Technology as a force for good, driving positive change – that’s how MIT scholar Andrew McAfee likes to see the influence of machines on the human world. In their international bestseller The Second Machine Age, McAfee and co-author Erik Brynjolfsson described a future powered by artificial intelligence and automation, concluding that “the transformations brought about by digital technology will be profoundly beneficial ones.”
Turning to sustainability, MacAfee argued in More From Less that we can achieve the seemingly impossible: consume fewer resources yet live more prosperous lives – thanks to advances in biotechnology, agriculture and renewable energy, among others. Most recently, he launched a research group named Tech For Good at the MIT Sloan School of Management to develop this idea further.
But progress also demands the right mindset. In a DLD Summer presentation aptly named “Shaking Up Europe”, McAfee chastised EU lawmakers for over-eagerly regulating technologies like artificial intelligence, at the risk of hampering the continent’s chances to stay competitive in the global economy.
In conversation with DLD, McAfee explains why he’s in favor of what he calls “permissionless innovation”, what could – and should – be done to battle climate change, and why intuition often leads us astray.
Andrew McAfee is the Co-Founder and Co-Director of the Initiative on the Digital Economy and a Principal Research Scientist at the MIT Sloan School of Management. He studies how digital technologies are changing business, the economy, and society. He is the author of More from Less and The Second Machine Age (with Erik Brynjolfsson).
Why does Europe continue to be so far behind Asia and the U.S.?
It’s actually a puzzle to me. Europe, the EU, is a very wealthy region of the world. It has very strong institutions. It has an excellent educational system. It’s got plenty of capital, the rule of law works here. These are things that would make you optimistic about the Union’s ability to thrive during this really profound transition.
But it’s not. What explanation do you see?
Part of the reason – and I want to be clear, this is only part of the reason why Europe is lagging behind – comes from the example that I gave in my talk about the approach the Union is taking toward regulating AI.
You mentioned that even a language-learning app like Duolingo could be seen as a “high risk AI system” under proposed EU regulation because the app uses sophisticated algorithms to evaluate its users.
It’s not clear to me why artificial intelligence is so powerful and so scary that using it to test someone’s English proficiency immediately puts the entire piece of technology in a high risk category that has all these requirements associated with it. I think that’s the wrong approach. I think that will not help Europe catch up to this period of intense innovation and disruption.
What would be a better approach?
One mental trick that I use is, whenever I see the phrase AI, I mentally substitute the word spreadsheet.
Because a spreadsheet is an extraordinarily powerful, very widely used technology that is opaque. In that way, it is similar to AI. If you look at a complicated spreadsheet model, you don’t know why it gave you the answer that it did. Yet we don’t have high risk spreadsheet applications where you have to satisfy all these requirements before you can go release that spreadsheet out into the world.
Can you really compare these technologies?
To be clear, there are areas where we need upstream governance. We need it for releasing drugs out into the world, for example, so that a factory doesn’t get to pollute until someone says, “Hey, you need to stop!” But there’s a deep difference of philosophy between permissionless innovation and upstream governance.
“Permissionless innovation” means…
First we’re going to let people do things, and we’re going to deal with the harms if, and when, they arise. With upstream governance you’re always trying to make sure that you can identify, and deal with, all the important possible harms in advance. I am not confident in the ability of any group of people to do that upstream governance well.
If you just let people innovate, don’t we risk a lot of unintended consequences?
I want to be very clear: Permissionless innovation is not without risks. It is not without costs. Absolutely. It’s not perfection. At all. There will be harms, and governments will be slow to catch up. That said, I am not sure that upstream governance will avoid all those harms, or will lead to faster government. But there’s always a spectrum between governance and permissionless innovation. And I am very far over on the permissionless side because I think that gives us better outcomes.
How do you see Europe’s chances in the transformation of the world economy towards more sustainability?
I think Europe has taken the lead in some really important areas. But we need a global response to climate change and resource depletion. And my great frustration is that we know how to simultaneously increase human prosperity and tread more lightly on the planet. I described the playbook for it in More From Less. But we’re not doing a great job of following that playbook right now.
What is it that we should be doing?
The two best weapons that we have in our possession, we are not using. Number one is a price on carbon. The United States doesn’t have a high enough price on carbon, Europe in general doesn’t. The Swiss just voted to roll back their increased price on carbon. So around the world in general, we’re not putting a price on carbon.
You’re saying we should increase the price to make carbon emissions painfully expensive?
Absolutely. Put a price on carbon and watch businesses respond when you change their cost structure. This has worked with all kinds of pollution in the past. Cap and trade has been a huge success for other kinds of pollution. As you know, economists love to argue, but almost every economist would agree that making carbon emissions more expensive is the single best weapon that we have to change behavior, to change what energy systems we use, and to change the scope of innovation in general.
What else are we ignoring?
We have one power source – one! – in the world today, that is potent, scalable, not intermittent, readily available and – here’s the controversial part – safe. Very, very safe. That is called nuclear power. And in most of the world, we are running away from it as fast as we can. So my great frustration is that we are not using the two best weapons that we have to fight global warming, because the first one disagrees with our pocketbook. And the second one disagrees with our intuition. Those are not great reasons.
We’ve seen a number of accidents with nuclear power plants. Why do you consider this technology safe?
My intuition is that nuclear power is scary and bad. I was a child when Three Mile Island happened, I was in college when Chernobyl happened. These are events that stay with you. But even taking those events into account, the evidence is overwhelming that nuclear is just about the safest power that we have. So we have to stop listening to our intuition. A lot of people feel genetically modified crops are scary, harmful violations of nature. Despite the fact that, from all we know, they are safe, and we need them to feed more people on fewer acres of land. So again, a big part of the reason that we’re not doing the right things: They hit our pocketbook. And they don’t correspond to our intuition. Those are two tough problems to solve.
Where do find hope that ultimately all will be well?
I’m massively optimistic on the technology front. Look at what happened over the course of the pandemic: We developed, manufactured and distributed a brand new kind of vaccine against a brand new deadly pathogen in the space of a year. That is a triumph, a triumph of science and technology. You have to be optimistic about that. I think we have every technology that we need to fight global warming, to reduce pollution around the world, to use fewer acres of land, to feed humanity and satisfy humanity’s needs, to make our air and our water cleaner. We have the technology toolkit, and that was painful and slow to develop. Green energy is not easy, and it’s not fast. But we have it now. It’s getting better very quickly. So on the technology front, I’m an optimist. The evidence says, we can do these things.
You still sound doubtful, though.
The main thing that I’m pessimistic about is this period of deep polarization that we’re seeing in many places around the world, and people turning away from evidence and rationality and reason in the face of these big challenges.The anti-vaccine movement is real, the people who refused to wear a mask while the pandemic was raging. These are real phenomena. And I don’t know the magic technology to solve that problem.
Could it be that there is none, because we’re talking about human vagaries, something that cannot be solved by technology?
Yes, looking to technology to deal with this human situation might be the wrong approach. We need to start listening to people who understand how to bring somebody along and change their opinion, and get them to look at the world differently. What we do know is that bludgeoning people with facts and calling them names, because they have other opinions, that doesn’t work. So we better figure out what does work.
Shaking Up Europe: Watch Andrew McAfee’s DLD Summer presentation