Don’t Worry About a Thing, AI's Gonna Be Alright

I hope you are having a lovely Friday morning as you read this, because the first few paragraphs are a bit dark. Grab yourself a cup of tea and a chocolate Hobnob before diving in.

Depending on who you listen to, artificial intelligence will either save humanity by curing cancer, solving climate change and improving crop yields, or destroy humanity by acting as a force multiplier on other technologies, such as robotics and nuclear weapons. Quite the difference of opinion.

The truth is, no one really knows. The Economist recently asked a group of domain experts for their probability estimate that, in a five year period, by 2100, 10% or more of humans will die due to AI. Their median estimate was 12%. The same experts put the risk of human extinction due to AI at 3%! By comparison, superforecasters put their median probabilities at 2.1% and 0.38%.

The inevitable foggy view forward makes forecasting hard. Though it’s a shame David Bowie isn’t with us any more, as his internet predictions were scarily accurate.

But as far as I’m concerned, these probabilities are uninteresting. We are notoriously awful at understanding probabilities. If you look at historical patterns in the data, you’ll notice that the chance of human extinction is precisely zero (I am aware that the same could be said of the dinosaurs until they bit the dust). In 2100, when most of us will be dead, who will look back at a July 2023 edition of The Economist and evaluate the predictions for accuracy? Think about it logically, either we are extinct, and those experts who predicted extinction won’t be around to gloat, or we are not extinct, but those experts who didn’t predict extinction won’t be around to gloat. It’s all pointless.

But it does point to the human psyche, from the days of the Luddites smashing mechanised looms, of a fear of the uncertainty of new technology. I’m sure when the trusty barcode came along, someone was against it.

I’m continuously baffled at how unreceptive many are to new tech and how we underestimate how everything will be wonderful. At the risk of sounding like a Billy Joel song, over the last century, we’ve got through World War One, hyperinflation in the Weimar Republic, the Great Depression, World War Two, the Cold War, mutually assured destruction, the Sick man of Europe, the fall of the Berlin Wall, Black Wednesday, the Asian Finance Crisis, Y2K, the dot-com crash, 9/11, the Global Financial Crisis, the Eyjafjallajökull volcano eruptions, the coronavirus pandemic and eleven Fast & Furious movies.

Sure, we’ve had some near misses along the way. In his must-read book Command and Control, Eric Schlosser details how close we came to nuclear catastrophes, such as the 1966 Palomares B-52 crash, where a B-52G bomber crashed into the Mediterranean Sea whilst carrying four thermonuclear bombs. Or the 1980 Titan II ICBM explosion in Arkansas.

But we didn’t incur a nuclear catastrophe. We’re still here. It gets lost in the daily economic news and stock market fluctuations, but the world has done fine.

Let’s have a look.

From March 2005 to December 2022, the FTSE All-World index averaged annual returns of 9.95%. £10,000 put into the index in March 2005 would be worth just under £54,000 at the end of 2022. Think about everything that has happened since March 2005: 7/7, Lehman Brothers collapsing, the Arab Spring and the Fukushima nuclear disaster, to name a few. But if you’d sat and done absolutely nothing for 17 years (tempting), in absolute terms, you’d be five times richer.

Go back a little longer, to the start of the 90s, and you can see that the world defining, catastrophic events of our time, the GFC and the coronavirus pandemic, are merely blips on the otherwise glorious upwards trend of global GDP. The last few decades have been immensely prosperous.

If you worry that I’m too focused on economic indicators and stock market returns, well, I’m not. The relationship between GDP per capita, human development, and life expectancy is strongly positive. It’s why you hear smile-raising stories like Belize becoming the 42nd country free of malaria.

The benefits seen on the stock market tickers of Wall Street and Canary Wharf are the same benefits seen everywhere. It’s a measure of how economic and technological progress works that yesterday’s new inventions become mundane: at one point, the barcode, magnetic stripe, and microchip were all revolutionary. Now, they are ordinary. The same will happen with artificial intelligence as it permeates every corner of business and the economy, like SQL and HTML.

Somehow, the mundane is exciting, and the exciting is mundane. But it means that despite the wars, the natural disasters, the failures and everything else, we’ve done just fine. Scrap that, we’ve done more than fine, we’ve done incredibly well. We did during the birth of the nuclear era, and we will do just fine in the artificial intelligence era. Don’t worry about a thing…

Thanks for reading, I’ll see you next time!