Let’s talk about DeepSeek a little more, since it’s all anyone seems to be talking about anywhere right now.
On Monday, China shocked the world when a company announced that they’d produced an AI to rival ChatGPT and Claude in much less time (two months) for much less money ($5 million) and with much less advanced chips (sorry, Nvidia) than its American competitors.
The tech world was a bloodbath all day Monday, with Nvidia in particular hardest hit, losing nearly $600 billion (with a “b”) in market value after the shocking announcement.
The move sent shockwaves through Silicon Valley, and everyone had something to say.
One of the most succinct and memorable descriptions came from venture capitalist and outspoken Trump ally Marc Andreessen, who described the debut of DeepSeek as “AI’s Sputnik Moment.”
If you’re missing the analogy, here’s a quick history lesson:
Sputnik 1 is the name of the first manmade satellite to enter space and orbit the earth, and it was produced by scientists in the Soviet Union, launching in October 1957.
At the time, everyone was aware there was a nascent “space race” between the U.S. and the U.S.S.R., but it was generally assumed that the same country that had been far ahead in producing nuclear weaponry would be far ahead in the space race.
The launch of Sputnik changed all that. It sent America a crystal clear and chilling message: “kick your space program into high gear, or lose your spot as the world’s leading superpower.”
Twelve years later, Apollo 11 would become the first crewed mission to land on the moon, reestablishing the U.S. as the leading power — but without the “Sputnik moment,” many hypothesize we might never have gotten there.
Now, how good an analogy is that for the DeepSeek moment? There are a few things to keep in mind.
First, like the launch of Sputnik, the launch of DeepSeek was clearly planned by China to shock and embarrass the U.S.
DeepSeek went from not existing to leading the Apple AppStore charts in a matter of days. But China must have been working on some version of this secretly for years.
With that said, there has been remarkably little suspicion directed towards the claims about DeepSeek coming out of China, a country that we know lies, steals, manipulates numbers, deceives, and does a whole lot worse to pursue a competitive advantage.
Without reading any headlines, ask yourself this question: “Do I really think it’s possible that China produced a better rival to ChatGPT with worse chips in a matter of months and for just $5 million?”
Phrased like that, if it doesn’t trip up your B.S. meter, you might need to have yours calibrated.
This tweet puts it fairly well:
You don’t have to agree with everything he says to recognize that China is a hostile player on the world stage trying to establish dominance in the most important technology of the future.
“Trust, but verify” seems to be the bare minimum standard here, but many are skipping the “but verify” part (not that China would allow anyone to “verify,” anyway).
With Sputnik, there was an actual, metal orb circling the globe. You could see it with telescopes. It was relaying radio messages back to planet Earth.
Yes, DeepSeek’s code is “open source,” meaning you can technically see the code. But no one has a lens into the actual amount of time, research, effort, and resources that went into producing, nor does the code reveal how the AI model has been trained and developed, i.e., to censor information about the Tiananmen Square Massacre.
And while Palmer Luckey (from the tweet above) might not speak for everyone, he’s not the only one to question the $5 million number claim…
Now, with all of that said, I think we should hope that this is America’s Sputnik moment on AI.
When Sputnik went into space, it snapped the U.S. out of a slumber, galvanized and unified its smartest people, and produced incredible, world-changing results that inspired humanity across the globe and changed the course of the Cold War (by the way, if you want to see an amazing depiction of the effect of the first Moonwalk, watch the episode of Netflix’s The Crown on the topic. Season 3, Episode 7).
We should hope that the same effect plays out in A.I.
As I’ve talked about recently, the potential for A.I. is virtually limitless, but it needs the best and brightest minds shaping it.
It can be used for good, or evil, so the U.S. needs to be on the forefront and in the driver’s seat, restraining the worst impulses of A.I. and making it the greatest transformative force in the history of humanity.
John F. Kennedy famously gave a speech at Rice University in September 1962, on why the U.S. had announced plans to land on the moon within the decade (which, at the time, seemed impossible).
His passionate prose seems as fitting for the A.I. race now as it did for the space race back then.
“We choose to go to the moon. We choose to go to the moon in this decade and do the other things, not because they are easy, but because they are hard, because that goal will serve to organize and measure the best of our energies and skills, because that challenge is one that we are willing to accept, one we are unwilling to postpone, and one which we intend to win.”
Winning the A.I. race will be hard, but the challenge should be one that we are willing to accept, unwilling to postpone, and intending to win.
To your prosperity,
Stephen Ground
Editor-in-Chief, ProsperityPub
No comments:
Post a Comment