In mid-July I started posting more consistently about leadership or, better, the lack of it in the brave new world – the world of AI. On July 20th was a piece titled, “AI – Who Must Lead?” On July 22nd was another one, “AI – Who Leads”? And on August 13th was a third, “Who Do You Trust?” (The links are below.)
These are all short pieces, and none is rocket science – I am not a rocket scientist. Nor am I an expert on AI – not hardly. But I am an expert on leadership. And I am not an idiot. I can see that AI is taking off at unanticipated speed and steering us in directions that no one can predict or control. For example, this headline, this week, in the New York Times: “A.I. Stumps Handlers Trying to Contain It.” How’s that for reassuring?
Until recently scant thought was given to the sorts of questions that I raised about how to control what so far is out of control. Specifically, artificial intelligence that stumps even the best and brightest. But in the last half year this has started to change. Which is to say that I am not the only one on whom it’s dawned that “Houston, we have a problem.”
Specifically, a few of the world’s leading experts on AI have begun to think about how to lead and manage, how to control, if it’s possible, the genie already out of the bottle. Which is of course AI itself.
To this question, three of the greatest minds in technology have now devoted significant attention. They are Dario Amodei, the co-founder and CEO of Anthropic, which is positioning itself to have the biggest IPO in stock market history. And Bill Gates, the legendary founder and longtime CEO of Microsoft as well as a path-breaking philanthropist. And Demis Hassabis, the Nobel Prize winner and pioneering AI researcher and entrepreneur.
The first and most ambitious of these attempts to address the issue is Amodei’s essay of some 20,000 words, published in January. Titled, “The Adolescence of Technology,” it ranges far and wide.* Still, it’s apparent early in the essay that Amodei was motivated above all by his concern that “humanity is about to be handed almost unimaginable power, and it is deeply unclear whether our social, political, and technological systems possess the maturity to wield it.”
Though at the end of his essay, he turns to the question of how we might manage the seemingly unmanageable, he does so less unconvincingly. Amodei admits that if we want to succeed “we will have to step up our efforts.” And that it remains for us to “convince the world’s thinkers, policymakers, companies and citizens of the overriding importance and imminence of this issue.” Oh dear.
The second most ambitious of these efforts was this week when Bill Gates published his own shorter (6,000 words) but still ambitious essay titled, “The Turbulent AI Era is Here. The Choices We Make Now are Critical.”** Gates has been foolishly exiled in recent years because he foolishly got involved with Jeffrey Epstein. But the powers that be now seem satisfied that he can return from Siberia and participate in the nation’s discourse. About time. For Gates is among the most valuable human resources we have on several subjects, including, most pressingly, climate change and AI.
Gates names a few problems as the most threatening, one of which is “many jobs will disappear forever.” However, like Amodei, he does not seem, at least not at heart, sanguine that we, we humans, can solve them. Toward the end of his essay, Gates writes that “the world needs a plan.” Who can argue with that? But the best that he can do is to write that while we cannot expect AI companies “to propose solutions to challenges raised by their own technology,” we can develop solutions “through a public democratic process that includes elected officials, policy makers, educators, health workers, local officials and community leaders.”
Now there’s a good idea! OK, so which elected officials might take the lead on this? You got one in mind? Does Gates? Similarly, which health workers? Which community leaders? I don’t see a single individual who, as of now, qualifies. Who could take on the task that Gates suggests.
The only other name I’ll mention here is Hassabis’s. Unlike Amodei and Gates, Hassabis is not front and center in America. He is British born and he just resigned as chief executive at Google DeepMind. But he is deadly serious about this issue – the issue of how AI might be controlled – and he has given it much thought. Among his efforts in this regard is his proposal for an independent industry entity that would codify AI safety guardrails and best practices. He compares it to the International Atomic Energy Agency, a nongovernmental watchdog that promotes cooperation on nuclear energy.
We are past the point of worrying about a single AI agent going rogue. What we have to worry about now are multiple agents going rogue at the same time. Agents that communicate with each other and cooperate with each other while leaving humans behind. While leaving us in the dust.
You think what I describe is a high hill to climb? Think again. For the far higher hill is that whatever progress Americans might make on this issue, it will have to be matched by the Chinese. Controlling AI, leading and managing in the AI space, is not a game that Americans can, or should, ever play alone.
Did I already say, “Oh dear”?
—————————————————-
*https://darioamodei.com/essay/the-adolescence-of-technology
**https://www.gatesnotes.com/a-turbulent-ai-era-and-critical-choices-to-make