Turing expressed pride in powering the next generation of AI co-workers, noting that as AI models start learning real-world workflows, realism has become more crucial than ever before. It elaborated that Turing is building reinforcement learning environments and expert training data to drive this next generation of AI co-workers—enabling the models to learn not only what to say but also how to act.
At this year’s SuperReturn Asia (Singapore), Amino Capital’s Larry Li shared the firm’s AI-powered ARM + LP Expert System on the “From Innovation to Commercialisation” panel, redefining LP–GP transparency. Unlike traditional delayed updates, the system lets LPs access real-time investment info and act quickly, with features like automated startup tracking (10k+ startups, daily SWOT) and custom filters. It benefits founders (faster traction), LPs (earlier insights), and the industry (end to “black-box investing”)—all within LPA/NDA compliance to protect sensitive data. Amino calls this an LP–GP collaboration paradigm shift.
Replit shares another BIG NEWS: it has closed its $250M Series C financing round at a $3B valuation. This round was led by Prysm Capital, with American Express Ventures and Google AI Futures Fund joining its existing investors, including YC, Craft, a16z, Coatue, and Paul Graham, among others. It extends a huge thank you to its customers, investors, and team for their trust and belief in the company.
Woz advances the view “Why code quality must be the new benchmark for AI app building” and shares it is protesting AI slop at Woz. It notes today’s AI app builders prioritize speed and flashy demos over solid code—hiding brittle issues like inline styles and un-reusable logic—and emphasizes building the first version is just 1% of work, with 99% focused on debugging, scaling and maintenance. Woz adds good code is boring yet modular, and poses the key benchmark for AI coding: “Can a developer understand the code again 6 months later?” This, it says, is the future it’s building.
