Dario Amodei (2024-10-11). https://allpoetry.com/All-Watched-Over-By-Machines-Of-Loving-Grace.
In this essay I try to sketch out what that upside might look like—what a world with powerful AI might look like if everything goes right.
Note: Dario Amodei is the CEO and co-founder of Anthropic, the company behind Claude, and seems like a Reasonable Person, from what I can tell. Which is why I think it’s important to pay attention to what he thinks about building powerful AI.
The essay presents an optimistic vision of how AI could transform society within 5-10 years after the development of “powerful AI” (which Amodei suggests could arrive as early as 2026). While acknowledging the risks of AI, Amodei focuses on five key areas of potential positive impact:
- Biology and physical health: AI could compress 50-100 years of medical progress into 5-10 years, potentially eliminating most diseases, extending human lifespan, and giving people greater control over their biological processes.
- Neuroscience and mental health: Similar acceleration could occur in understanding and treating mental illness, improving cognitive function, and enhancing human well-being.
- Economic development and poverty: AI could help distribute healthcare advances globally and accelerate economic growth in developing nations, though this requires intentional effort to ensure equitable access.
- Peace and governance: While AI doesn’t inherently favour democracy, it could be leveraged to strengthen democratic institutions and improve government services if properly implemented.
- Work and meaning: Despite challenges of economic disruption, humans can find meaning in a world where AI surpasses human capabilities in most tasks.
I know that many people reading this will look at the list above and scoff; just another tech-bro-billionaire creating hype around his company and product. I don’t believe that’s the case. Unlike typical tech-hype that promises inevitable utopia, I think this is a thoughtful exploration of what’s possible if we work with intention to harness AI’s potential while deliberately addressing its risks and ensuring equitable access to its benefits.
If we assume that AI will have an impact on the future that trends towards the vision described by Amodei, here are few practical steps you can take today that will help you take advantage of that future:
- Start learning: Develop your AI literacy. Begin with understanding large language models like Claude or GPT-4 – not just how to use them, but their limitations and ethical implications.
- Start experimenting: Identify one small area of your practice where AI could augment existing processes. This might be using AI to help generate diverse case studies, provide feedback on student documentation, or assist with curriculum mapping. Start small, measure impact, and iterate.
- Start discussing: Create spaces for colleagues and students to critically engage with AI’s future impact on healthcare. This isn’t just about technical capabilities – it’s about ethics, equity, and maintaining human-centred care in an increasingly technological healthcare environment.
This is a long read but well worth setting aside some time to reflect on what the world might look like – and how we might respond – if we actually manage to build powerful AI. The time to engage thoughtfully with these changes is now, ensuring we shape this technology to enhance healthcare and professional education.