are we there yet?' It's the Holy Grail that promises to transform everything from how we work to how we binge-watch TV. Recently, Shane Legg of Google DeepMind, someone who's probably forgotten more about AI than most of us will ever know, tweeted that AGI is on the horizon. But before we all start stockpiling canned goods and learning how to barter, let's unpack what this really means.

The Tweet Heard 'Round the World

Shane Legg, co-founder and chief AGI scientist at Google DeepMind, recently tweeted about AGI being on the horizon, specifically suggesting a 50% chance of achieving minimal AGI by 2028. You can find the actual tweet Shane Legg's tweet on X, because, as they say in journalism, "trust, but verify." While Legg's tweet might seem like the kind of optimistic tech prediction you'd find in a late-night infomercial, it carries more weight given his credentials and proximity to the cutting edge of AI research.

Hiring for the Future

In addition to the tweet, there's chatter about DeepMind looking to hire a senior economist to explore the post-AGI economic landscape. This isn't just your run-of-the-mill "let's hire someone to make pie charts" kind of gig. The job posting, which can be verified DeepMind's careers page, indicates they're thinking seriously about how AGI could upend our economic systems—kind of like how the internet changed the way we pretend to work at our desks.

The Skeptic's Guide to AGI

Now, before we all start writing think pieces about robots taking over our jobs (and potentially our pets), it's important to note that AI systems today still lack key cognitive abilities, such as reasoning and long-term memory. Think of them as brilliant but forgetful teenagers—great at solving complex math problems but not so hot on remembering why they walked into a room.

DeepMind's move to hire an economist suggests they're planning for a future where the traditional economic equations (like productivity equals labor times hours worked) no longer add up. We're talking about a world where labor markets could look like the Wild West of the 1800s, but with more algorithms and fewer tumbleweeds.

A Walk Down Memory Lane

If you've been around the tech block as often as I have, you'll recall that this isn't the first time we've been promised a transformative breakthrough. Remember when Virtual Reality was going to change our lives? Or how about when everyone thought the Segway would revolutionize urban transport? Spoiler: It didn't. The point is, tech predictions often come with a side of hyperbole.

The Road Ahead

So where does that leave us? As with most things in tech, the journey to AGI is fraught with unknowns. Shane Legg himself admitted that timelines are notoriously hard to pin down, saying, "AI progress is inherently difficult to predict." And he's not alone in that sentiment. The landscape is riddled with "unknown unknowns," to borrow a phrase from a certain early 2000s political figure.

In the end, whether AGI arrives in 2028 or 2048, the conversation around its implications is worth having now. After all, as any good sci-fi movie from the '80s will tell you, it's better to be prepared for the future than caught off-guard by it.

As we continue to edge closer to a world where machines might think like us, perhaps the real question isn't, "When will we reach AGI?" but rather, "What will we do when we get there?"

By Mike Sullivan