Matt's Thoughts In Between - Issue #71

Forwarded this email? Subscribe here. Enjoy this email? Forward it to a friend.

The State of AI in 2019

Ian Hogarth and Nathan Benaich published the 2019 State of AI report (see here for coverage of last year’s edition). It’s a superb and comprehensive read that covers all aspects of AI from recent tech developments to the geopolitical dimension. Do read the whole thing.

The report hammers home that we remain in a period of rapid improvement in AI and, just as importantly, ever expanding scope of what AI can do. The apparent triviality of some past breakthroughs in AI - identifying cats in videos or beating 1970s video games - is long gone. Ian and Nathan highlight important new results in protein folding, molecule synthesis and warehouse automation.

Another striking conclusion is the centralisation of cutting edge AI in a small number of organisations - and the growing structural barriers to reversing that. DeepMind’s Starcraft model, AlphaStar, took $26m in computational resource to train, while remuneration for top AI talent is now close to $1m annually. As Ian and Nathan note, this makes the question of AI governance increasingly important - but there are not yet any successful models (and, as the public attitudes section makes clear, no consensus on what good looks like). This is troubling, and an important space to keep an eye on.

Why did the US stop caring about the future?

We’ve discussed a number of times Peter Thiel’s idea of “definite optimism” - the belief that (a) the future will be better than the present and (b) it’s possible to get there by laying out a vision and executing a plan to achieve it (See here and here for previous coverage). For Thiel, the loss of definite optimism is one cause of economic stagnation and, as his VC firm Founders Fund has put it, “We were promised flying cars; we got 140 characters”.

The economist Ray Fair has a new paper out that provides interesting empirical evidence for the thesis that the US became a lot less future-oriented starting around 1970. Fair constructs two datasets that seem to suggest a strong preference for consumption today over investment in the future. First, he demonstrates that US infrastructure spending has dropped significantly since 1970. Second, he shows that around the same time the US federal budget deficit became consistently large and positive.

What might have caused this policy “impatience”? Fair isn’t sure - though there are no shortage of hypotheses, from the idea that it’s just much harder to come up with new ideas now to a growing preference for self-expression over wealth generation. If pessimism doesn’t just reflect economic reality, but shapes it, this seems like an important and under-discussed question.

Software and the new political battlegrounds

"Software is eating the world” has become a truism. One of the many interesting questions is what political conflict looks like in a world devoured by software. TiB favourite Michael Nielsen has an excellent thread on the topic this week. He talks about the idea that as more and more human activities are mediated by software, political struggle migrates to being a conflict to control this mediation layer. Many of these conflicts are new.

Nielsen points to an interesting - if superficially low stakes - example: the fact that books can now be switched "off” remotely. Alex Danco’s excellent Snippets looks at another: the extraordinary story of Tether - a controversial cryptocurrency pegged to the US dollar, whose governance and business practices have been questioned by many. It’s worth reading the whole thing, but it's an example of major societal functions - central banking, fractional reserve banking, legal systems - starting to be mediated through (decentralised) software, with potentially enormous consequences.

Few existing institutions are well prepared for these changes. This week’s debates among Democratic candidates for the 2020 nomination barely mentioned technology (with the notable exception of Andrew Yang) and the UK’s Conservative leadership election is no better. As mentioned above, the intersection of tech and governance is growing in importance. We need leaders who understand that.

Quick Links

Fortune favours the foolish. Why sometimes getting a very big bet right is a sign of poor judgement (Full paper here)
The very rich are (not) like you and me. How do millionaires spend their time?
How do you add numbers in your head? A surprisingly interesting Twitter Q&A on how people think about arithmetic.
Remember, remember... remember. Fascinating results on how spaced repetition improves recall.
Renewed hope. Renewables pass coal in the US energy mix for the first time.

Your feedback

Thanks for reading Thoughts in Between. I'd love it if you'd forward this to someone who might enjoy it - it has a big impact on building the TiB community. And feel free to reply if you have any comments or talk to me on Twitter.

Until next week,

Matt