Bad Idea #35 "THIS is the Future" Why Forecasts Fail – with David Wallace-Wells

The fact that this large-scale global paper was so sensitive to this one particular analysis of this one particular country, I think tells us something about how complicated, dense, and absolutely full of assumptions almost every model we try to use to project almost anything out into the future really is.

Oh my god. Um, can we do like 2030? I'd feel maybe a little more comfortable. There will be changes that when we look at them from the earth, if we were to look at them in 2025, we would think that is really dramatic. But almost certainly when we're living through them in 2035 or 2045 or 2050, they feel a lot more normal. And I think ultimately that's the sort of meta prediction that I would make, that we won't be on the other side of a massive singularity, whether that's about AI or about, you know, Chinese hegemony or climate apocalypse.

I am thrilled and honored and humbled to be here. So, thank you for having me. Great to find the time to talk.

Yeah. And this is for me, it's a bit of a personal story. So I've been writing about climate now for almost a decade, which is a little bit hard to believe. I also wrote a lot about the pandemic when that happened. And I've sort of found my way to a focus on what I like to think of as the messy future. Sometimes I describe it as the near future. And the longer I've spent with that material, that is to say, trying to think about where the world as we know it is heading 6 months, 2 years, 20 years down the line, honestly, the more humbled I am by that process and the more impossible that project really seems.

I mean, we're talking this week in early December. Just recently in climate world, there was a big paper that was retracted that was trying to project economic damages from climate change over the course of the 21st century. And this paper had a big finding. It said that economic output was going to decline by, I think, 62% by 2100 compared to a no climate change alternate reality. And it kind of came under some criticism almost immediately after being published, and now maybe a year, 18 months later, they've actually retracted it. And that in and of itself isn't actually even the thing that I think is important or illustrative here. It's the fact that the key problem with the paper was that they had made a miscalculation in discussing the climate effects on economic data, and that alone had contorted their overall finding so much so that rather than predicting a 23% decline in economic output from climate change by 2100, they predicted a 62% decline in economic output by 2100.

And the fact that this large-scale global paper was so sensitive to this one particular analysis of this one particular country, I think tells us something about how complicated, dense, and absolutely full of assumptions almost every model we try to use to project almost anything out into the future really is. And we can talk about, I imagine we will talk about, a lot of ways in which that shapes the way that we think about climate science in particular, our sense of the climate future and where we're heading.

But I have to say, like, the experience that I had that most brought this home to me was not about climate. It was about COVID. Which was that, and this is, you just, it's another story. So, apologies for telling a second story right off the bat, but I think it shows things even clearer. In the first case, we're talking about a paper that's trying to project 75 plus years into the future. We're talking massive political social assumptions about the way the world will look in the year 2100. In this case, in the second case, we're talking about the way that we modeled the spread of COVID in the midst of the pandemic.

Here in the US, we had the CDC collecting 26 modelers' best estimates for where the disease would be, not 2 years, not 20 years, not even two months out, but two weeks out from the present day. They collected the people they thought were doing the best job of predicting just two weeks into the future and publishing them in real time on the CDC website so that you could get a sense for where things were heading if you were trying to make decisions about your own life or theoretically think about what you think the appropriate political social response would be.

And in January of 2021, the CDC published these 26 scenarios. And two weeks later, the actual outcome lay outside the 95% confidence interval of 24 of those 26 models. Which means they got together their best people, the best people that we had anywhere in the world projecting the spread of COVID. And though they projected just two weeks out, and two weeks later only two of those models had even treated the ultimate outcome as mathematically possible.

This is how difficult models of any kind are, even when we have an extremely constrained timeline. Even when we have a very narrow band of question we're asking, like how many cases will there be of COVID 2 weeks from now. And the more I write about science, the more I think about science, the more I think about the near future, the more I worry that absolutely everything we think we know about where we're heading is bedeviled by the same basic epistemic problems. We are building future projections so much off of our own projections that we really, really can't see that future all that clearly.

And I think, I mean, it's telling enough to try to transport back in time to the 1950s, but you can even transport back in time 8 years and say we weren't anticipating the pandemic and the way that that transformed climate politics, the way that it produced inflation and the way that that transformed our understanding of the relative priority and urgency of decarbonization. We weren't anticipating Russia's invasion of Ukraine and the way that that precipitated an energy price spike that transformed the politics of energy all around the world. I mean, that's not 50 years in the past. That's 6 or 8 years ago. And those are some of the major world historical forces that have shaped the present in 2025.

They were not at all on the horizon a decade ago when the Paris agreement was signed or when the 1.5 degrees report was put out in 2018, which is the moment that really gave the sort of climate left, the activist left of the climate movement, the talking points that it used to mobilize people the world over. So much has changed even in that amount of time.

But I also think it's, I want to be careful because I'm guessing you'll agree, but I think that in the climate area, in the climate space and the climate story in particular, humility of this kind has often been weaponized by people who say that we don't have to worry at all about what's coming. Sometimes because they say that the science is going to prove to be hysterical itself. Sometimes because they say that when we try to incorporate even good science into models of the human future, we're leaving out so much that we're telling ourselves a sort of apocalyptic fable rather than allowing for human ingenuity and adaptation.

But I think at an emotional level, what that message really tells us is that we can discount futures that scare us because we don't know for sure that they're going to happen. And I think, I don't know exactly how to assess the math, but I would say it's something like as likely that we're as wrong in a scary direction as we're as wrong in a comforting direction. And that we need to take that very much seriously, too. And any effort to humble ourselves and to be modest in what we can and can't know about the future, we should take seriously the warnings of people like Marty Weitzman who said it's not the stuff we know that scares us. It's the stuff that we think is probably unlikely but could happen that's really quite scary.

It's since been surpassed, but it was at the time, yeah.

Yeah. Well,

Well, it's actually interesting on some of these points. So I wrote the article which came out in 2017, was a self-conscious survey of worst case scenarios and it was meant to scare the reader among other impacts, but really to showcase that some things beyond what we were being promised were in store for us were possible, that things could be worse than the median projections. And even when policymakers and scientists and journalists talked about median projections, they tended to sort of underplay what that meant, as you know as well.

And the original line in that piece, which was about worst case scenarios, was actually, 'It is, I promise, worse than you think.' The book, which was more in line with those median scenarios, I actually rewrote the first line in a scarier direction, so that is, as you say, 'It is much worse than you think.' I think I would still sign on to some version of the original first line, which is to say, 'It's worse than you think.'

I think that to some large degree the world has since then tried to wrap their minds around the fact of global transformation at the hands of industrial civilization and everything that we're putting into the atmosphere. I think that we have reckoned with, at least readers of mine, readers of the New York Times, readers of yours, have started to think about climate change as one of the grand forces of this century which will shape the human future, have thought about how it may shape geopolitics, have thought a little bit about what it means for the fate of developing nations, and have probably thought a little bit about what it means for their lives and the lives of those around them.

And that means that we're living in a bit of a different world socially and culturally, to some extent politically, than we were 8 or 10 years ago when I think a lot of that reckoning had yet to be done. That said, I do think that the future that we're heading for now in 2025 looks in a lot of ways like a more manageable future than the one that I was writing we might fear in 2017 or 2019. I don't have all that much certainty about that for some of the reasons that I mentioned at the top.

I think we can say with some certainty that the 21st century is not going to be dominated by coal in the way that most energy modelers and climate scientists said would be the case a decade or 12 years ago, producing a whole raft of science that was based on what were basically unreliable future energy projections. I think we can say with some certainty that the green energy rollout is going to be much faster than most people predicted 6 or 8 or 10 years ago.

There's a lot more green tech than you look at all these. Every year the International Energy Agency makes its projections for solar and every year it gets obliterated by that reality. Those projections are shown to be too pessimistic every single year. So I think we can also say that green energy is proceeding and being installed much more rapidly and expansively than most people predicted.

But there are still huge unknowns about where that leaves us. We don't know at what pace we're going to be genuinely retiring fossil fuels. I'm actually considerably less confident in the decarbonization of agriculture than I might have been 5 years ago because a lot of the stuff that people seemed to think was promising then doesn't seem nearly as promising now. And on the flip side, we are doing better on what used to be called the hard-to-abate sectors, industrial emissions from things like making cement and steel. And on balance, maybe those two things balance out, but I don't really know.

And then beyond all that, I would just point to something you said a few minutes ago, which is just this unbelievably complex Earth system, climate system that we are toying with. We know we are toying with it. We're messing with it much more dramatically and much faster than humans ever have before. In fact, much faster and more dramatically than the planet has ever been messed with by any force in the entire planetary history of Earth. But we don't know exactly what that means.

We don't know whether a certain amount of emissions is going to land us at 2° or 2 and a half. We don't even really know for sure that a certain amount of emissions will land us between two and three as opposed to north of four. We can say that with some confidence, but again, I'm trying to be humble about everything that we do and don't know. And the truth is we have really big questions about how sensitive the climate system is to these perturbations, which are historically enormous.

And I'm sure your listeners know, but just not to put too fine a point on it, the data point that I like to cite about just what scale of experiment we're running is that the carbon that's currently hanging in the atmosphere from our emissions weighs more than everything that humans have ever built on the surface of the earth. It weighs more than all the biomass on Earth combined. We've already done that much damage to the atmosphere. We've already transformed the atmosphere that much.

We just don't know what it is going to mean in precise detail for sea level rise in 2100 and sea level rise in 2800. We don't know exactly what it'll mean for the rate and pace of heat waves across Europe or across the Middle East. We don't know exactly what it means for the ocean circulation system. We are learning more, and I hope that the message here isn't that we should toss science out the window and not pay attention to it.

But I think every time now when I read a new scientific paper projecting what the planet will be like at three degrees, I try to file it alongside other papers on the same subject, not as a definitive and comprehensive new statement, but as one of many contested and contestable hypotheses, probably all of which we have to take into account if we're projecting that future.

Now, that all said, yes. You're asking me in 2025, am I the alarmist that I was in 2019 or 2017? The answer is no. Part of that is relatively good news from the energy system. Part of that is my maturation, my understanding that there's more to this story than the climate impacts, that humans can respond. And part of it is my understanding, which I wrote a little bit about in the book but has come to be a bigger part of the way that I think about how humans adapt to this ragged future, is that we just normalize a lot and accept a lot, and that is probably going to govern a whole lot of our future too.

There will be climate disasters beyond the kind that we've seen in the past. Probably to some extent we will be able to limit the human damage of those disasters, but to some extent we won't. And probably we'll sort of define the gap between those two as the acceptable drift of the climate system and find ways to live relatively normal lives even on a world that 5 or 10 years ago we would have considered horrifying and if not destroyed.

And I think about that with regard to the particular disasters that we've seen in recent memory. I mean, I'm working on a story right now about the Los Angeles fires of last January, which is not a narrow, simple climate story. It's basically an urban firestorm. Most of the fuel was the homes and the properties of the homeowners there. Nevertheless, it had some climate component.

And even beyond the question of climate's relative responsibility, you just, this is an absolutely mind-bending, horrifying, tragic human disaster. A very visible, large, very wealthy, well-connected, well-resourced neighborhood in what you could say is the cultural capital of America or even possibly the world being incinerated overnight with 1,600 homes destroyed and dozens of people killed and literally a part of the city just getting wiped off the map.

And here we are less than a year later and it just isn't a big part of the way that we're talking about what happened in 2025. And that is an indictment of us. It's an indictment of our human empathy and our capacity for compassion. But it's also, I think, a sign of the way that we'll navigate a world pockmarked by so much more climate disaster. Noting it quickly, maybe being moved by it, being angry by it for a time, but then moving on.

And if I had to tell you my mental model for what the climate future is going to look like, it's that there's going to be a lot more human suffering than we would have wanted, than was necessary, and probably many billions of people living on the planet will hardly even notice.

No, no, it's an incredibly normal reaction. My reaction, which is to like a sociopath keep turning back to the pandemic, is I think the unusual one. But I do think in a really profound way we can say that that experience really reshaped a huge amount of our society and politics and culture, and that one of the things that we're missing by wanting to move on and forget about it is just how differently we're living and thinking about one another and our obligations to one another than we did before then.

And I think speaking for myself, I think we're living in a world that is considerably more selfish. I think we're living in a world in which we are unwilling to do much for others and unwilling to tolerate safety precautions, even those that are undertaken on our behalf. We consider the intrusion, the burden of those kinds of actions to be too much almost no matter what the human cost is. I think we are more atomized and isolated.

We regard each other, this term that became popular over the course of the pandemic, lifted from gaming, but it went far beyond that. We regard each other as NPCs, as non-player characters. Our politics, I think, have been deformed by those forces, too, where we regard the other side in even less humane ways than we might have a few years before. And I think we basically, for someone like me who has the kind of political commitments that I have, it's really dispiriting.

We started this period, the pandemic period, in a spasm of incredible solidarity, health solidarity. We basically all around the world, not just in one society or another. Americans like to talk about our experience as though it was exceptional, but if anything, we had a looser pandemic experience than most people in Europe did. All around the world, we undertook a dramatic response to try to defend ourselves against this threat, to protect ourselves and those around us.

I think at some point in the spring of 2020, there were more than a billion school children out of school. And considering there were only 8 billion people on the planet, it's a remarkable fact. Basically, nobody was going to school anywhere on the planet. We suspended our romantic lives, our social lives. This was a major burden. But it was also to me inspiring to think that in response to a threat that we didn't even know existed a few months before, we had radically upended our way of life in an effort to limit the damage.

And that really wore on us. Most people decided, especially after the availability of the vaccines, that they wouldn't want to do that again. And they wouldn't want to do that kind of thing for anyone no matter what the threat. Here in the US, we now have laws in a bunch of states prohibiting public health officials from ever instituting future guidance against mitigation, which is to say, even if we had a disease as infectious as COVID, that was 50 times as deadly, they are prohibited by law from imposing restrictions on movement.

Now, I would assume that if we truly saw a pandemic of that scale, people would change their minds and get scared again. But even so, it is a kind of sad development that we had this incredible gesture of human solidarity and then collectively basically decided we don't want to do that ever again. It's not worth it. We should have just skated through the pandemic, letting the old and vulnerable die so that the rest of us who are relatively less old and vulnerable could thrive.

And I just want to say for clarity, like that paradigm about who is vulnerable and who is not is self-serving. We are the survivors. There were plenty of young and middle-aged people who got very sick. Some of them died. It's true that the age skew was quite dramatic and the older you were the more vulnerable you were. But we have a distorted view of the risk landscape because all of us who are talking now endured the pandemic.

And I think it's really made our society uglier. And I say that about our geopolitics. We have a much more nationalistic world than we had before the pandemic. I say that about our politics where, I mean maybe most vividly in the US, but not only there, much more combative, selfish politics. But I also think it's really true at the level of anecdote and personal lived experience.

I just think when I talk to my friends about the obligations they feel towards one another, the obligations they feel to the homeless in their city or the drug addicts in their city, in 2018, 2019, those would have been sources of, at least if you didn't feel like you should do something more for those people, you at least felt guilty about saying that. And now I talk to people who will half jokingly say, 'I just want those people thrown in jail.'

I've even had friends, again half jokingly, say, 'I think homeless people should just be killed.' And this is like the kind of thing that maybe some sociopaths thought in 2017 but nobody would have felt comfortable speaking out loud. And now I think we've given much bigger berth for that kind of thinking up to and including the American president and his followers who are considerably uglier, considerably more self-interested than they were even in the first go-round.

It's a really unfortunate and uncomfortable story. I would say a couple of things about it. The first is in the US we have had a major backlash against the COVID vaccines. There's a large minority of the country who thinks that they either had a null effect or had a negative effect. It's not a majority of the country, but it's a disturbingly large portion, something like 25% of the country feels that way.

But in 2021, 95% or more of American seniors got their COVID shots, which is to say the large, large majority of the most vulnerable people chose to get those shots. We actually had relatively good uptake among the young and middle-aged compared to anybody who would have been guessing before the pandemic how tolerant and willing to submit to such a shot the American public would have been.

We did less well than the UK. We did less well than most of the countries in Europe who got their seniors well above 99% and all adults above 90% relatively quickly. And the US failed that test compared to those peer nations. But starting from scratch and thinking, if you're like in 2019 and someone says to you there's going to be this novel virus that's quite scary. It's really infectious. It's pretty deadly but it's really infectious.

We're going to invent new vaccines using a platform that's never been even trialed in humans before. We're going to design those vaccines in two days over the course of a weekend in 2020. In January 2020 actually, before any American had even died of COVID. We're going to design them in two days, we're going to have them being manufactured within two months. Would you be willing to get that shot within a year?

The fact that we got 95 plus percent of American seniors to say yes to that and something like two-thirds of American adults to say yes to that is on some level inspiring and testament to, at least in the context of pandemic fear, the level of trust that people did have in the scientific authorities even as they were telling pollsters and ranting on social media that they didn't.

So I think that there is a kind of divide here between, you know, revealed preference and how you articulate your political views particularly on social media and particularly after the fact. So a lot of the people who are really ranting about these shots got the shots. That's one thing to note.

Another thing to note is that while overall vaccination rates have declined in the US since the pandemic, they actually haven't declined that much. We went from about 94% of kindergarteners getting all the full slate of shots to like 92%. That's the effect of the pandemic. And that's epidemiologically consequential because measles requires a higher level of herd immunity protection than 92%.

And there's certainly parts of the country that are well below 90 where we're now seeing outbreaks and that's quite troubling. But while it's epidemiologically significant, I'm not sure how much meaning to make out of the cultural change there. If we go from 94% to 92%, are we really living through a new dark ages when it comes to vaccines or are we living through something that would have looked like noise in the data in a slightly different time?

And on that point, I would also say despite the fact that the US has an unusual, maybe even unique culture war fight over science unfolding right now and it's unfolding in a way that I hate. Nevertheless, when I look around to our peer countries, I see vaccination rates falling similarly in the UK, in Europe, in Canada, in Mexico, and in fact, in all those places, there have been bigger and more deadly measles outbreaks than in the US.

Now, I don't want to downplay what's happened in the US. I think it's actually quite corrosive and quite horrifying that RFK and his band of COVID contrarians are empowered now. But I think the key point there to emphasize is that it's COVID contrarians.

And what I mean when I say that is that in what's called the MAHA Alliance, the Make America Healthy Again alliance, there are all of these contradictions. There are people who think that drugs are being approved way too quickly and that it's a handout to the pharma companies and we need to be much stricter about what we allow patients to take. And there are other people who think the FDA approval system is broken because we can't elect to take drugs that haven't yet been approved and that we should be allowed to undertake human challenge trials that will accelerate drug development. Those are intellectually, ideologically incompatible positions and yet they are...

Both part of the Maha Alliance. We have people who think that the great document of the contrarians, the Great Barrington Declaration, was at least nominally built around the idea that we need to protect the vulnerable, but we need to focus on the vulnerable in order to protect them and not pretend that we need to protect everybody. That was the central pitch of that document, which was published in the fall of 2020. Now, one of the central authors of that document, Jay Bhattacharya, is the head of the NIH, and he just wrote a big essay in which he said the best way to prepare for a future pandemic is to make America healthy again. We need to eat better. We need to lose weight. We need to stop smoking. In other words, there's no protecting of the vulnerable at all. He's just focused on getting people to become not vulnerable. So even in that one person, in his one narrative, in his development, there's a contradiction there. And the lesson that I draw from this pattern, where you have all of these basic disagreements: there are people who advocated for the vaccines as an absolute end to the pandemic in 2021. We don't have to worry about anything after you get that shot. Forget all these worrywarts and anxiety mongers. Just get the shot and forget about the pandemic. And then there are people within this coalition who insist that most of the deaths that we recorded were actually the result of shots from the vaccines. The fact that we have all of these contradictory elements in this coalition tells me that the coalition is fundamentally about the pandemic, not about science. Because the only way that you could organize these people together is through opposition to the public health establishment and a conviction that they, the powers that be, mismanage things. Once you get even one level deeper in that critique, it all is a mess and doesn't suggest a clear path forward. And I think you're seeing some of that in the way that RFK and his team are managing the question of vaccines, where they're doing a lot to sow doubt. You could tell that they want to really unwind the vaccine schedule. They really want to reduce the number of vaccines that kids take, but they actually haven't taken a very aggressive posture, fundamentally changed the nature of the sort of vaccination bargain in the country, because I think they understand that if they do that, they're going to be like whatever it is, 75, 80, 90% of parents who are like, actually, I want to get my kid vaccinated against these diseases. And that means I think that we're still living in the aftermath of COVID again in ways that we don't entirely recognize. If we fast forward five years, I don't think that this same coalition of people could be in charge, even if JD Vance is the president next time around, because I think the memory of the pandemic will have faded. The resentments that it produced, the frustrations it produced would have faded, and people will be asking themselves slightly more pointed questions about whether it really makes sense to talk about routine vaccination against measles, mumps, and rubella in the same way that we talked about the mRNA vaccines for COVID. Now, personally, I'm pro both of them. I think they're both miracles, but even if you object to the first, it doesn't necessarily mean that you're going to object to the second. And I think at the moment, we're living in an unusual period where our politics allows us to pretend that all that stuff is getting thrown out at the same time. Ultimately, I think, hope, pray that we're a brief period and we'll end up on the other side of it.