How do we get serious about defending the planet from the changes that are causing climate instability?
What you hear from a lot of people working on these problems is that we have to be able to identify what we’re looking at, and project into the future, in order to do groundbreaking work on climate change.
It’s likely that as new AI technologies come online, they’ll be fundamentally helpful in this regard.
Here are some thoughts from experts who are looking into what we can do with modern systems.
The Fish Detectives: Identifying Activity and Taking Action
Nick Wise is the CEO and cofounder of the sustainability firm OceanMind. Talking at a recent conference, he laid out how AI can help evaluate and catch activity that is harming the environment.
“We need a healthy ocean to survive,” he said, citing illegal fishing and destructive fishing practices as a trigger for climate change.
Using the example of tankers hitting whales and passing through marine protected areas, he took aim at the idea that enforcement is too hard.
“Weak enforcement allows the destruction to persist today,” he said. “Marine enforcement is hampered by the belief that you can’t see what’s going on over the horizon.”
New AI systems, he said, can look at the ocean in “unprecedented detail,” and help regulators to enforce the law.
“AI is used to identify the key characteristics and information from the images, to correlate insights between data sources and to understand the infrastructure and the activities to support law enforcement,” he explained. “It’s not enough to know that an activity happened. You need to know where it started and precisely when it stops, you need to know the rules and regulations, and then you need to work out where these might have been breached. This is why we have developed an extremely accurate variant of AI that is very precise (in) understanding where fishing is happening. It understands 20 different types of fishing. It understands refueling and resupply operations, trans shipments, different characteristics of shipping, and many other vessel behaviors that you can compare with the regulations to identify illegal activities and damage to the marine environment over the past few years. Then we’ve used these capabilities to help Interpol capture most-wanted pirate fishing vessels, or the Thai government to reform their fisheries enforcement regulations, reducing overall (harmful) fishing by 30%, and they’ve started to see stocks begin to recover.”
Going into detail about work in places like Costa Rica and Britain, Wise cited massive amount of satellite data as supporting real change. This is one pillar of how AI is likely to aid climate science.
Research and Development Roles
Here’s an idea brought forward by Vanessa Chan, a data analytics intern at PwC, who talks about the difference between entrepreneurship and innovation.
She gives the example of Thomas Edison versus Joseph Swan, in the invention of the lightbulb.
“Breakthrough innovation isn’t the finish line,” she says. “It is a starting gun. … So many people think of research is being creative, but deployment is the ultimate creative act.”
As for Edison:
“He was able to take the light bulb and scale it. He was able to manufacture it at a lower cost. He was able to create relationships with utilities so that they would not just deliver gas to homes, but deliver electricity. And he understood the system, which is why history remembers him as a father of the light bulb.”
Talking about new programs at the Department of Energy, she mentions hydrogen hubs and other renewable projects.
It’s easy to see how AI might help here. Deployment is a logistical thing, where innovation is much more creative in a sense. So there is a good point to be made about how the technology will bring us from a dream project, to a reality in a prototype or beta phase.
Crossing the Valley of Death
I’ve seen Andrew Lo speak a number of times on key issues around AI and its applications to our world.
In some of his climate talks, he mentions trying to get to a net zero environment, and how AI might help us get there.
First, he concedes that data centers are going to use more energy, and that adds to the challenge.
He also mentions another big hurdle to getting where we want to go – the valley of death.
“It’s not a geographical location,” he says. “It is the area between early stage scientific discovery and the translational process that brings those discoveries into commercial applications. And we are being challenged with funding that early stage development, and that’s what the valley of death is all about.”
Why do we have this problem?
“The valley of death exists in many different industries, biotechnology, energy, scale up and manufacturing, semiconductors, they all have a version of this valley of death, the missing middle, the scale up problem,” Lo adds. “And it turns out that there is a common theme. The common theme is increasing risk and uncertainty. Risk is the kind of randomness that you can parameterize, probability of success, statistical distributions, but uncertainty is the unknown unknowns that you can’t parameterize. And both have been increasing over the last several decades, particularly in the areas of hard tech, which includes clean tech.”
Giving the issue the example of portfolio rebalancing, he says scientists are working on better organization, partnerships, and policy for clean technology, with things like nuclear fusion and geothermal.
One fundamental idea, he suggests, is to break the whole thing down into components. Lo contends:
“We need to have a process by which you can break down this big challenge into sub components that have clear milestones, and where investors can come in and come out, and where there is some independent third party that certifies that you’ve reached those milestones.”
Capturing Hearts and Minds
Another good insight, I thought, came from Andrew Jones, who talked at the Planet Action conference recently about the tragedy in Asheville, North Carolina.
He explained what happened in a scientific sense…
“The Gulf of Mexico, over the last century, has just gotten warmer and warmer, and warmer and warmer oceans make hurricanes bigger and more powerful. So Helene was able to travel 450 miles on a single tank of gas, twice the width of Katrina, more powerful than Katrina 2020 years ago.”
Sadly, he noted, data seems to show that giving people the numbers just doesn’t work in making them sit up and think about tackling climate science.
On the other hand, he suggested, simulations can be real wake up calls for viewers. He mentioned a study by Juliet Rooney Varga at UMass Lowell.
“New experiences change hearts and minds, and it’s incredibly powerful when we can help people see, through simulation, what it could look like out in the future if we really take on these important issues,” he said. “My friends, the future is not yet written. We can shape it. Take this simulation. Take other simulations, to your top leaders, to the businesses, to communities, to the schools.”
All of this was inspirational to me as I thought about how we can move forward and tackle the biggest problem of our age with the technologies that we are building in discovering. Let’s think about that as we bring 2024 to a close.
