View profile

TiB 190: Combinatorial innovation; Rawls and AI; climate and VC and more...

Revue
 
 
November 9 · Issue #190 · View online
Matt's Thoughts In Between
This week: How combinatorial innovation explains the economic history of the world; how AI might change our ideas of justice; a TiB podcast conversation on climate tech and VC; and more…

Welcome new readers! Thoughts in Between is a newsletter (mainly) about how technology is changing politics, culture and society and how we might do it better.
It goes out every Tuesday to thousands of entrepreneurs, investors, policy makers, politicians and others. It’s free.
Forwarded this email? Subscribe here. Enjoy this email? Forward it to a friend.
So many ideas, so much growth, so little time?
A compelling theory of long-run economic growth (a favourite TiB topic) needs to be able to explain at least two important empirical observations: first, the explosion of growth around the time of the industrial revolution, after millennia of apparent near stagnation; and, second, the slowing down of this growth beginning 50 or 60 years ago. In a superb new post Matt Clancy suggests a plausible candidate for a “deep law” of growth: combinatorial innovation. We’ve discussed this idea before (see TiB 124 and 138) in the context of technological progress, but Clancy cites literature that suggests that it could provide a profound explanation for the shape of human history so far.
The basic idea of combinatorial innovation is that combining two or more ideas sometimes produces a new and more powerful one. In theory, the more ideas there are to combine, the more possibilities there are. Clancy cites this paper that shows that explosion of new ideas could be self-sustaining - and, indeed, produces a chart that looks strikingly like the pattern of long-run economic growth. So why don’t we get exponential growth forever? What causes stagnation in a world of combinatorial innovation? Clancy introduces two sources of slowdown. First, as the number of possible ideas explodes, we run out of resources to explore them all (more recent evidence here).
Second, if we assume that the economic usefulness of combinations is normally distributed (which seems reasonable) and that the average combination is useless, then finding combinations that are better than the status quo will get harder and harder as the status quo improves. As Clancy says, taken together, this “suggests the ultimate rate of technological progress depends on how rapidly we can increase our resources for exploring the space of ideas”. As I’ve said before, AI might be one way to do this and I’d love to see more work on this - but, as Clancy says, even machines will run into hard limits in the face of the sheer cosmological vastness of the possibilities.
Towards a Theory of Justice... by gradient descent?
Political theory has had relatively little to say about AI; a new paper from DeepMind researcher Iason Gabriel, “Towards a Theory of Justice for AI”, aims to start to redress this. A Theory of Justice(ToJ) by John Rawls is almost certainly the most influential work of political philosophy of the twentieth century. It’s not too much of an exaggeration to say that since its publication in 1971, academic political philosophy has been, for better or worse, a series of replies to it (there’s also - doubtless for worse - Theory of Justice: The Musical).
ToJ puts the idea of fairness at the heart of its account of justice. Gabriel argues that it’s time to apply Rawls’ frameworks to the impact of AI on society. For example, core to ToJ is the notion that each person should have a set of basic liberties that protect fundamental interests. Gabriel notes that Rawls believed that this set of liberties would vary according to the technological character of society, and suggests that, for example, an AI-dominated society may require a right to privacy (I wonder, too, about Albert Wenger’s “right to be represented by a bot”)
One the most famous, and controversial, ideas in ToJ is the Difference Principle, which (in part) states that social inequalities are permissible only to the extent that they benefit the worst off (e.g. incentives to get rich might lift societal wealth enough that the poorest are better off than in a more equal world). Gabriel notes that this is an exacting standard but that AI could have a positive role to play in fulfilling it. I’m not a Rawlsian, but papers like this seem an important avenue for further research. AI will remake society in profound ways and public debate that goes beyond the short-term and political will be an important part of ensuring it works for all of us.
TiB Podcast: Christian Hernandez on VC and climate
This week’s TiB podcast episode is a conversation with Christian Hernandez. Christian is one of the founding partners of 2150, a venture capital firm dedicated to backing startups that are reinventing how cities are built, run, and maintained - with a particular focus on sustainability and reducing carbon emissions. In this discussion we explore what role venture capital and startups can play in addressing the climate crisis.
In my conversation with Christian, we discuss how - after a long career in big tech and generalist venture capital - he came to see addressing climate change as the most important goal for the next phase of his career. One of my favourite things about talking to Christian is that he’s a great book recommender. In our discussion, he lists several that had a big impact on him on his path to starting 2150, including All Hell Breaking Loose, Uninhabitable Earth, and How to Avoid a Climate Disaster. We also talk through several topics that Christian has written about, such as why climate change is a national security issue.
Climate is a topic I’d like to explore more on the podcast. My guest next time will be Michelle You, co-founder and CEO of SuperCritical, a platform that helps businesses achieve true net zero with carbon removal. If you have ideas for other guests who know this topic well, please let me know. In the meantime, I hope you enjoy this conversation with Christian.
Quick links
  1. Goodbye to all that. Striking (UK) polling evidence on the decline of optimism.
  2. How to learn chess (and other things). Not a quick read, but an excellent one from my friend Alex on the challenges of adult learning.
  3. If at first you don’t succeed… Corporate recruiters prefer failed founders over successful ones.
  4. Don’t be born in August. Interesting / disturbing study suggests the youngest children in a class are much more likely to be (mis)disagnosed with learning disorders.
  5. “Meh”? New data on how highly people rate their own happiness (why is America so different?)
How you can help
It’s my birthday tomorrow and all is want is… for you to forward this to a friend or share it on social media. Thank you!
And if you have comments, questions or recommendations, just hit reply.
Until next week,
Matt Clifford
PS: Lots of newsletters get stuck in Gmail’s Promotions tab. If you find it in there, please help train the algorithm by dragging it to Primary. It makes a big difference.
Did you enjoy this issue?
In order to unsubscribe, click here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue