View profile

Matt's Thoughts In Between - Issue #18

Revue
 
This week: what democracy is good for, the hotel that's a vision of the future, and whether we're mak
 
June 12 · Issue #18 · View online
Matt's Thoughts In Between
This week: what democracy is good for, the hotel that’s a vision of the future, and whether we’re making progress in AI…

What is democracy good for?
The LA Review of Books has an interesting essay by a Chinese graduate student at Oxford that looks at how Chinese liberals in the West often become disillusioned with democracy. It’s worth reading in full, but the essential idea is:
They became disillusioned with Western liberal democracy, because it did not live up to the promise of being a liberating force
Historian T. Greer has a good response, which unpacks “what democracy is good for”. It’s a fairly pessimistic conclusion: Greer says democracy is (empirically) pretty good at ensuring freedom, not so good at providing equality and terrible at promoting “self-government” - but it’s worth it nevertheless, as it at least channels political conflict into non-violent competition. 
It’s hard to dispute the idea that democracy isn’t living up to its billing (I highly recommend Democracy for Realists, which Greer links to, or at least this review, for a sobering dose of data on just how unresponsive the US political system is to the preferences of ordinary people). Nevertheless, if - as discussed in previous Thoughts in Between - liberal democracy faces a crisis crisis, it feels as though we need a more vital vision than the mere avoidance of violence and the secret police… 
EA Hotel: you can check in, but you can never pay
My friend and colleague Elspeth pointed me to this fascinating project: a hotel funded by an Effective Altruist (a movement we’ve discussed before) to provide free board and lodgings for two years to people who want to have time and space to maximise the amount of good they can do in the world. 
There are a number of remarkable things about this. One is the extremely low cost per person - around £5,400 per person, per year - which is an interesting datapoint for discussions about the cost of a viable Universal Basic Income (see previous discussion). What’s the catch? Well, depending on your perspective, one is that the hotel is in Blackpool, a sadly non-thriving seaside town in the north of England.
How much does that matter? The general version of this question is, I think, one of the most important of the coming century: how much does technology - for now, the host of services associated with remote work, but eventually seamless VR - eliminate the importance of place? Right now, cities command an extraordinary premium. If VR puts an end that, as my colleague Alex believes, perhaps the EA Hotel Blackpool is a mini-vision of the future.
AI winter or AI snowball?
It’s only six months since DeepMind’s AlphaZero seemed to make a major AI breakthrough by mastering chess using an apparently general game-learning technique. This and other advances prompted some people to wonder aloud whether artificial general intelligence was nearer than we thought. 
But in recent weeks, skeptics within the AI community seem to have had the upper hand. Filip Piękniewski wrote a much-shared post on what he sees as the coming “AI winter” - a historically cyclical phenomenon in which hype dies down and funding falls away. And Judea Pearl, an early AI pioneer, has a new book and a number of interviews that claim AI progress is stuck in a rut.
What to make of this? The critics have important substantive arguments, but to some extent these interventions represent esoteric methodological skirmishes. Some recent criticism seems to be attacking the dominance of deep learning (DeepMind and others’ favoured technique - though even they are not averse to other approaches), as much as identifying fundamental barriers to AI progress. Whether you’re on Team Musk or Team Zuck with respect to whether we should worry about superintelligence, everyone can agree the stakes are high - and internecine warfare is unlikely to do anyone any favours.
Quick Links
  1. The most important piece you’ve ever written. Kevin Simler solicits excellent links here. (I particularly liked this one on avoiding polarisation)
  2. IMF on Bitcoin. The IMF has written a (very good) report on monetary policy in the cryptocurrency age
  3. Rome wasn’t built in a day. Not even this gorgeous scale model.
  4. Easy mistake to make. How the UK Government “forgot” an island off the coast of Wales (Bonus semi-related anecdote)
  5. Cognitive outsourcing, spider edition. Do spiders outsource cognitive tasks to their webs?
Your feedback
Until next week,
Matt
Thanks for getting this far! If you’re enjoying Thoughts in Between, please share it - either by forwarding to friends or sharing the link on Facebook or Twitter.
Did you enjoy this issue?
If you don't want these updates anymore, please unsubscribe here
If you were forwarded this newsletter and you like it, you can subscribe here
Powered by Revue