Ilya Sutskever: “Here’s a perspective that I think might be true. The way ML used to work is that people would just tinker with stuff and try to get interesting results. That’s what’s been going on in the past. Then the scaling insight arrived. Scaling laws, GPT-3, and suddenly everyone realized we should scale. This is an example of how language affects thought. “Scaling” is just one word, but it’s such a powerful word because it informs people what to do. They say, “Let’s try to scale things.” So you say, what are we scaling? Pre-training was the thing to scale. It was a particular scaling recipe. The big breakthrough of pre-training is the realization that this recipe is good. You say, “Hey, if you mix some compute with some data into a neural net of a certain size, you will get results. You will know that you’ll be better if you just scale the recipe up.” This is also great. Companies love this because it gives you a very low-risk way of investing your resources. It’s much harder to invest your resources in research. Compare that. If you research, you need to be like, “Go forth researchers and research and come up with something”, versus get more data, get more compute. You know you’ll get something from pre-training.”
FT: “India has doubled steel production in the past decade, overtaking Japan to become the world’s second-largest steelmaker. Prime Minister Narendra Modi’s government is now aiming to boost India’s steelmaking capacity to 300mn tonnes by 2030, almost double its level in 2024. In a global industry grappling with slowing demand in major markets such as China, where consumption has plateaued, India’s steel use bucks the trend, forecast to rise about 6 per cent a year through 2035. “India is the best story as far as the steel industry is concerned, globally, there’s no other major market growing at this pace — in fact, most markets are shrinking,” says TV Narendran, chief executive at Tata Steel, one of the country’s biggest producers of the metal. “India is underinvested in infrastructure.”…Booming production may buoy the domestic economy, but is causing environmental damage and trade tensions.”
WSJ: “The overall goal of sovereign AI is to build out a localized computing infrastructure and potentially possess domestic capabilities in everything from large language model development and chipmaking to cloud storage and engineering talent. Many also hope to regulate and manage the data used and generated by AI on their own terms…South Korea and others believe their homegrown tech sectors are strong enough to build up their AI capabilities.”
Henry Oliver: “The classical liberal project has become too narrow. It is too often perceived as overly economic and technocratic: as something that ignores the deepest problems of human life, and is unable to speak to the whole human experience. This is wrong! Alongside its advantages for growth and prosperity, liberalism offers greater meaning, dignity, and a sense of community than any of the political alternatives…Liberalism is already gaining new cultural and practical relevance. New interest in the liberal arts is coming from technologists and others online. Classical liberal answers are emerging to solve today’s problems. AI is disrupting the old order of intellectuals, just as the internet once did. The Pursuit of Liberalism is an experiment in this spirit. We are as committed to advocating for human flourishing, as to describing economic reality. We will argue for the enduring relevance of the arts and philosophy to classical liberalism. Our aim is to revive the broad classical liberal tradition. We will argue for its place in an innovative world. We don’t need a new liberalism. We need a new way of helping people to see it for what it is.”
