Infinite Jerk

Hey there,

Yes, the title is a play on “Infinite Jest,” the encyclopedic novel by David Foster Wallace that operates as a nigh-perfect schism: you either love it or you hate it. Clocking in at ~900 pages with 388 footnotes, many of which have to do with the storyline itself as opposed to citing references or offering additional but not entirely necessary details, reading the book, as I did a few years back and likely will again soon, simulates what I imagine the experience of trying to digest the entirety of an IPCC report or other similar climate science synthesis reports would be like. All that’s to say that keeping up with the latest in climate science, especially if endeavored across all climate science, would be a Sisyphean affair. Still, the headlines of recent climate science are relatively clear: Global warming may well be accelerating. Let’s discuss what that means and what it requires of us.

The above is a quote from Andy Jankowiak, a NASCAR driver who quipped it casually over his headset to his crew towards the end of a race. He proceeded to make good on that commitment, pushing his car to the max with complete disregard for other drivers and any semblance of control. At max velocity, he careened into other cars and scraped against and along the wall while pieces of his car flew off—all without taking his foot off the pedal—until he did, indeed, see the checkered flag and crossed the finish line (the checkered flag denotes the final lap of the race). While he didn’t win, the “strategy” vaulted him and his team from tenth place into fifth. You can watch a replay here.

Reckless abandon may have a place in NASCAR racing; indeed, a modicum of it is likely a prerequisite for participation. It shouldn’t, however, characterize technological development or the way we manage our relationship to the foundational systems on which society depends.

Last month, a paper in Geophysical Research Letters reported that since 2015, the planet has warmed faster than ever previously recorded. Specifically, the rate of change of global warming, that is, not the warming itself, but the rate at which warming is accelerating, has nearly doubled since 2015. At the risk of pedantry, to spell that out a bit more, we’re dealing not just with a question or consideration of velocity, i.e., the direction and rate of warming, but of acceleration, i.e., the rate of change of that velocity.

Sitting here in Silicon Valley as I write this (where it is also currently ~90°F in Winter), there’s more “acceleration” conversation outside of climate circles than within them. If you spend much time online or in the tech and venture capital-adjacent spaces where I sometimes swim, you’ll quickly hear a cacophonous rabble of technologists, investors, and their hangers-on who spend a lot of breath talking about acceleration in the computational capacity and abilities of each new and improved AI model. The vision of those who celebrate “accelerationism” is perhaps best characterized by the belief that AI and advances in other technologies will solve more problems than they will cause. Sure, there may be a thin, albeit super fat, tail in the distribution where an artificial superintelligence leads to doom, deciding, for instance, that humans are superfluous if not detrimental to its designs.

For many, that seems to be a risk worth taking. But I’m not here to opine on the probabilities of that one way or another. Still, it’s not lost on me that we’re entering an accelerationist paradigm, one in which we’re either explicitly or implicitly taking the “until I see God or the checkered flag” attitude, across several seminal domains.

While the jury is out on the real risk distribution of more advanced AI, there’s far less uncertainty regarding the evolving nature of climate risk. I am also confident that most people aren’t cool with espousing the “God or checkered flag” attitude with respect to Earth’s climate and how we prepare for, mitigate, and manage climate change. But even as there is growing recognition of the risks of acceleration in climate change (just read the raft of comments on this recent New York Times article if you need some encouragement around whether climate risks are entering the common knowledge), as of yet, the lion’s share of investment and work in climate continues to focus on mitigation and adaptation. Those are very necessary pillars of climate work. And we urgently need to build another pillar focused on near-term climate stabilization, i.e., a confluence of strategies to arrest climate change acceleration.

Disambiguating climate destabilization

More important than AI accelerationism and everything that “movement” subsumes is the physical reality of what accelerating climate change portends. There is notable pushback against the results of the paper referenced above, which found that warming rates over the decade since 2015 have nearly doubled. As much debate as there is on this topic—about whether and to what extent climate change is accelerating and how much of the observations surrounding acceleration are driven by El Niño periods vs. not—we need look no further than the drivers of warming and climate change to appreciate why rates of change are increasing:

Greenhouse gas emissions: Emissions of most major greenhouse gasses globally remain at or close to all-time highs. Annual carbon dioxide emissions have increased more than 50% in my lifetime (since 1995 or so). And rates of increase for emissions of other, more powerful greenhouse gasses, like methane, are even higher: in terms of relative increase, atmospheric methane concentrations have increased much more than carbon dioxide concentrations have since 1800 (methane concentrations are up more than two-fold compared to a roughly carbon dioxide’s 50% increase.) The relationship between higher rates of emissions and faster rates of warming and climate change shouldn’t be particularly opaque.

Loss of natural reflectivity: As the planet warms, we also see losses of reflectivity that exacerbate warming; as reflective surfaces and particles are lost, more energy is retained in Earth’s net energy balance. The Arctic summer sea ice extent is a well-documented example, and a similar pattern holds across many Arctic and Antarctic regions. Reductions in the summer Arctic sea ice extent already account for ~25% of observed global warming.

Loss of human-induced reflectivity: Another factor driving additional warming is the reduction in sulfur dioxide emissions. When sulfur dioxide emissions, whether from coal power plants, ships, or otherwise, enter the troposphere (the lower atmosphere), they oxidize to form sulfate aerosols, tiny particles that scatter incoming solar radiation back into space. Sulfur dioxide pollution is decreasing globally as ongoing efforts to reduce it continue, given that it is an air pollutant that’s deleterious to human and animal health. Still, the loss of the artificial shield that atmospheric sulfate aerosols previously provided is estimated to have “masked” ~0.4°C of the ~1.5°C of warming that’s already latent in Earth’s climate system. Put differently, the world is already much hotter in terms of retained energy than it seems or feels. We’ve just engineered a substantial atmospheric cooling effect to offset some of that warmth. Now, we’re consistently diminishing that cooling. 

Diminishing buffers: Alongside the loss of artificial or additional cooling from sulfate aerosols, many other ecosystems and climate systems that have historically acted as buffers against greenhouse gasses and the warming they drive may be approaching or at their “carrying capacity.” The world’s oceans are a good example: so far, the ocean has absorbed 90%+ of the excess heat introduced into Earth’s climate system by anthropogenic global warming. Other ecosystems, ranging from rainforests to peat bogs, are also exhibiting signs of stress and of being at capacity, as it were, to continue acting as buffers. There are limits to the extent to which any of these systems can offset climate change; beyond those limits, the impact of additional warming will accrue and compound elsewhere across Earth’s climate system.

Additional reinforcing feedback loops: Many elements of Earth’s climate system are complexly interconnected and coupled. Changes in one invariably impact many others. As global warming proceeds or even accelerates, climate change not only becomes more pronounced. It also risks “tipping” many stabilizing systems out of equilibrium. Some of what we’ve discussed above, such as the loss of sea ice or the shift in ecosystems’ ability to function as buffers against additional warming, is emblematic of this risk. And there are many other such systems, changes in which could accelerate warming and climate change even further. Another good example is thawing permafrost; at present, Arctic permafrost stores 1,460-1,600 billion metric tons of organic carbon. If even a fraction of that carbon is released as warmer temperatures thaw permafrost, it will set back greenhouse gas reduction efforts and could hasten other tipping points, forming a vicious cycle.  

There are dozens of other such reinforcing feedback loops I could introduce, but I don’t imagine it will necessarily help my overarching project, which is to create a desire to do something more than it is to engender despair. Further, it is worth noting that truly “runaway” warming is unlikely, as explored in much more scientific detail than I can muster here. That said, with respect to tipping points and their feedback loops, the risks of significant change to Earth’s climate, well beyond what we’ve seen from global warming and climate change thus far, aren’t particularly remote, either. As described in an excellent report, “Parasol Lost”, from the Institute and Faculty of Actuaries:

While the best estimate for when the 1.5°C warming threshold will be reached points to 2035, recent analysis suggests it could happen on or even before 2030. Further, some tipping points, such as large-scale die-off of coral reefs, may have already been breached. Taken together, all of the above threaten to drive step changes in climate change and its impacts. It’s possible we’re already on a path to cross thresholds that tip the Earth into an irreversible, fast slide into a permanently altered state. The graphic below, taken from the same report mentioned above, visualizes this risk well:

Ultimately, the changes to Earth’s climate that this version of climate change could reap are too chaotic to predict perfectly. But they will almost certainly render the Earth far less supportive of human and non-human flourishing than it is in its current state, to say nothing of what it was like circa 1800. As such, the paradigm of action we need to shift into is one of stabilization, of arresting acceleration and destabilization. This requires responses beyond mitigation, adaptation, and carbon removal, all of which, even taken together, and even if deployed at much larger scales, won’t suffice to meaningfully reduce the risk of accelerating and cascading climate change we’ve outlined thus far. 

There are quick wins within mitigation and adaptation that should be prioritized, given that they deliver benefits faster than reductions in long-lived atmospheric carbon dioxide do. Examples include reducing methane emissions and halting deforestation; we should emphasize these as much as in the near term, if not more than, scaling cleaner energy and electrification (not that these things should be in competition, anyway). Further, we will need to take even more proactive steps to evaluate how to intervene directly in regions such as the Arctic to stall the loss of reflective sea ice and ice sheets (as one example among many). Fortunately, that work—whether we call it climate interventions or stabilization—is coalescing and gaining traction as we speak, and I’m excited to tell you more about programs I’m helping develop soon as they have concrete news to share. 

Whereas acceleration is the first derivative of velocity, the second derivative of it is “jerk.” Acceleration is more familiar to us because it is associated with reasonably quotidian experiences, like hitting the gas in a car (especially in an EV!). While we don’t think of “jerk,” i.e., the acceleration of acceleration, as much, it’s probably time to start, as there’s a chance it becomes an apt analogy for the changing shape and scope of climate risk. Climate change is already such a complex, global, and often abstract concept that it’s tough for people to wrap their minds around it, let alone to remain enduringly concerned about it, or, even more ambitiously, dedicated to doing something about it. We will need to build the muscles to conceive and communicate about it as such, though, as that’s a foundational capacity to being able to then elevate a step further to respond—whether with technology, policy, investment, or otherwise—to the even more complex epoch of climate change that we’re stepping into. On that front, if not on many others, I’ll gladly commit to not lifting pen from pedal until I see God or checkered flag.