Every single essay I’ve read about the impact of AI on the future of work since January 2023 has either been biased or reductive. I’ve thought about how to pen my thoughts and ideas about this topic for the past 6 months. It took me a few weeks to research and draft. It was far from perfect. Naturally, I completely forgot about it. Yesterday, I rediscovered it in my drafts and thought now is as good a time as any to press “Publish”. Enjoy!
In my opinion, we’re having the wrong conversation. So far, the debates I’ve heard or participated in on the topic of AI and the future of work have centred on whether people will have jobs in the future. Needless to say, this is an incredibly important question. However, it shows that the way our collective imagination conceives of the nature of work is incredibly narrow. As Daniel Susskind–one of the foremost thinkers and economists in this space–says in his book A World Without Work, “The risk, in focusing on jobs alone, is not so much failing to see the proverbial wood for the trees, but failing to see all the different trees in the wood.”
Not only are most essays having the wrong conversation, but when they’re closer to having the right conversation, they’re biased by some ulterior motive (*cough cough* venture capitalists waxing lyrical about how AI will save the world).
My aim with this series of essays is to broaden our minds to acknowledge what a “protopian” future of work could really look like. For those unfamiliar with the concept of a protopia, it’s a concept that Kevin Kelly, the founder of Wired, debuted in his 2010 book, What Technology Wants. He used it to refer to a society that, rather than presenting a highly optimistic, perfect vision of the future (as in a utopia) or falling into dire dysfunction (as in a dystopia), makes incremental progress over a long period of time — thanks to the ways in which technological advancement is enhancing the natural evolutionary process. In exploring this potential protopia, I will present a vision for how we can reorient human potential in a way that will enable us to craft a future fit for human snd ecological flourishing.
But let’s not get ahead of ourselves. In Part I of this essay, I want to discuss how we’ve got to where we are and where the current conversations are falling short.
“Automation anxiety” is not a modern phenomenon. Throughout history, technological progress has led to fear-mongering about the potential displacement of jobs. AI is often compared to Prometheus, the Greek titan who, in bringing fire to humanity as an essential tool that warms and protects us, also dangled the keys to unlock utter destruction. In other words, technology's innate promise and threat have always been two sides of the same coin.
It's crucial to recognise that while there are genuine concerns about job displacement due to AI, history has shown that those who are fearful of huge job displacement are largely proven wrong. On the contrary, it often leads to the creation of new professions and spurs demand in unexpected areas. For example, the Industrial Revolution, while initially disruptive, eventually paved the way for myriad professions that previously didn't exist.
My take is that some job displacement in the short to mid term is inevitable as those at the bleeding edge of the AI adoption curve make decisions to automate away or outsource large elements of their workflows to AI. This is known as the Substitution Effect, when new technologies replace human workers entirely. However, what’s often missed is that there is simultaneously a complementary force at play when new technologies like AI emerge into the mainstream. There are three complementary forces which have the biggest effect on job increase.
Firstly, we have the Productivity Effect, which refers to how technology makes humans more productive and effective at the tasks they already do. One example is when AI replaces specific, contained tasks completely e.g., administering paperwork, allowing humans to focus on other, more high-leverage tasks that require uniquely human skills such as creativity, context-dependent problem solving or nuanced relationship management. While this is commonly referred to as the “productivity effect”, I’m increasingly terming it the “centaur effect”, in recognition of the augmentative effect AI can have on humans’ innate capabilities.
The most famous example of the centaur effect in action is in chess, where the best performing teams are not exclusively human or exclusively AI. The best teams are hybrid; they combine the computational speed and power of a machine with the natural intuition of a human to generate a better overall outcome. Chess aside, if the centaur effect plays out in the mid to long term (which I’m quite certain it will), productivity increases from AI/human hybrid teams will be passed on to customers via lower prices or better-quality services, leading to a rise in demand.
Secondly, we have the Bigger-Pie Effect. The pie, in this case, refers to the economy. The more new technologies are integrated into working practices, the bigger the economy grows. The Bigger-Pie Effect works in direct contrast to the Lump of Labor Fallacy, which states that there is a fixed amount of work to be done in the economy. The logic goes that if machines do the work for us, there won’t be anything left for us to do. Questionable logic, to say the least. 60% of jobs that exist today did not exist in 1940 (Autor, 2022). Technology-driven metamorphosis and growth in jobs has been happening since the dawn of time, leading to a larger overall pie.
Lower prices = greater spending power = higher demand = more production = the creation of new jobs in new industries we haven’t created or envisioned yet
But the pie doesn’t stop rising there. When you combine the Bigger-Pie Effect with the Changing-Pie Effect–which describes how the economy is constantly in flux–you really start to see how reductive some of our conversations have been. 500 years ago, Britain’s economy consisted primarily of farms. 300 years ago, it was factories. In 2019, it was offices. Today, it’s changing literally as you read this. The rise of the gig economy, flexible work, portfolio careers and fluid roles underscore just how difficult it is for us to anticipate how fast and dramatically the pie can change.
To boil this all down: on centre stage, we have AI, that will replace some jobs, while simultaneously making us more productive. Backstage, the AI-fuelled economy will continue shape-shifting and growing in ways that are hard to predict.
So why might all this not happen? Oh, it’s simple:
Pretty easy if you ask me.
But rather than furthering the false dichotomy–between the idea that automation will spell the end of work for humans and the argument that technologies will always increase the demand for labour as they have done in the past–I believe that the Substitution Effect and the complementary forces that accompany it will interact in ways that are hard to predict. Historically, the complementary effect has won out over the Substitution Effect, ultimately resulting in a net positive increase in jobs. The reason why human labour has historically prevailed is connected to our ability to adopt and acquire new skills by means of education (Goldin and Katz, 2009).
But what makes AI different from previous technological breakthroughs is its exponential nature. This, then, is the ultimate question: Once AI hits the point of exponential intelligence, will our ability to adopt and acquire new skills still allow us to prevail?
To be continued…
