I read the Giga article "Command from Heights" about strategy in a world with AI, I want to share some of the main trends from the article. Much of the text is devoted to the strategy of American AI. The article was read as an echo of Lenin's idea of ​​commanding heights as a way to transition to NEP - the state must control the main thing, that is, industry, finance, transportation. In the case of artificial intelligence, what awaits us is not the NEP, but a spread into the economy, but still the records remain. Yesterday it was labs, chips and data centers, tomorrow it's the "rail" to deliver AI to the economy. Multi-billionaires will be the ones who will create almost monopoly systems to create value in the economy. The article cuts to a section on exactly what it is, but the obvious things are payments and new financing (stables as a way to export dollars), distribution to billions of people, legalization of liability (licensing and jurisdictions), roll-ups and transformation of old companies through AI. In addition, there are several trends that are difficult to argue with, but important to remember in planning: - Education ceases to be a social elevator, the idea of ​​"get a diploma, find a job" is dead. Learning is still important, but it is no longer a guarantee. There is no longer a need for young experts in old positions. Either you will make a leap to post-scarcity (Idk, 10-20-30 years?), or you will find something to do yourself instead of looking in the market. Entry-level jobs are the first to suffer: companies don't so much lay off as they stop recruiting young people. - Citizenship is becoming more expensive, immigration as a closed exit: rich countries no longer need additional workers. A US or EU passport becomes one of your most valuable assets (although the EU is going straight to hell). - Data has value now, but not for long. Today, labs pay for unique data, especially for entire datasets of many years of work by large companies. But when the models are trained enough + start continuous learning, most datasets will decrease, so you should sell now. - Inequality is increasing, but among the rich.

I read the Giga article "Command from Heights" about strategy in a world with AI, I want to share some of the main trends from the article. Much of the text is devoted to the strategy of American AI. The article was read as an echo of Lenin's idea of ​​commanding heights as a way to transition to NEP - the state must control the main thing, that is, industry, finance, transportation. In the case of artificial intelligence, what awaits us is not the NEP, but a spread into the economy, but still the records remain. Yesterday it was labs, chips and data centers, tomorrow it's the "rail" to deliver AI to the economy. Multi-billionaires will be the ones who will create almost monopoly systems to create value in the economy. The article cuts to a section on exactly what it is, but the obvious things are payments and new financing (stables as a way to export dollars), distribution to billions of people, legalization of liability (licensing and jurisdictions), roll-ups and transformation of old companies through AI. In addition, there are several trends that are difficult to argue with, but important to remember in planning: - Education ceases to be a social elevator, the idea of ​​"get a diploma, find a job" is dead. Learning is still important, but it is no longer a guarantee. There is no longer a need for young experts in old positions. Either you will make a leap to post-scarcity (Idk, 10-20-30 years?), or you will find something to do yourself instead of looking in the market. Entry-level jobs are the first to suffer: companies don't so much lay off as they stop recruiting young people. - Citizenship is becoming more expensive, immigration as a closed exit: rich countries no longer need additional workers. A US or EU passport becomes one of your most valuable assets (although the EU is going straight to hell). - Data has value now, but not for long. Today, labs pay for unique data, especially for entire datasets of many years of work by large companies. But when the models are trained enough + start continuous learning, most datasets will decrease, so you should sell now. - Inequality is increasing, but among the rich.