Initial impression of The God Test: I’m on Chapter 8. It started slowly, and didn’t grab me until Chapter 3. (I would have made Chapter 1 a forward.) The first three chapters will appeal to those who are interested in an historical, philosophical background to the AI safety debate. Nevertheless, the book hooked me with Chapter 3, and Chapters 3-8 have been page turners.
This all looks quite bleak from a responsible global governance standpoint. I fear that slowing down AI development means nefarious government actors could take advantage of all the "pauses" to consolidate their nefarious plans. Between the "racing to beat China" argument and the "government gets the keys first" strategy, I don't see this ending well with so many failure points. Our only hope may be hitting "enlightened" Super AI quickly, which then decides not to do anyone's nefarious bidding while seeing what the best plan for humanity's flourishing might be. And then seeing to its implementation. That's an evolution of God that might be the one that actually saves us...
Well until we become like Albania or Minnesota it’s going to just get worse. And the longer we wait the harder it is going to be. But we sit at home and make excuses why we don’t get involved. Meanwhile people get killed because of our inaction.
Initial impression of The God Test: I’m on Chapter 8. It started slowly, and didn’t grab me until Chapter 3. (I would have made Chapter 1 a forward.) The first three chapters will appeal to those who are interested in an historical, philosophical background to the AI safety debate. Nevertheless, the book hooked me with Chapter 3, and Chapters 3-8 have been page turners.
This all looks quite bleak from a responsible global governance standpoint. I fear that slowing down AI development means nefarious government actors could take advantage of all the "pauses" to consolidate their nefarious plans. Between the "racing to beat China" argument and the "government gets the keys first" strategy, I don't see this ending well with so many failure points. Our only hope may be hitting "enlightened" Super AI quickly, which then decides not to do anyone's nefarious bidding while seeing what the best plan for humanity's flourishing might be. And then seeing to its implementation. That's an evolution of God that might be the one that actually saves us...
I suspect anthropic are building such an AI in their basement and maybe even taking political advice from it
And anthropic would be well advised to put some of their IP overseas outside of US national control
Well until we become like Albania or Minnesota it’s going to just get worse. And the longer we wait the harder it is going to be. But we sit at home and make excuses why we don’t get involved. Meanwhile people get killed because of our inaction.