100 years progress in 5-10 years
My take on Dario Amodei's essay 'Machines of Loving Grace'
As the rain pours down and the wind whips through my garden, I find myself staring out, feeling powerless against an unstoppable force. For many, this is how watching the rapid progress of AI feels—distant, yet overwhelming.
For those deeply involved in AI development—building models and training data—there’s a sense that a storm is brewing in Silicon Valley. That’s why Dario Amodei’s recent essay, Machines of Loving Grace, arrives at such a critical moment. Poised at the intersection between sense and madness - this essay walks the immersive line showing us what is possible through the lens of ‘what could go well, or what could go badly’.
Reacting to a perception
For those new to AI, Dario is the CEO of Anthropic, founded in 2021 by former OpenAI employees Dario and Daniela Amodei. His company focuses on AI safety and alignment, ensuring AI systems behave ethically and are aligned with human values and goals. Anthropic is the flip side of the AI-progress story, working in opposition to rapid expansion with a more cautious stance. For context, Anthropic was founded by a team who left OpenAI, the creators of ChatGPT, over safety concerns. This is significant, as it occurred in March 2021, around the same time a letter, signed by 18,000 people, called for a six-month pause on AI development due to fears of existential risk.
That ‘pause’ never happened and today we are facing rapid AI development on a scale never seen before. This essay then has come at a pivotal time. The discourse that surrounds AI has been poisoned by hype agents - people selling a vision of the future, one that in truth, is yet to arrive. Popular media sells us an image of two extremes - the AI utopia, and the AI dystopia. Perhaps the truth lies somewhere in the middle. I feel Machines of loving grace captures this reality perfectly.
As far as Silicon Valley CEO’s go - Dario is not one to make wild speculations about uncertain futures. I would say he is a cautious optimist - a value his company embodies at its core. There is a perception that Anthropic is a company of AI-Doomers always warning about the dangers whilst ignoring the potential. This is not the case. Dario’s essay does speak about risk, but reminds us that risks are important but manageable, and should not overshadow potential.
Let’s breakdown the main points.
What if all AI does is allow us to make 100 years worth of progress in the next 10 years? What would that change about the world?
In a part called ‘the compressed 21st century’ - Dario hints we might get powerful AI (AGI) by 2026. Sketching out a vision for this future, he focuses on what a world with powerful AI might look like if everything goes right. His thinking hinges on the arrival of powerful AI, his alternative to AGI, or Artificial General Intelligence.
*AGI refers to a type of artificial intelligence that possesses the ability to understand, learn, and apply knowledge across a wide range of tasks at a human level or beyond.
Here is a summary of the five key points, referred to as Radical Upsides in the essay:
Biology and Health: Powerful AI could eradicate most diseases, extend human lifespans and create enhanced biological freedom.
Neuroscience and the mind: Powerful AI could cure mental illness, improve cognitive function and enhance human emotional capabilities.
Economic Development and Poverty: Powerful AI has the potential to alleviate poverty through accelerated economic growth, particularly in the developing world.
Peace and Governance: Powerful AI can strengthen democracy, promote individual rights and work to restore a more peaceful global order.
Work and Meaning: Powerful AI can adapt to an AI-driven economy, finding meaning in a world where AI surpasses human capabilities.
This vision sounds almost utopian, and it's clear how these radical benefits could enhance life for all of humanity.
When discussing powerful AI, he compares it to "an AI smarter than a Nobel Prize winner in any field"—an astonishing thought. He also challenges the notion that AI is a disembodied force causing chaos, pointing out the irony that it is, in fact, a creation of our own relentless pursuit of progress. In short, it’s a bold and optimist view of AI’s potential.
Critics might point to the timing of this essay to co-inside with a new round of funding for Anthropic. Or, the attention grabbing title of the piece. Both are fair judgements but distract from the importance of the ideas contained within. Dario is attempting to show us both sides of a story here. On one hand he is saying its’ ok to be a ‘Doomer’ when it comes to AI, and this can help us evolve with this technology. By contrast, he paints a very different future for us. In this near future (2026) powerful AI will give us 100 years of biological progress in 5-10 years.
Whatever the timeline AI takes, it is important to envisage what a good world with powerful AI would look like. Dario’s essay does this. Check it out here and let me know what you think.