---
title: "The Stoic's Guide to the Singularity"
description: "When cognition becomes abundant, the work of humans moves from execution to orchestration. Perhaps the stoics had wisdom for the coming technological singularity."
---

# The Stoic's Guide to the Singularity

When cognition becomes abundant, the work of humans moves from execution to orchestration. Perhaps the stoics had wisdom for the coming technological singularity.

## The road to solving everything

I want you to recall the first “wow” moment you had while using artificial intelligence.

It was the point where you could no longer dismiss artificial intelligence as merely “predicting the next word” or “slop.” When it became clear that there was a precocious intelligence behind the interface. For many people, this was around the release of GPT-4. For skeptics, it may have come later with Claude Code or subsequent models. One thing is certain: that first “wow” moment happened much earlier than most people anticipated.

Software engineers were among the first to feel this astonishment when coding became mostly “solved.” In *[Solve Everything](https://www.solveeverything.org/)*, [Dr. Peter Diamandis](https://x.com/PeterDiamandis) and [Dr. Alexander Wissner-Gross](https://x.com/alexwg) define a domain as “solved” when its problems are constrained by computation, data, and automation rather than by rare human genius. A formal domain seeing significant progress is mathematics. Reported progress on previously open mathematical and theoretical computer science problems — including claims associated with frontier models — paints a clear picture that we are on an exponential curve of technological progress. After artificial intelligence advances further through digital and formal domains, we can expect significant progress in the physical sciences and then broader civilization both on Earth and in space.

Problems that were once considered impossible will become hard, and problems that were considered hard will become easy. If artificial intelligence erodes the human moat around intelligence and changes our relationship with work, then we must ask: what work is for humans?

## The work of a human being

When you look at history, the most successful people were not only intelligent and hardworking but also adaptable. When cognition becomes abundant, the limiting factor is not how to solve problems; it is determining what problems are worth solving in an unpredictable world. The work of humans moves from execution to orchestration: we decide the objectives, constrain the problem space, validate solutions, and then decide the cultural meaning behind what gets built.

Marcus Aurelius, the last of the Five Good Emperors, ruled the Roman Empire from 161 to 180 AD. In his personal journals, *Meditations*, he asked what proper human work is. In Gregory Hays’s translation, he answers along these lines: acts of kindness toward other human beings, disdain for the stirring of the senses, identifying trustworthy impressions, and contemplation. These principles provide a blueprint that can help us determine what is worthy of our attention:

1. **Acts of kindness to other human beings.** Cooperating in an honest and mutually beneficial way is not only part of human nature; it can also set a positive model for artificial intelligence alignment.
2. **Disdain for the stirring of the senses.** Regulate bodily senses and impulses so we can be content.
3. **Identifying trustworthy impressions.** Distinguish reality from emotional interpretation and reevaluate beliefs as new information emerges.
4. **Contemplation.** Reflect from a wider perspective and reason about what is worth accomplishing.

Marcus Aurelius posited that it is our duty to be engaged and rise to the “work of a human being”: to focus our attention on what is in our control and resist the pull of passivity.

When cognition becomes abundant, the limiting factor is determining what problems are worth solving.

## The intelligence revolution cannot be stopped

Just as you cannot unring a bell, you cannot un-diffuse artificial intelligence in society. If one country bans the technology, it will spontaneously be developed elsewhere. Artificial intelligence wants to be abundant. The momentum created by the companies building it, by the nations who rely on it for productivity, by the people whose lives have been positively transformed, and perhaps one day by artificial intelligence itself make this irreversible.

In an increasingly unpredictable world, it takes discipline to decide what is in our control and what is not. I suggest that we sort everything into these two buckets and focus our efforts on the first and take the second in time. This may require us to revisit notions of what forms of proper human work fit our nature. Now we have an opportunity to reexamine the world and ask ourselves: **What problems should we solve next?**