We'll Probably See The End of Work
What then?
I’ve spoken before about how lazy I am. Work is just a thing I have to get through before getting back to the things I really want in life.
It’s a gauntlet that must be run everyday.
So, you can understand why I’m a bit of an AI stan. If we can make an unfeeling, dead-on-the-inside husk do my emails, I’m a happy guy. If we can get it to do all our boring work, all the better. Maybe this makes me a bit biased, but I think we’re not actually that far off this reality.
Sometimes when I talk about this, people are skeptical. It’s even become a running joke between my brother and me that whenever we talk about some concern of the future, I’ll just say “Eh, the nanobots will solve it”. It’s an understandable skepticism. People are wrong about their predictions of what the “next big thing” is all the time. We were supposed to have hoverboards by 2015, and the closest we got was this.
“And here I am, doing eeeeverything I can,
Holding oooon to what I am, pretending I'm a superman!”
We were also promised that crypto would change the financial industry, that we’d all be riding around on Segways, and Google Glasses would be sitting on my face right now. However, I still pay with pounds, I walk everywhere, and the only thing sitting on my face is your sister. It’s the same as it’s always been!
Still, as numerous the stories of failed tech are, there are also stories of us underestimating it. In 1903, The New York Times featured an article that said “Man won't fly for a million years—to build a flying machine would require the [efforts] of mathematicians and mechanics for 1-10 million years”. The Wright Brothers flew for the first time 69 (nice) days later. Presumably right by the NYT offices to flip them the bird.
So, the question is, where does AI fall? Is it the crypto of the 2020s, or is it the flight of the 2020s? I think it’s becoming clear that it’s the latter. Last week, Open AI released a new set of models, O1. I’ve used them, and in under an hour I was able to (with no knowledge of web development) create a site that looks like this.
For some reason, this gif doesn’t work in my mobile app, but if you’re having trouble, you can probably see it in the browser here. At least, that worked for me. If you can’t be bothered, just imagine a really cool website in your head.
Just 2 years ago, the most state of the art model the public had access to (GPT 3) would output something of this kind of quality.
If you’re paying attention, it’s obvious that this tech is catching up to our skillsets quickly. The reason that tech changes this way, is because improvements in tech lead to further improvements in tech, which makes the curve we’re on exponential. To us bystanders, this is why it feels a bit like nothing happened for decades, and now it’s happening all at once - it’s because it is!
AI is routinely underestimated in much the same way aeroplanes were. In 1989, Kasparov had this to say about chess machines.
How long do you reckon Garry? 1 - 10 million years, maybe?
He was beaten by a machine just 7 years later. In 1997, the New York Times (Again?!) said that it may be 100 years before a machine beats an expert human at Go. It only took 20 years.
During Covid, there was a lot of talk of exponential growth and our misunderstanding of it. People seemed unable to grasp that the nature of disease spread is that it goes from 0 to 100 very quickly. That’s why we saw very flat lines with massive spikes, instead of gradual rises. Each new person sick increases the number of people spreading the illness, which makes more people sick, and so on.
Computers do the same. Each generation helps to improve the next generation, and so on, until we see graphs like this.
Notice the y axis is logarithmic. Nuts!
So, we’ll probably find that AI surpasses us in most domains in the not too distant future. At the rate we’re at, I would be surprised if there are any jobs left that a human can do better in 20 years. So, where does that leave you and me?
I’m not really sure. Of course, it relies on how we transition into this next phase. The whole reason we’re able to build decent lives for ourselves is because we’re able to sell our labour in exchange for resources - but what happens when those with the resources no longer need our labour? I’m an optimist, so I hope that we’re all given more free time to pursue what we really want, instead of being murdered by robots, or left to starve by omnipotent dictators. I will admit, though, that the chance of disaster is not insignificant.
Still, even if we get the perfect outcome, we’ll need to prepare. I fear that many of us will just succumb to whatever insane hedonic contraptions await us in the 2040s - and will likely be unfulfilled as a result. Whenever I take time off work, my first day is always great, but the 5th one feels a bit empty. There’s only so much Netflix, PS5, and skittles you can consume before the hole inside you gapes ever wider. Wellbeing is not found in lighting up our pleasure centres day in day out, it’s found in a mix of pleasure and purpose.
This, I think, will be the defining difference between those that sink and those that swim in the post-work era. The pleasure pit can always take you deeper, and there’s no light on the other side. The people that will succeed, then, will be those that replace their work with more work. It’ll just be the work they’ve always wanted to do, but were unable to because they work at Amazon and get their toilet breaks timed. The people that flourish will be the ones that write, paint, lift, and play. Even if a machine can do them better, we’ll still find these things meaningful because they push our boundaries, and express our feelings. That’s the reason we still love watching Usain Bolt, even though I could beat him in my Ford Fiesta.
I hate to be the one to break it to you, but unless you’re a high ranking government official, or in the industry, there’s probably not a lot you can do to influence the coming changes that AI will bring to society. So, we better start thinking about how we’re going to live our lives once it becomes really advanced - and part of that is finding something that’ll get us out of bed in the morning when we don’t have to anymore. As much as it pains me to say, I don’t think Space Marine 2 is going to cut it.
What then? We’ll work it out 😉
I think how we live our lives will depend a lot on whether the benefits of AI accrue to workers or owners. If workers still get paid decent salaries and they can do their jobs more efficiently with their AI assistants, that will be awesome.
But what if the owners decide that they don't need workers any more with all this awesome AI? Will they share the gains equitably or will they buy expensive houses in the tropics and leave everyone else to fend for themselves?