Andy Konwinski, a founding figure behind both Databricks and Perplexity AI, has unveiled a bold new initiative aimed at reshaping the future of artificial intelligence. On Monday, he announced the formation of Laude, a new AI research institute seeded with a personal pledge of $100 million.
Unlike traditional research labs, Laude is structured more like a funding hub for computer science breakthroughs—offering grant-style investments to researchers tackling pressing challenges in AI. The initiative brings together an influential board that includes UC Berkeley’s award-winning computer scientist Dave Patterson, Google’s chief scientist Jeff Dean, and Meta’s VP of AI Research Joelle Pineau.
To kick things off, Laude is funding its first flagship initiative: a $3 million annual grant over five years to establish a new AI Systems Lab at UC Berkeley. The lab will be led by renowned professor Ion Stoica, who also co-founded both Anyscale and Databricks. Set to open in 2027, the lab will focus on foundational AI systems research, powered by a team of celebrated academics.
Konwinski emphasized in his launch blog post that Laude’s mission is deeply academic, stating it was “built by and for computer science researchers.” His goal? To push AI research forward in ways that prioritize long-term societal benefit—not just product rollouts or corporate profits.
While he didn’t directly mention OpenAI, the contrast is clear. Once an AI research nonprofit, OpenAI now operates a sprawling commercial business. And similar shifts have caused concern across the research community—especially when benchmark tools, such as those developed by Epoch, get entangled with corporate interests and controversial aims like replacing human labor entirely.
Laude aims to sidestep that dilemma by straddling both academic and entrepreneurial worlds. It’s set up as a nonprofit foundation, with a parallel public benefit corporation to support scalable ventures. Research funding will be funneled into two tracks: “Slingshots” for early, high-potential projects, and “Moonshots” for big, long-term bets—think AI for healthcare, scientific discovery, civic trust, or workforce reskilling.
Laude is already collaborating with key benchmarking efforts like terminal-bench, a Stanford-led project used by leading AI companies like Anthropic to evaluate agent performance in complex tasks.
But Laude isn’t just writing research grants. It also runs a venture arm, which quietly launched in 2024 with former NEA partner Pete Sonsini. That fund includes over 50 top AI researchers as limited partners and has already backed startups like Arcade, a $12M seed-stage AI agent infrastructure company. Though operating separately from the nonprofit, both arms share a vision of aligning research with real-world impact.
Konwinski’s ability to bankroll this vision comes from the staggering success of his startups. Databricks, where he was a co-founder, was recently valued at $62 billion following a $15.3 billion funding round. Perplexity, the fast-growing AI search company he also helped build, hit a $14 billion valuation just last month.
While the AI ecosystem is increasingly crowded with “ethical” research initiatives, Laude’s combination of deep academic roots, transparent funding structures, and commercial pragmatism offers a refreshing new model. As benchmarks become marketing tools and AI labs chase product launches, independent research backed by respected leaders like Konwinski, Stoica, and Dean might be just what the field needs to reset its moral compass.
