Today it costs a fraction of a fraction of a cent per megabyte to store data, but it is less than 30 years since the cost of data storage dropped below one US dollar per megabyte. It seems that one dollar was a psychologically important price point as, from then on, the world economy significantly shifted to capturing, valuing and transacting on information.
Despite the language we use, the production and trading of information is fundamentally different to producing and trading goods or even services. Physical goods are tangible, predictable and have clear rules of ownership. In contrast, information is intangible, almost infinitely reproducible at negligible cost and the rules of ownership are inconsistent and often ambiguous.
Also, unlike goods and services, information behaves in unpredictable ways. Just look at the unexpected influence information improperly handled through social media has had on elections, or the ability to identify personal information from apparently anonymised medical, transport and retail data.
By the time the global financial crisis arrived in 2008, the transition to the information economy, and its implicit valuing of information, was well underway. Although the financial crisis had a horrific impact on a huge number of people across many countries, central banks. regulators and governments worldwide did an amazing job of mitigating what could have been far worse.
However, the years subsequent to the global financial crisis have been far from smooth. The economy simply doesn’t behave the way economists would have expected. It has become common to say that “uncertainty is the new normal”.
I think much of this uncertainty can be attributed to the differences between the response of information-centric products and their comparable physical products to economic settings.
Consumers have gone from buying phones to buying data plans, from buying razor blades to shaving subscriptions and increasingly from buying groceries to delivery services which are incredibly information intense. Similarly, few manufacturers rise to the lofty heights of becoming major companies in developed economies, rather that badge of honour falls to businesses focused on developing and managing intellectual property which are largely information assets.
Because our knowledge of markets was evolved during the latter half of the twentieth century, it’s also useful to look at what futurists of that same time, particularly in the 1960s and 1970s thought of the twenty-first century. You can get a great summary from Alvin Toffler’s Future Shock published in 1970. The writers of the time got much of our world right, largely predicting the advances we’ve seen in computing, the internet, as well as many of the social and geopolitical moves even if not the specifics.
What futurists of the time got universally wrong was their predictions that the great challenge of our era would be what to do with all our leisure time! Although they expected mass automation to remove much of our need for work, they didn’t allow for the massive complexity that we would add to our organisations with the move to our information economy (see Why aren’t I working a four hour day?).
Rather than simple streamline a 1970s-style airline, where ticketing and check-in where simple but manual – we’ve created dozens of virtual platforms to buy our tickets and automated the check-in process through mobiles, text messages and kiosks. All of these sources of information interact in an almost infinite number of permutations.
Even more dramatically, governments and banks haven’t simply harvested the benefits of the information economy with virtual doppelgangers of their 1970s services and products, such as unemployment benefits, healthcare, mortgages and savings accounts. Rather they have leveraged the opportunity of almost infinitely flexible technology to create almost increasingly complex rules, regulations, exceptions and options.
If we find navigating today’s society hard, spare a thought for the challenge artificial intelligence faces! Today’s AI is really good at doing the things that humans do by rote. What it doesn’t do well is cope with chaos, unpredictability and unique exceptions. When today’s complex welfare services collide with our equally complex financial products the number of potential permutations make almost every circumstance unique. AI works best when each scenario is at least relatable to one it has been trained on.
Far from fearing AI will take our society away from us, perhaps our economic certainty and quality of life is aligned with the interests of making the world more predictable for our AI-assistants. The AI future might be more human friendly than we ever thought!