Posts

Comments

Comment by robert-miles on AI Alignment Open Thread August 2019 · 2019-08-27T16:27:40.529Z · score: 5 (3 votes) · LW · GW

Yeah, nuclear power is a better analogy than weapons, but I think the two are linked, and the link itself may be a useful analogy, because risk/coordination is affected by the dual-use nature of some of the technologies.

One thing that makes non-proliferation difficult is that nations legitimately want nuclear facilities because they want to use nuclear power, but 'rogue states' that want to acquire nuclear weapons will also claim that this is their only goal. How do we know who really just wants power plants?

And power generation comes with its own risks. Can we trust everyone to take the right precautions, and if not, can we paternalistically restrict some organisations or states that we deem not capable enough to be trusted with the technology?

AI coordination probably has these kinds of problems to an even greater degree.

Comment by robert-miles on Stories of Summer Solstice · 2018-07-12T06:43:40.686Z · score: 7 (4 votes) · LW · GW

My experience of that sudden abrupt silence at the exact instant that the very last sliver of sun dropped below the horizon wasn't a hushed and numinous rapture, so much as it was like "Holy Shit, that WORKED? Did we really all just stop dead silent at the exact same moment?" I've been in orchestras that struggle to do that.

Comment by robert-miles on Idea: OpenAI Gym environments where the AI is a part of the environment · 2018-04-14T13:12:57.452Z · score: 9 (2 votes) · LW · GW

The "Whisky and Gold" environment is particularly relevant

Comment by robert-miles on The Steampunk Aesthetic · 2018-03-08T23:31:03.777Z · score: 17 (5 votes) · LW · GW

If you would like to see this lifestyle/philosophy lived out fully in real-time, you can watch Jamie Mantzel's YouTube channel: https://www.youtube.com/user/JMEMantzel

He is living on an island in Panama, in a concrete building he built single-handedly, using a solar powered bulldozer he designed and built himself, collecting materials and components from the mainland on a solar-powered cargo boat he built himself, etc etc.

https://www.youtube.com/playlist?list=PLCcG44tLg9nUsQ38rPMbLg1uiUHZcmUMB

Comment by robert-miles on set of cards · 2018-03-07T16:37:46.375Z · score: 13 (3 votes) · LW · GW

This idea reminds me very much of Oblique Strategies. I guess the idea of that set of cards is to help with creative work - you draw a card when you're feeling stuck or uninspired, and the cards say something oblique like "not building a wall, making a brick", allowing you to give your thoughts a big shove in a pretty arbitrary direction in thought-space, which can push you out of a local optimum and get things moving again.

Inspired by this, I've thought about trying to do exactly what you're doing, and I can share the list I wrote down (with no claims whatsoever about their quality or suitability):

  • Compare the outside and inside views
  • Has this problem been solved before?
  • Is this the right problem to solve?
  • Have you tried the obvious things?
  • Ask someone else for obvious things to try
  • Your future self visits to tell you your plan failed. What went wrong?
  • Who would be better at this than you? Pick a specific person. What would they do?
  • Stop and make a list or two
  • Consider the opportunity costs
  • What are you avoiding thinking about?
  • Be more specific
  • Name three examples
  • Come up with a concrete example
  • Why are you drawing a card? Ask yourself "why?" to the response, 4 more times.
  • What would convince you that you are wrong?
  • What assumptions are you relying on?
  • What other problem is this most similar to?
  • Do a Fermi calculation
  • Go meta
  • Separate the parts of you that disagree, and let them have a conversation
  • Break the problem into smaller sub-problems
  • Take the contrapositive
Comment by robert-miles on Confidence Confusion · 2018-02-16T11:39:18.857Z · score: 18 (4 votes) · LW · GW

Perhaps the principled way is to try representing your probability to the same number of significant figures as a probability, as a log probability, as odds, and as log odds, and then present whichever option happens to fall closest to your true estimate :p

Comment by robert-miles on "Just Suffer Until It Passes" · 2018-02-12T13:26:25.992Z · score: 25 (8 votes) · LW · GW
a lot of problems aren't huge problems in-and-of themselves, and it's the flight to distraction and stimulation that compound the problem and create bad ongoing habits

This reminds me of a technique I use sometimes called Doing Nothing, described in this blog post, where instead of what you call 'flying to distraction' , you just don't do anything at all. You say "I don't have to write this paper right now, but I'm also not going to do anything else". You have the freedom to sit there blankly doing nothing at all for as long as you want, and generally it's not long before you stumble upon some new ideas or motivation to carry on.

Comment by robert-miles on Moloch's Toolbox (2/2) · 2017-11-20T12:22:11.011Z · score: 10 (3 votes) · LW · GW

The reason you can't just vote to change the voting system actually seems a lot simpler to me, based on my experience of the UK AV referendum. The system actually worked well enough to get the idea of switching to a basically sane voting system put to a public referendum. Then look at how that referendum goes. Who stands to benefit?

  • The public (but they don't know that)
  • The small political parties (who lack funding and are unable to mobilise large numbers of voters almost by definition)

Who stands to lose out

  • The major political parties (who have vast, well funded propaganda machines, good contacts, expertise etc)

So the red party told red voters that using a sane voting system would be bad for the red party, and the blue party told blue voters that using a sane voting system would be bad for the blue party, and the referendum came out as a 'No'.