Futureproof by Kevin Roose came to my attention after listening to him on the Rich Roll podcast. This is a great place to start if you’re curious about AI but not sure you want to read a whole book (link).
AI is hugely complex and technical. It’s a field populated with serious mathematicians. Whilst this is good for AI, it’s not so good for those of us who are trying to learn, but cannot speak “double PhD in computer science & statistics”.
Luckily, you don’t need PhD to absorb, enjoy and learn from Roose’s book.
He offers the reader 9 (do-able) steps to make ourselves more resilient to the potential negative impacts of AI. This left me feeling that I could actually make a difference, despite my absence of technical ability.
Two discussion areas, in particular, stuck in my mind, months after reading the book.
The first is that AI is already taking our jobs, but we don’t seem to notice or care too much. The second is that, just because we can use a machine to do something, doesn’t mean we always do…
AI is Already Taking (& Creating) Jobs
There is a fear that AI “will take jobs” but people rarely recognise that AI has already “taken” many jobs. We don’t seem to care though because enough jobs have been created in other places to absorb the workers. We haven’t (yet) seen mass, socially destructive unemployment directly due to AI.
As an example, Roose discusses insurance brokers for personal products like travel insurance. Small insurance companies used to have teams of brokers making sales calls, calculating quotes and premiums and calling customers to bill them. Most of these processes now happen more efficiently with AI systems.
Think about the last time you purchased travel insurance: we go to whatever insurance website pops up first on google. We enter our needs in a few clicks. Within seconds, thanks to AI, hundreds of quotes are received, we pay online, and are immediately sent our policy documents.
This means that small insurance firms employ fewer brokers today.
However, have you heard of the plight of insurance brokers losing their jobs to AI? I have not.
Enough new jobs have been created in the industry to account for this. Perhaps those brokers are working for the insurance aggregators such as Go Compare or Compare the Market.
Or perhaps they’re not in insurance at all.
The Unimagined Jobs
Roose points out that, when he was a kid, jobs such as “App Developer, Social Media Manager, Podcast Producer or Drone Cinematographer, SEO specialist” did not exist and were unimaginable.
Similarly, it’s reasonable to predict that there will be AI-related jobs in 25 years’ time that we simply cannot imagine today.
An example of something we can imagine today is an AI Auditor. We know that AI built on biased or incorrect data will create biased or incorrect results. Therefore increasingly AI companies employ auditors to check that their AI is working correctly.
Perhaps this will create a new opportunity for today’s auditors who normally focus on business risk and finances. Ironically, an auditor at an accountancy practice who may lose their job due to automation may get a job with an AI company instead.
But, Will AI Grow Our Economies?
Historically, technological progress has been accompanied with increased productivity. In short, people have more money to buy more stuff so more people are employed to make that stuff.
Roose is concerned that the AI in our world is reducing costs without expanding the economy. He references the term “So-So technology”.
Q// So-so technologies describe the type of machine that is good enough to replace human workers but isn’t good enough to generate new jobs.//
(Roose Quoting Acemoglu & Restrepo, 2019)
Take supermarket self-checkouts. Technically they are an early form of AI. But they don’t encourage us to buy more or use a particular shop more frequently. In fact, they are often faulty and frustrating. But they are not frustrating enough to stop us from shopping or make the shop improve them.
Roose believes that these “boring bots” are a bigger risk to our societies than killer drones or rogue self-driving cars. Boring AI cuts costs, cuts jobs but doesn’t improve our world or grow businesses.
These are the silent risks that we need to watch out for.
Just Because We Can Automate, Doesn’t Mean We Do
The second idea that stayed with me, is that just because we can automate something, doesn’t mean we will.
Take coffee, for example.
We all now have the ability to make very good coffee at home.
And yet…the coffee shop market continues to grow. In the USA the Coffee Shop industry grew by 3% each year between 2017-22 (source here).
This is impressive, particularly given that many coffee shops shut down or closed during COVID.
Why Do We Leave Our Homes for Coffee?
Why are we “rational” humans choosing to leave our houses, brave the elements and spend more money to drink an identical product we could get without leaving our homes?
It’s because we are not just purchasing a delicious cup of coffee.
We are purchasing the service, the cheerful interaction with the barista, we are purchasing the feeling of treating ourselves, and perhaps we are purchasing an incentive to get us out of bed in the morning.
Humans value more aspects beyond the “efficient”. It’s fair to assume this will continue when more of our lives are run by AI.
Roose brings in a psychology theory called the “effort heuristic”. Humans generally prefer goods and experiences that have obvious human effort behind them. Importantly, we prefer them even when the final product is exactly the same.
Q// No matter how hard you work, you simply cannot outwork an algorithm. If you try, not only wll you lose, but you will sacrifice your unique human advantages in the process…Instead of trying to hustle our way we should refuse to compete on the machines’ terms, and focus instead on leaving our own, distinctively human mark on the things we’re creating.//
(Roose, 2021, Rule 4)
Copying the Barista Model
Roose believes that being a barista is one of the most AI-proof jobs. Most lists of “AI-resistant jobs” put all types of hospitality right at the top.
We can look to this example (which has technically already been automated) and think “where can I maximise my humanity in my work”.
This could be as simple as making a phone call rather than pinging off an email. Or sending a handwritten thank you letter once a piece of work has been completed.
Humans have spent the centuries since the industrial revolution trying to be more like machines. More specialised, more efficient.
However, we won’t be able to beat the AI machines and therefore it is time to reverse these trends, to be more human in every interaction, every day.
Thank you Kevin Roose for writing such a useful and absorbable book. Here‘s his website to read the book and see more from him!
People | Spaces | Places 