Saving the planet is a software challenge too

Ask a technologist to tackle a problem and their instinctive reaction is likely to be: there must be a technological fix for that. So it is with climate change, as technologists devise ever more inventive solutions to combat global warming.

We can sow clouds with fine particles to block out the sun and cool the planet. We can invest in experimental nuclear fusion that generates energy by pushing atoms together, rather than splitting them apart. We can genetically modify coral reefs to make them more adaptable to warming seas. Reinventing the future is so much more fun than succumbing to the stop-the-clock rhetoric of other parts of the environmental movement.

Bill Gates, the billionaire philanthropist and Microsoft co-founder, has emerged as the most eloquent champion of such techno-optimism. In his book, How To Avoid A Climate Disaster, he argues we must invent and deploy breakthrough technologies if we are to stop pumping 51bn tons of greenhouse gases into the atmosphere each year.

With the resources that come with a $137bn fortune, Gates has mobilised support for Breakthrough Energy, a coalition of private investment funds, philanthropic initiatives and advocacy groups, that is backing more than 40 companies working on such technologies. Some of these ventures may be crazy, Gates admits, but others may prove transformative.

It is admirable that Gates spends his money in this way, rather than splashing out on a yacht, football club or Caribbean island, as so many of his fellow billionaires are inclined to do. And who knows? With luck, one of the technological solutions backed by Breakthrough Energy might yet achieve spectacular results and help save the planet.

Yet my concern about this fixation with technology is that it might encourage inaction and distraction. If we believe that ingenious inventors are going to save us tomorrow, why bother agonising over contentious environmental accords today? If we think that nuclear fusion is just around the corner, why build wind farms?

The critics of techno-optimism argue that the challenge is not so much one of technology, but of policy. We need to change the features of the economic system, rather than fix its bugs.

Mark Jacobson, a climate change expert at Stanford University, has over the past 30 years developed computer models that show that existing wind, water and solar technologies could fully decarbonise the electricity supply if deployed at sufficient scale. That would take us a very long way to dealing with the environmental challenge.

Achieving that goal, according to Jacobson, involves a lot of unglamorous work, winning over hearts and minds, opposing vested energy interests and campaigning for politicians who support green policies. But this campaigning is paying off as several countries, and some US states, have signed up to renewable energy road maps.

To my mind, this environmental debate echoes a parallel discussion among computer scientists about how to tackle the theoretical risk of runaway artificial intelligence. The fear is that we may one day create an AI system that surpasses human intelligence across most domains and poses an existential threat to humanity. 

Stuart Russell, a computer science professor at the University of California, Berkeley, is one of the leading voices in the debate on how to control AI. Intriguingly, he suggests a kind of precursor to runaway AI already exists in the form of multinational corporations. We have “programmed” these companies to prioritise shareholder returns and ignore externalities, such as resource depletion or environmental pollution. 

“If you think of the fossil fuel industry as an AI system, we set it up with the objective of maximising profit, and it has won,” he told me. “It is a machine that has human components but functions as a machine. It has outwitted the rest of the human race.”

This industrial “machine” has done an extraordinary job of mobilising resources and reducing the cost of harmful fossil fuels. As Gates highlights in his book, the price of a gallon of oil is cheaper than a gallon of soda sold in Costco ($1 versus $2.85). 

For AI experts, the challenge is one of alignment. As Russell puts it, we have to build machines to ensure “their actions can be expected to achieve our objectives”. Just as we need to design AI systems to remain human-centric, so we must rewrite the software for how our market economy operates.

Encouragingly, this is happening, if in a somewhat messy way. Governments are reshaping market incentives by rewarding environmental “goods” (subsidising wind farms) and punishing “bads” (imposing carbon pricing). Even Wall Street is coming around to the mantra “green is good”. Powerful investors are intent on “reprogramming” our corporations with a new environmental function. 

To be fair to Gates, he accepts that we need both the supply and the demand sides of the equation to change for radical innovation to succeed. Policy change is essential even if technological breakthroughs remain indispensable. “Techno-fixes aren’t sufficient, but they are necessary,” he writes.

Yet it is ironic that such a famous technologist, synonymous with the productive capabilities of software, is now betting so heavily on hardware.

[email protected]

Source link

Related Articles

Back to top button