Keep Calm and AI On

Are you a glass half full kind of person? Or does your glass tend to be half empty? I’m of the half full variety and it seems to be part of my natural temperament. And I also pride myself on being a realist (“an optimist with a pinch of pessimism”), and I don’t shy away from hard truths about myself or the world.

Unless you’ve been incredibly effective at avoiding it, you’ve been hearing a lot about the rise of artificial intelligence and machine learning ... and may have recognized the glass half full/half empty dynamic on full display. On the one hand is the latest techno-utopian vision coming out of Silicon Valley telling us that AI will solve all our problems and usher in the Jetsons age we’ve all been waiting for; on the other hand is doom, gloom, the rise of the robots, Terminator in real life, and humanity’s days are numbered. Both are problematic.

What are we to make of this? Earlier this week, I sat down with one of our own Providence parents, Leonard Arul, a Principal Program Manager at Microsoft who works with AI every day. With his wife Subha, he recently presented his insights to our Technology Program students. In our wide-ranging conversation, we talked about his work at Microsoft, AI tools, and their exciting possibilities and potential dangers. When I asked what he wishes every Providence parent would know about AI, Leonard said, “First, I think we all, as parents, need to start conversations with our kids about emerging technologies like AI, and explore together the ethical implications and desired boundaries for these technologies. Second, much like preparing our kids for the responsibilities of driving, at some point we have to get in a car, put our kids behind the wheel, and teach them how to drive. I think we all need to try these platforms, like Copilot, for ourselves so that we can learn alongside our kids.”

If only more of us thought like Leonard. The influential venture capitalist, Marc Andreessen, posted "The Techno-Optimist Manifesto” last October. To call his vision grandiose would be an understatement. “We believe that there is no material problem – whether created by nature or by technology – that cannot be solved with more technology. . . Give us a real world problem, and we can invent technology that will solve it.” If you have a concern about artificial intelligence, our ability to control it, or align it with human values, Andreessen has the answer: “We believe any deceleration of AI will cost lives. Deaths that were preventable by the AI that was prevented from existing is a form of murder.” Ok then.

For techno-utopians like Andreessen, even slowing down the acceleration of artificial intelligence to assess the risks is unacceptable. But many of the technologists building and funding AI think a pause is a good idea, especially if you see the value in exploring potential unintended consequences. As Elizabeth Spiers put it in a guest opinion piece in the New York Times last year, “the real problem with Mr. Andreessen’s manifesto may be not that it’s too outlandish, but that it’s too on-the-nose.”

On the other hand, people like Eliezer Yudkowsky, founder of the Machine Intelligence Research Institute (MIRI), and Yuval Noah Harari, historian and author of Sapiens, take an almost perfectly equal and opposite position from Andreessen. Last April Yudkowsky wrote in a post on the MIRI website (later published in Time): “Many researchers steeped in these issues, including myself, expect that the most likely result of building a superhumanly smart AI, under anything remotely like the current circumstances, is that literally everyone on Earth will die.” Breathe.

Although Harari readily admits that AI will bring large scale benefits to humanity, like Yudkowsky he warns that recent advances in AI spell danger. In the Economist last year he wrote “We have just encountered an alien intelligence, here on Earth. We don’t know much about it, except that it might destroy our civilisation. We should put a halt to the irresponsible deployment of ai tools in the public sphere, and regulate ai before it regulates us.”

When it comes to AI right now, it appears that people’s cups are either overflowing with superabundant techno-blessings or are empty, dry, and dangerously parched.

I am drawn to Leonard’s advice because it is balanced and practical. I would never argue that AI is equivalent to previous technologies created by humans. It isn’t. But from the printing press to radio and television, every generation displays both blushing optimism and existential dread when it encounters the latest and greatest. Here are four simple suggestions:

• Get to know the extremes but avoid them in practice. In other words, it’s important to take seriously both the current and future benefits and the real dangers of AI but avoid approaches that treat AI as a techno-savior (by letting your kids use it without any guidance or restrictions) or world-destroyer (by banning all uses of AI tools).

• Try AI on for size. All of us should experiment with AI tools like ChatGPT, Copilot, Midjourney, Dall-E, etc. Experimentation with current platforms is a safe way to help us understand these platforms through experience.

• Read, listen, watch. Nerd out on AI by reading books, listening to podcasts, and watching YouTube videos. You don’t need any technical expertise to do this and there are tons of sources that are reliable and accessible. See below for my recommendations.

• Talk to your kids about AI and open a dialogue about their interest in and use of AI tools. The MIT Technology Review has six recommendations. You can even get your own parents involved and share AARP’s seven suggestions!

These are simple and straightforward suggestions that we can all take seriously. The reality is that these new AI tools, despite their impressive powers, are narrow, prone to error, and safe and exciting to use. There’s a LOT more for us to discuss when it comes to AI so stay tuned for more in the months and years ahead. In the meantime, have a great weekend and remember, keep calm and AI on.

See you next Saturday,

Scott McLarty Head of School