Listen now | Noah Smith & Brad DeLong's 30:00 < [Length of Weekly Podcast] < 60:00. We get nerdier! Be warned! Key Insights: Brad cannot, in fact, reliably and accurately multiply two-digit numbers in his head… When people comment on twitter that we are a nerdy podcast, we respond by going nerdier..
Towards the end, there is an interesting discussion of whether infantry should be robots driven by AI or controlled by a human. It could well be a combination of both. In terms of speed, we have seen that an AI simulation air battle reliably beats a very good human pilot. OTOH, we have legions of players of "shoot 'em up" video games with quite realistic combat situations. It would be very "Ender's Game" for nations to train kids to be really good infantry housing videogames, then have them control battlefield robots, while letting AIs take over for some functions to increase speed of response against similarly equipped battlefield robots. If robots fight robots, are we back to wars of attrition? Will robots fight humans, e.g. wealthy nations fighting poor nations? I can't help feeling this is a poor use of robots.
If you have such technology, why not apply it in more work areas once the technology becomes cheap enough? The awful movie "Surrogates" shows a world where everyone is effectively controlling a human-like robot. A good movie is "Sleep Dealer" where poor country workers control robots in dangerous jobs in the rich world (the pay differential is still advantageous).
While wealthy people might well prefer real human servants, might most other people pay for cheap labor to control household robots to do all those things that you want to make life easier without having to have appliance automation? Asimovian robots but with much human wetware rather than positronic brains controlling the robot, or at least carefully managing it. Allowing an AI to take over for the tedious tasks, a single person could control many robots all from their place of residence. True "Mechanical Turks" using telechirics plus AI could have a huge future.
Finally, a reasonable discussion of Rainbow’s End. I’ve been wanting to talk about Vinge’s AR vision of the near future, but never encountered anyone who’d heard of it, let alone read it. It seems to me that “provenance” or “certification” of transactions were a memorable part of that vision.
Also, Brad could definitely do Pfizer ads and I hope he doesn’t.
Great reference to electricity at Hatfield House. Tech matured through the accretion of layers of abstraction
Towards the end, there is an interesting discussion of whether infantry should be robots driven by AI or controlled by a human. It could well be a combination of both. In terms of speed, we have seen that an AI simulation air battle reliably beats a very good human pilot. OTOH, we have legions of players of "shoot 'em up" video games with quite realistic combat situations. It would be very "Ender's Game" for nations to train kids to be really good infantry housing videogames, then have them control battlefield robots, while letting AIs take over for some functions to increase speed of response against similarly equipped battlefield robots. If robots fight robots, are we back to wars of attrition? Will robots fight humans, e.g. wealthy nations fighting poor nations? I can't help feeling this is a poor use of robots.
If you have such technology, why not apply it in more work areas once the technology becomes cheap enough? The awful movie "Surrogates" shows a world where everyone is effectively controlling a human-like robot. A good movie is "Sleep Dealer" where poor country workers control robots in dangerous jobs in the rich world (the pay differential is still advantageous).
While wealthy people might well prefer real human servants, might most other people pay for cheap labor to control household robots to do all those things that you want to make life easier without having to have appliance automation? Asimovian robots but with much human wetware rather than positronic brains controlling the robot, or at least carefully managing it. Allowing an AI to take over for the tedious tasks, a single person could control many robots all from their place of residence. True "Mechanical Turks" using telechirics plus AI could have a huge future.
IMO you started at the wrong end. The easy route is:
50 x 50 =2500
47 x 50 = 50 x 50 - 3 x 50 = 2350
47 x 47 = 47 x 50 - 3 x 47 = 2209
But perhaps the hard route was the point. Because it's there!
Finally, a reasonable discussion of Rainbow’s End. I’ve been wanting to talk about Vinge’s AR vision of the near future, but never encountered anyone who’d heard of it, let alone read it. It seems to me that “provenance” or “certification” of transactions were a memorable part of that vision.
Also, Brad could definitely do Pfizer ads and I hope he doesn’t.