There's a lot to be worried about. And EAs have a seemingly boundless capacity to worry about the world's most painful challenges. AI risk, the looming specter of climate change, the sharp pangs of animal suffering. Excruciating and pointless diseases. Institutions lacking the capacity to execute on their mandates, or the incentives to enact change quickly enough. Unbridled ego and greed. It's enough to make the most balanced person despair! But here, amidst it all, I want to offer some modest reflections on hope.
I've been facilitating an AGI Safety fellowship for the past few weeks, and I've rarely been so inspired as I am every time I walk into the meeting room. Our cohort is full of deeply technical engineers and thoughtful scientists; people who have known for a while their commitment to making AI go well and vital newcomers; people who reflect in all aspects the breadth and diversity of talent, background, and thought that in my view grounds the best of human ingenuity.
And it's this cohort, I think, that gives a convincing case for hope. In small ways: when they anticipate a reading's argument, or pose a clever question. And in large ways, too: when they dive into a research topic with ardor, or refocus their career on steering lifeboat Earth away from gloomy rocks. To be sure, there's the hard challenges: timelines, core unresolved conceptual gaps. What if it's all for naught? And what right do we have to think we'll meaningfully contribute?
Yet I keep coming back to the feeling I get when I walk into the conference room each week. "We can solve this, somehow!" "It's possible. And if it's possible, we'll find a way." Sometimes in despairing moments, this is a good anchor for me. It's a reminder that the world will look brighter someday, and that the best of us are arrayed to face the challenge.