Watching → the quiet transformation of the data pipeline Something underdiscussed is the fundamental shift in how AI is labelled & trained for the next era of intelligence.. As we move toward Superintelligence or 'ASI', AI models don’t just need more data.. they need high-quality, expert-generated data, not bulk-labeled outputs from gig workers. If V1 AI had its greatest impact on lower-skill knowledge work (transcriptions, note-taking, basic marketing), next-gen AI, 'Superintelligence,' will be defined by its ability to operate in fields requiring extreme abstraction, multi-disciplinary synthesis, & high-stakes reasoning. Think: theoretical physics, advanced mathematics, AI safety, neuroscience, cryptography, aerospace, climate science, synthetic biology, & geopolitical strategy Industries that have real impact on frontier technology.. leading to true 'Superintelligence' Superintelligence, n. “An intellect that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom, and social skills.” - Nick Bostrom . . . "The thing that I think will be the most impactful on that 5 to 10 year time frame is AI will actually discover new science. And this is a crazy claim to make but I think it is true. And if it is correct then over time I think that will dwarf everything else." @sama on Jack Altman's Uncapped FWIW I don't think it's crazy to think ASI will make new discoveries, but I do think one major test of ASI will be its ability to invent new science & technologies. And that'll need industry experts. . . . The Financial Times just spotlighted this shift, covering top players like Scale AI, Turing, & Toloka moving away from gig-based labeling & toward highly specialized, expert-driven data pipelines “Deep-pocketed AI companies are now willing to pay for more sophisticated datasets & experts from around the world.” “As leading AI groups such as OpenAI, Anthropic, & Google attempt to develop models that they claim will exceed human intelligence, there is a new push to focus on the quality of these datasets & hiring experts to examine complex problems.” This is where Turing seems to be leading the pack. As @turingcom CEO @jonsid puts it, the goal isn’t just to replicate human expertise, it’s to transcend it, leading to 10x productivity: “The result of this is the model’s not just going to be better than a physicist. It’s going to be better than a superposition of somebody who’s at the top in physics, computer science, & data science,” said Turing’s Siddharth.