Stay up thus far with free updates
Simply log in Artificial intelligence myFT Digest – delivered straight to your inbox.
Even the neatest experts find it difficult to predict the long run of technology. Take the instance of Bob Metcalfe, the inventor of Ethernet, who in 1995 boldly predicted that the Internet would experience a catastrophic collapse – or “gigalapse” – the next 12 months.
But when he got it fallacious, Metcalfe literally chewed up his own words. To chants of “Eat, baby, eat!” At a tech industry event, Metcalfe tore open a replica of his forward-thinking InfoWorld column, put it in a blender and ate the resulting pulp.
Metcalfe's unlucky experience – which he accepted with grace and humility – is considered one of dozens of examples of false predictions included within the insightful online resource Pessimist Archive. The archive covers the invention of the camera, electricity, airplanes, television and computers, and chronicles the various imaginative ways by which subsequent generations of technology experts got it completely fallacious.
It's price browsing the archives considering the flood of predictions in regards to the miracle technology of our time: artificial intelligence.
The only certain prediction is that the overwhelming majority of those predictions might be exaggerated. Those optimists who predict that AI will soon usher in a wonderful recent era of radical abundance are more likely to be upset. But the pessimists who predict that AI will soon result in human extinction aren’t any less more likely to be fallacious. On the opposite hand, nobody might be there to congratulate them in the event that they are right.
With AI, it could be easier to find out the direction of travel than the speed of travel. Just because the Industrial Revolution enlarged muscles, the Cognitive Revolution enlarged the brain. “AI is best viewed as the newest general-purpose technology that might be used for an infinite variety of purposes,” says Arkady Volozh, founding father of Amsterdam-based startup Nebius, which builds AI infrastructure for model makers in various industries.
“AI is like electricity or computers or the Internet,” he says. “It's like a magic powder that may improve anything. More and more functions are being automated more efficiently. Just as an excavator is more powerful than a human with a shovel, AI might be used to automate routine processes.”
However, earlier general-purpose technologies corresponding to railroads and electricity often took a long time to extend productivity. A brand new infrastructure have to be built. New ways of working should be introduced. New services have to be dropped at market.
Meanwhile, the introduction of recent technologies may very well impact productivity for some time as firms and their employees adapt to recent ways of working. In fact, recent technologies may even result in a rise in unproductive work: How many pointless emails have you ever read today?
Some economists have described this phenomenon as a J-curve – productivity initially falls before later rising again.
“General-purpose technologies like AI enable and require significant complementary investments, including the co-invention of recent processes, products, business models and human capital,” said economists Erik Brynjolfsson, Daniel Rock and Chad Syverson argue in an article from the National Bureau of Economic Research. These complementary investments are sometimes poorly recorded in official economic statistics and might take a protracted time to translate into higher productivity growth.
Zooming out even further, it may be fallacious to speak about AI as a revolution in its own right, somewhat than a continuation of the knowledge technology revolution that began within the Seventies. According to at least one Essay this 12 months by economic historian Carlota Perez: “A revolutionary technology is just not the identical as a technological revolution.”
In her 2002 book, Perez identified five major technological transformations, starting with a wave of creative destruction, followed by mass diffusion of innovation and a golden age of economic growth. This pattern has repeated itself often: starting with the Industrial Revolution within the 1770s; followed by the steam and railway ages of the 1830s; the electricity and engineering era of the 1870s; the era of mass production within the 1910s; and our own current IT revolution.
All of those technological revolutions were accompanied by changes in government and society, resulting in the creation of recent institutions corresponding to unions, regulatory agencies, and welfare states to assist manage turbulent changes.
According to Perez, we are only starting to assume the institutions we want to administer our current IT revolution and counter economic inequality, autocratic populism and climate-related disasters. “Changing this broader political-economic context has grow to be probably the most urgent task of our time,” she argued earlier this 12 months.
Designing suitable recent institutions might be a serious challenge – even with the assistance of AI.
