Sam Altman: The Visionary or the Hypocrite?
If you’re not living under a rock, you probably know about Sam Altman’s $7 trillion plan to massively increase global production of semiconductor ships for the AI industry.
The goal is a massive one. Here are some stats from Vensy on X to put that into perspective:
- $0.53 trillion: 2023 Global semiconductor sales
- $3.2 trillion: Microsoft’s market cap (the largest company in the world, as of 10/04/24)
- $4 trillion: the US’ cost for WW2, adjusted for inflation
- $6.13 trillion: US’ 2023 federal budget
- $18.0 trillion: China’s 2022 GDP
The move has been met with widespread criticism. Gary Marcus, a prominent psychology and AI scientist, has publicly lambasted the proposal. He’s raised seven reasons on his Substack page on why the world should ignore Altman:
the strain on the world’s energy and climate, natural resources, economic resources, financial risk, impositions on human intellectual capital, and negative externalities.
These issues with the proposal are strikingly obvious.
Somewhat unsurprisingly, that’s actually prompted some critics to argue that Altman’s not actually serious about the proposal. Sam Lessin, an early Facebook exec, dismissed Altman’s move as playing a game of “one-upmanship” with Elon Musk, where both sides are “making promises and clearly absurd but exciting statements which select for meme-propagating-cult-following loyalists”.
But Altman has shown, at the very least, some level of seriousness regarding this proposal. He’s been in talks with various investors, including TSMC, Softbank, and the United Arab Emirates government, to raise that money.
He’s also taken to X to dismiss his critics as anti-progress, skeptical laggards who are too blind to see the promise of Gen AI. He retorted in a one-liner X post:
you can grind to help secure our collective future or you can write substacks about why we are going fail.
Fair enough, Sam. Let’s take your arguments seriously. But they’re frustratingly biased to address.
What does “secure our collective future” even mean? It’s such a vague, nebulous, and unsubstantiated term.
Altman’s arguments are built on the implicit assumption that AI will make our future more secure. He begs the question as follows: Surely, if AI were guaranteed to make our future secure, there’d be no reason why we won’t all be chasing after it with all our collective energy right?
The problem with Altman’s claims is, that assumption just doesn’t hold true.
As Gary Marcus retorts back at Altman:
“$7T might serve OpenAI well, and raise your public profile, but would it serve humanity? How certain of that bet can we be? What’s the rush? (Also, what has the $100 billion grind on driverless cars brought society? Might people have rushed in too soon, technology-wise?)”
True, LLMs and other AI applications have brought significant benefits to our lives. But they also have significant risks. And plowing such an unfathomable amount of money into them without any sort of clear-cut guarantee of their benefit to humanity is obviously worrying.
Altman’s recent comments also don’t line up with his past comments on AI, which have generally been cautious and stressed the need for AI regulation. He has repeatedly acknowledged that if AI “goes wrong, it can go quite wrong”, and that the AI industry needs to “work with the government to prevent that from happening”.
In a February 2023 post, OpenAI’s blog itself has also stressed the need for a “gradual transition” for AI which “gives people, policymakers, and institutions time to understand what’s happening”.
But it seems that Altman has changed his tack on AI in the past year. By past-Altman’s logic, we would need to have significant global AI regulation in place to launch anything close to current-Altman’s $7 trillion plan.
But there’s been no comprehensive global regulation launched in the past year. Sure, some regional standards (e.g., the EU AI Act) have been implemented. But many other places still lack regulation anywhere near stringent enough to deal with the potential dangers of advanced AI.
Is a $7 trillion plan to turbocharge AI chip development really a “gradual transition”? I think not. Altman has seemingly forgotten his previous desire for regulation. Instead, he just points to a vague notion of securing humanity’s “collective future” as justification to let him do whatever he wants. That’s precisely the thing he was cautioning against just over a year ago.
That’s not to say that I think Altman is an inherently greedy, evil tech genius seeking to conquer the world for his own personal benefit. But within a span of a year, he has somehow managed to adopt a grandiose “rules for thee but not for me” hero complex. (Or, he revealed the complex that was always there).
Like many other tech CEOs, Altman was a particularly precocious student. At eight years old, he self-taught programming and disassembling Macintosh computers. At eighteen, he enrolled into Stanford to study computer science, before shortly dropping out to launch Loopt, a mobile app which shares locations, as Y Combinator’s first batch of startups. He so impressed Y Combinator Co Founder Paul Graham that he eventually became a Y Combinator partner, and eventually its President.
The trope of the precocious, intense, and successful techie is almost overused by now. Steve Jobs, Bill Gates, Elon Musk, et cetera.
All these men had a huge impact on the world, reinventing our way of life and turbocharging new innovation. In the process, they cultivated huge egos and hero complexes.
There’s nothing wrong with that in isolation. The issue arises when such techies are given the freedom to fervently pursue new technologies with little regard for their associated dangers, and no regulation to restrain them.
No one should be given the power to do whatever they want with little to no oversight.
This is especially true for a realm like artificial generative intelligence, which poses large existential risks to humanity at large.
Sam Altman touts his benevolent aims, and he may very well believe in them. But at the same time, we cannot as a society let him do whatever he wants, without sufficient regulation. It’s simply not fair or safe to allocate so much power to so few people.
Past Altman has warned us about the need for AI regulation. We must heed that warning.
If you enjoyed this article, please follow me here on Medium for more wide-ranging stories about the ever-changing business world.
