“Let’s go and invent tomorrow rather than worry about yesterday,” Steve Jobs once boldly declared. “Show me a problem, and I’ll look for a technology to fix it,” said Bill Gates. These words, and many like them from other tech luminaries, perfectly capture the techno-optimism that has become so ascendant today – a shared vision where the power of technology can overcome all sorts of limitations and advance the greater good.
But is technological advancement always for the greater good? And does it always bring progress?
Not necessarily, say MIT economists Daron Acemoglu and Simon Johnson in their big new book, Power and Progress: Our 1000-Year Struggle Over Technology & Prosperity. For Acemoglu and Johnson, there is nothing automatic or inevitable about technological innovation leading to better lives and shared prosperity.
That’s a sharp challenge to common wisdom. Many take it as given that tech advancements spur productivity growth and eventually, higher wages and living standards – even if there are temporary periods of difficult adjustments for some. But Acemoglu and Johnson say this is not always so. To get past blind techno-optimism, we need to ask two questions: Who controls the path of a technology’s direction? And who will benefit from it?
A new technology can go in many different directions, say the authors, but for better or worse, it’s usually the powerful and influential who determine its path. Those who stand to gain through personal enrichment or otherwise often claim their vision is for the common good – but the many people affected by their choices rarely get consulted. And once a vision takes hold, it can be hard to break out of it.
Take the Chinese Communist Party’s decision to create a social credit system that tracks the nation’s billion-plus citizens through massive data collection and AI-driven surveillance. Only a few senior members of the Party decided on this route, but it affects 1.4 billion people who must live with its consequences every day. What’s more, AI investments there have become so heavily focused on facial recognition and censorship that many see China lagging in other areas of AI like natural language processing.
Another example is when Facebook’s top two executives decided to change the platform’s algorithm to increase user engagement so that more user data could be gathered for more profitable targeting of advertisements. The change affected Facebook’s 2.5 billion global users — and many more indirectly — because the effect was to prioritize the most inflammatory posts that kept people online and increased social and political polarization. Social media did not have to go this way, but somehow, we let two people at the top of Facebook make it so.
As Acemoglu and Johnson march through history’s major breakthroughs, from early crop cultivation through AI, we see how in each case, progress depends on the choices made. The early years of Britain’s industrial revolution enriched a few and impoverished many who experienced lower incomes, poorer health, and shorter lifespans for a century. In contrast, the spread of electricity and the Ford Motor Company’s productive use of new machinery in the early days of auto manufacturing led to gains that were widely shared.
The worry today is that we’re in a bit of a funk. While we may believe we’re in a golden age of digital innovation, it hasn’t shown up in the numbers. The authors point out that total factor productivity growth in the U.S. averaged 2.2% from the 1940s to the 1970s, but since 1980 has fallen to less than 0.7%. Low productivity growth has been a mystery to all – something Acemoglu readily acknowledges in public interviews. But one possible explanation may be what the authors call “so-so automation,” or technologies that simplistically remove tasks from humans without otherwise making them more productive. It’s very hard, Acemoglu has said, to see how self-checkout counters at supermarkets will move the productivity needle. More meaningful actions – like upgrading the work of nurses to take on more responsibility with medical technology – haven’t happened because this requires big dollar investments when we’ve mostly been focused on cost-cutting.
Again, it comes down to choices. And now that we’re facing a surge in AI, the choices are consequential. Alphabet CEO Sundar Pichai recently said that AI could be more profound for humanity than electricity or fire. But what will we do with AI? Slow it down? Regulate it? And who should be in charge? Acemoglu and Johnson would say that all of us – rather than a chosen few – should have a say.