I've been thinking a lot recently about AI and how it will affect the way our organizations and companies work. Most organizations are structured like this: At the ground level, there are the Doers, the people who actually do the work and produce stuff. These are the designers, the writers, the nurses, the programmers, the support agents, the marketers, the tellers. Although these people have technical skill, they are often the lowest-paid. At the middle level, there are the Managers, who guide the Doers and organize the way they work together. The Manager gets paid more than the Doer because it's their responsibility to make sure that groups of Doers produce the thing that the organization needs. At the top level, there are the Talkers, and they spend their entire day just communicating with people inside and outside the organization. Talkers get paid the most because although they don’t have the technical skill of the Doers or the management skills of the Managers, they can make big changes across entire organizations or industries by saying the right things and using their connections to make large deals.
Now, I think that AI is going to make a lot of the Doers obsolete in many industries. These industries include basically anything that has a digital footprint, a lot of administrative work, or both: marketing and sales, customer support, software engineering, logistics, healthcare, and any business that produces art, music, or content. Any Doer role that uses technical know-how in a specific way is going to be challenged by AI. Why? Because no matter what that technical skill might be, AI will learn to do it better. In some cases, it already has. Other examples may not be as obvious: think about the hundreds of small print shops, the thousands of recording studios, the hundreds of thousands of part-time models. All of these are being actively obsolesced by image, voice, and music generators. And we haven’t even talked about what AI is going to do to finance. (Hint: BloombergGPT is just the tip of the iceberg.)
Over the next 5 years, we’ll see a complete overhaul of the traditional organizational structure, because you simply won’t need as many Doers. Even now, we see people claiming on Twitter that they are seeing 10x increases in their productivity with AI. Think about that. If people in your company are becoming 10x more productive because of AI, then a business that’s prioritizing its finances would probably just fire 8 or 9 people and then give the remaining guy a $20/mo ChatGPT Pro subscription. And then of course, if you don't have that many Doers in your company, then you probably won't need as many Managers either.
Where will all these Doers go? I’ve had this discussion with many fellow technologists, and a few of them have reminded me that technology has always made industries obsolete in the past. And society has always improved because of it. Think about how cars replaced the kalesa, they argue, or how the Internet triggered the decline of newspapers. On balance, these technological improvements were better for humanity.
Now I disagree with that comparison because we’ve never seen anything like the speed (“5 years”) or the scale (“everything”) at which AI is going to take over. Why do I think it’ll affect “everything?” Well, because we moved “everything” that we could onto the Internet, so AI now has a very convenient channel to access all of that “everything.” Unlike the car replacing the kalesa, AI is being unleashed on … well, everything, everywhere, all at once. Why do I think it’ll take only 5 years? Because the only thing faster than the speed of the Internet is the speed of greed. With every major tech company racing to get to market, the AI explosion is already here and is limited only by how fast they can roll out the Windows updates, safety be damned.
As I continued to ponder AI’s impact on society, I realized that there’s one arena that is probably going to survive relatively untouched. That arena (pun intended) is esports. Why do I think that? Because AI defeated humans at every conceivable mental game a long time ago, beginning with chess as far back as the 1990s, then Go in 2016, then Starcraft 2 in 2019. But none of those games have gone out of fashion, and in the case of chess, you could argue that it’s never been bigger than it is today. The thing about esports (or sports in general) is that we can restrict the competition to humans only, and it doesn’t really matter if AI can beat all of the League of Legends pros without even breaking a sweat. In sports, we only really care about how humans fare against other humans. That being said, the entirety of the global sports + esports industries is barely larger than Johnson & Johnson’s marketcap, so it’s not like we have saved a major chunk of humanity’s remaining viable jobs.
Now, you might be tempted to argue that art is also a human-specific endeavor, but the difference is that sports are closely monitored and regulated against cheating. It’s difficult if not impossible to do that with art and music. As we’ve seen with that recent AI-generated image that won at the Sony World photography awards, there is currently no surefire way to differentiate a human-made piece of art from one synthesized by Midjourney. And you’ve probably already heard the songs generated with AI-simulated voices of deceased artists like Michael Jackson and the Notorious B.I.G. I don’t think we’ll find a reliable way to detect AI content any more than our cybersecurity industry has managed to eliminate hackers. None of the current tools that claim they can detect AI-generated text are reliable, with one of them even falsely claiming that the US Constitution was AI-generated. They’ll continue to improve, of course, but it’s a neverending battle, and there will always be examples that slip through the safety nets.
So where do we go from here, and how do we answer the question that I began this essay with? The first thing I’d suggest is to think long and hard about how this stuff might affect your line of work. My feed is full of writers, artists, content creators, gamers, traders, and tech founders, and every single one of those areas are going to be affected massively. This is not the time to romanticize human-made stuff, because it’s too hard to forecast whether that will continue to flourish or whether it will eventually become a niche industry like the kalesa. As the cost of digital output approaches zero, we need to keep ourselves flexible and either move up the food chain or find a sustainable exit. Interestingly, it’ll probably be the physical jobs — the chefs, the stage performers, the artisans, the caregivers — that will continue to prosper well into the next decade. I don’t know where I’ll personally end up in this massive societal transition, but last month, I started playing Starcraft 2 seriously again, and then yesterday, I bought my first two baby bonsais. Even if my esports dreams never come to pass, I might still do well as the world’s most overthinking gardener.
Keep safe out there, cryptofam!
I'm a software developer by trade and we'll either be the first on the chopping block or the last. Woodworking, baking, cooking and/or gardening were my "retirement" fallback and I've been accumulating these skills slowly. Maybe it's time to speed it up a little bit?