David Egan Philosophy

Artificial Intelligence (AI) chat models have rapidly become ubiquitous in today’s digital landscape. With the advancement of natural language processing technology, AI chat models are now able to carry out conversations that feel surprisingly human-like. This technology has been adopted across a wide range of industries, from customer service to healthcare to finance, offering companies the ability to automate conversations at scale, while also providing a more seamless user experience for their customers. The rise of AI chat models has transformed the way we interact with technology, and the impact of this technology is only continuing to grow.

I didn’t write that last paragraph. The text was generated by ChatGPT, the chatbot developed by Open AI, which has become the fastest-growing consumer application in history. I gave it the prompt: “Write the opening paragraph to an essay about the rapid rise of AI chat models.” The text in the paragraph above was the result.

ChatGPT is only the best known of an array of programs and services built around “large language models,” or LLMs. These models train AI systems on huge quantities of online text so that they learn how to generate plausible-sounding responses to queries or conversational prompts. This technology has advanced at a speed that has surprised even insiders in recent years. Responses have ranged from enthusiasm to alarm and from skepticism to (in the words of Sundar Pichai, CEO of Google) “the most important thing humanity has ever worked on.”

The ”this changes everything” mentality sees in artificial intelligence not just a possible major advance in technology—something with the transformative effect of the Internet, electricity, or the wheel—but something that might radically reshape what a human life is. We seem to be on the cusp of an encounter with a hyper-intelligent alien civilization—except that that civilization was made on Earth by us.

I’m mildly skeptical that we’re mere steps away from creating sentient artificial beings. Learning to generate an exceptionally plausible pastiche of human thought isn’t the same thing as learning to think. I could well be wrong about this. But the point I want to explore in this essay is a different one. Even if we’re a long way off living alongside silicon souls, recent and near-future advances in artificial intelligence will challenge our ideas about what makes us human.

Let me start with a pair of recent achievements in artificial intelligence and robotics. In 1997, the IBM chess-playing system Deep Blue defeated Garry Kasparov in a six-game match, making it the first time that a machine had defeated a reigning world chess champion. Last autumn, a two-legged robot set a new record by running 100m in 24.73 seconds. As a point of reference, that’s about the speed of a relatively athletic five-year-old.

Twenty-five years ago, we had a chess machine more powerful than any human being. We’re still working on developing a robot that can beat a five-year-old in a footrace.

We don’t normally think of our bipedalism as involving a great cognitive achievement but it’s actually remarkably tricky. With every step you throw yourself off balance. Keeping going in a straight line without falling over requires sophisticated kinaesthetic awareness and the coordination of a large number of muscle movements. But because most of us get the hang of it in the first couple years of life, we tend not to reflect on what a marvel walking is.

What makes humans special? The people who ask that question have tended historically to draw a contrast with animals. The thing about us that stands out most immediately in comparison with other animals is our minds. Other animals can run faster than us, see, smell, or hear better than us, and outmatch us in strength. But none of these animals can outsmart us. It’s not just that Garry Kasparov could mop the floor with the smartest chimpanzees in a chess match. Even I could.

We’ve maybe given outsize importance to our smarts because they’re what most distinguish us from our closest non-human cousins. But what happens when we start to live alongside some inorganic cousins who make our chess-playing or mathematical skills look infantile? Suddenly being able to run without falling over seems a lot more impressive.

We’re used to being the smartest things on the planet. What happens when that’s no longer the case?

There are a number of different paths we might take as we adapt to this change. One path that some techno-optimists already eagerly anticipate is to transcend our fleshy human limits altogether. They aspire to a cybernetic future where we cease to be organic beings and use technology to transform our bodies and minds. This approach leans ever more heavily into the idea that found such clear expression in the Second Meditation of René Descartes: that we are essentially thinking things and not bodily things. One possible future for humanity is a techno-Cartesian dream of sundering our dependence on the body altogether.

Another possibility is that we might come to appreciate our bodily existence more fully. I fear we’ve done tremendous harm to ourselves, and to the world we inhabit, because we’ve tried so hard to deny our animal nature. Perhaps the crowning achievement of our distinctive human intelligence will be to create machines whose own intelligence so humbles our own that we can once again embrace the simple but perplexing predicament of having a body, of being a body, and of sharing that bodily life with animals that are more like us than we once supposed. 

1. With over 120,000 weddings per year, Las Vegas is the second most popular city in the world for getting married. Number one? Istanbul. (source)
2. Canada is the second-most tornado-prone country in the world, after the United States. But this fact is rarely remarked upon, even by Canadians, because most Canadian tornadoes touch down in sparsely populated areas. (source)
3. When the British TV channel Channel 5 launched in 1997, the first ad it featured was for the perfume Chanel No. 5. (source)
4. Rifles get their name from the “rifling” of the inside of the barrel: a series of spiral grooves that cause the bullet to spin as it leaves the barrel, giving it greater accuracy over a longer range. You see this rifling design in the classic opening to James Bond films where you look down a rifled handgun barrel at Bond but he turns and shoots first. (source)
5. If animal agriculture were a country, its emissions would outrank every other country besides China. (source)
6. The US federal government imposes a 10–11% excise tax on the purchase of firearms. That money provides the bulk of wildlife restoration funding in the US. (Back in 1937, when the law was passed, most gun owners used them for hunting.) (source)
7. Contrary to contemporary English usage, the ancient Greeks spoke about the past as something that lies in front of us (because we can see it) and the future as something that’s behind us (because we can’t see it). (source)
8. Canada is such an attractive place for money laundering—notably through Canada’s hot real estate market—that it even has its own name: “snow washing.” (source)
9. The artificial flies used in fly fishing are too finicky to be made by machine. About 60% of the world’s supply is hand-made in Kenya. (source)
10. Because of volcanic activity earlier in its history, the moon has a network of gigantic tunnels just beneath its surface, with tunnels as much as a kilometre wide and several hundred kilometres long. (source)
11. It seems like dinosaurs may have had lips. (source)