V interesting piece. Would be interested in a follow up piece focussing more on the transition from the current economy to whatever the new economy is. While there may be a point at which the economy is doubling rapidly, that won’t be day one. If you have, say, a decade in which data centres, robotics factories, mineral infrastructure, dismantling of existing human-focussed regulatory structure, etc is being built out, with AI gradually eating a larger and larger share of the economy, then it will take time to get to the instant doublings of the economy that will sustain high levels of UBI/abundance. If you’re a highly paid accountant, who loses his job to AI, how long before UBI gives you the equivalent of your salary? In the West, people take on high levels of mortgage debt. Unless UBI operates at a salary replacement level from the beginning, I would expect: 1) high proportions of people who lose their jobs losing their homes; 2) a credit crunch in the housing market, as lenders are unwilling to lend over the long periods e.g. 25 years, at which most mortgages are offered, due to threat of widespread job losses; 3) a substantial drop in house prices, leaving a lot of home in substantial negative equity. AI will not only erode labour as a source of income, but will erode the value of the main store of wealth most people have, during the transition to the new economy. As AI, and access to intelligence, reduces the moats of existing real world companies, and exposes them to unprecedented levels of disruption, it will also become difficult for many existing businesses to borrow. Unless they are seen as part of the AI disruption - or as undisruptable by AI - lenders will be unwilling to lend for anything other than very short periods at very high rates. Otherwise going concerns will go to the wall because the cost of finance becomes so high. This will bring job losses forward - you may lose your job before the AI is ready to replace you. If this kind of transition happens over 50 years then it’s so slow that nobody notices, and if it happens over 6 months, it’s incredibly painful, but then you’re through it. If it happens over a decade, then it’s potentially the most challenging- you can offer people the hope of an incredible future of abundance (if the big picture can be solved), but they will experience huge loss and pain in the immediate term. There’s an element of Turchin’s elite overproduction that you can throw into the mix - unemployed young people, but also unemployed middle class professionals who have lost their social status and their source of wealth. That’s a difficult political mix to deal with
Agree with most of this and very glad that you are thinking about it - my question is about how universality can work in a world divided by nation states? It seems entirely feasible that there could be a world where certain publics own models and other publics do not - societies where labor is abundant and freely accessible but where the wealth created is automatically transported to the society that controls the cheapest compute. What do we do to make universal mean universal?
In answer to the title question. When the lower class, and this includes myself, run out of food. We take desperate measures and hope the army is hungry too. Extreme perhaps, but this a real fear. I believe that decentralized AI is perhaps the best and only chance mankind has at true equality.
This is the first time in a long time I have been excited for technology in our modern world. I believe it will be the great equalizer amongst all humans.
I am glad you do not like the sound of Russian Roulette :), the middle ground is good. 50/50 nice.
After listening to you on Moonshots today. definitely worth watching.
I have come to the conclusion, that you may be the most level headed Artificial Intelligence commentator. I have watched many discussions about this topic.
The ones you are involved in seem to stay on track, with all participants hitting that 50/50, pure emotional vs pure technical bell-curve. A little trip to the Fat Tails is nice once in a while though.
I have printed off your last two essays. I think they will be a great reference point in the future, as well as a great tool for people just starting to take a closer look at this Global Changing Technology. It will change all disciplines from Computer Science to Habitat Building...
I do not see AI as a net negative for society.
Just in the fields of Engineering and Metallurgy, this will have profound developments that will spread from these fields into the everyday lives of all humans.
I feel in the early stages of your new project, the folks that subscribe to this Substack, should be 50/50 Nice.
I wish you the best in your endeavors.
I truly believe you are on the right path.
You have the ability to bring people along with you. An admirable skill that few care to practice in our fast paced environment that never shuts down.
You are definitely one of the top people to listen to on these topics. However, P(the-current-state-of-human-society) being so far from P(writings-of-emad) makes me wonder what your plan is.
V interesting piece. Would be interested in a follow up piece focussing more on the transition from the current economy to whatever the new economy is. While there may be a point at which the economy is doubling rapidly, that won’t be day one. If you have, say, a decade in which data centres, robotics factories, mineral infrastructure, dismantling of existing human-focussed regulatory structure, etc is being built out, with AI gradually eating a larger and larger share of the economy, then it will take time to get to the instant doublings of the economy that will sustain high levels of UBI/abundance. If you’re a highly paid accountant, who loses his job to AI, how long before UBI gives you the equivalent of your salary? In the West, people take on high levels of mortgage debt. Unless UBI operates at a salary replacement level from the beginning, I would expect: 1) high proportions of people who lose their jobs losing their homes; 2) a credit crunch in the housing market, as lenders are unwilling to lend over the long periods e.g. 25 years, at which most mortgages are offered, due to threat of widespread job losses; 3) a substantial drop in house prices, leaving a lot of home in substantial negative equity. AI will not only erode labour as a source of income, but will erode the value of the main store of wealth most people have, during the transition to the new economy. As AI, and access to intelligence, reduces the moats of existing real world companies, and exposes them to unprecedented levels of disruption, it will also become difficult for many existing businesses to borrow. Unless they are seen as part of the AI disruption - or as undisruptable by AI - lenders will be unwilling to lend for anything other than very short periods at very high rates. Otherwise going concerns will go to the wall because the cost of finance becomes so high. This will bring job losses forward - you may lose your job before the AI is ready to replace you. If this kind of transition happens over 50 years then it’s so slow that nobody notices, and if it happens over 6 months, it’s incredibly painful, but then you’re through it. If it happens over a decade, then it’s potentially the most challenging- you can offer people the hope of an incredible future of abundance (if the big picture can be solved), but they will experience huge loss and pain in the immediate term. There’s an element of Turchin’s elite overproduction that you can throw into the mix - unemployed young people, but also unemployed middle class professionals who have lost their social status and their source of wealth. That’s a difficult political mix to deal with
Just ask AI to look at it and predict ahead...
Agree with most of this and very glad that you are thinking about it - my question is about how universality can work in a world divided by nation states? It seems entirely feasible that there could be a world where certain publics own models and other publics do not - societies where labor is abundant and freely accessible but where the wealth created is automatically transported to the society that controls the cheapest compute. What do we do to make universal mean universal?
In answer to the title question. When the lower class, and this includes myself, run out of food. We take desperate measures and hope the army is hungry too. Extreme perhaps, but this a real fear. I believe that decentralized AI is perhaps the best and only chance mankind has at true equality.
Emad,
This is the first time in a long time I have been excited for technology in our modern world. I believe it will be the great equalizer amongst all humans.
I am glad you do not like the sound of Russian Roulette :), the middle ground is good. 50/50 nice.
After listening to you on Moonshots today. definitely worth watching.
https://youtu.be/lY8Ja00PCQM?si=HaCB0JUfAgh6sxZA
I have come to the conclusion, that you may be the most level headed Artificial Intelligence commentator. I have watched many discussions about this topic.
The ones you are involved in seem to stay on track, with all participants hitting that 50/50, pure emotional vs pure technical bell-curve. A little trip to the Fat Tails is nice once in a while though.
I have printed off your last two essays. I think they will be a great reference point in the future, as well as a great tool for people just starting to take a closer look at this Global Changing Technology. It will change all disciplines from Computer Science to Habitat Building...
I do not see AI as a net negative for society.
Just in the fields of Engineering and Metallurgy, this will have profound developments that will spread from these fields into the everyday lives of all humans.
I feel in the early stages of your new project, the folks that subscribe to this Substack, should be 50/50 Nice.
I wish you the best in your endeavors.
I truly believe you are on the right path.
You have the ability to bring people along with you. An admirable skill that few care to practice in our fast paced environment that never shuts down.
You are definitely one of the top people to listen to on these topics. However, P(the-current-state-of-human-society) being so far from P(writings-of-emad) makes me wonder what your plan is.
P.S. Thanks for the SDiff toys.
Good luck 🤞