Andy Konwinski, co-founder of Databricks and Perplexity, is making a bold bet on the future of artificial intelligence—with a clear focus on its public benefit. On Monday, Konwinski announced the formation of the Laude AI research institute, backed by $100 million of his personal wealth. Unlike traditional AI labs, Laude aims to blend philanthropy and frontier research, offering grant-like funding while maintaining a flexible, startup-friendly structure.
Konwinski is assembling some of the most influential minds in AI on Laude’s board. Among them: UC Berkeley legend Dave Patterson, Google’s Jeff Dean, and Meta’s Joelle Pineau, all highly respected for their leadership in foundational AI systems research.
Laude’s first major commitment is a $15 million flagship grant to establish a new AI Systems Lab at UC Berkeley, which will be led by Ion Stoica, a renowned AI researcher and co-founder of Databricks and Anyscale. The lab is expected to open in 2027 and will bring together some of the best talent in AI infrastructure and systems design.
Beyond Benchmark Chasing: Building for Long-Term Impact
Konwinski emphasizes that Laude is not just another research vanity project or startup feeder. Instead, he says it’s “built by and for computer science researchers”, with a mission to guide AI in directions that offer widespread human benefit. That subtle framing may be a quiet critique of OpenAI and other labs that began with altruistic promises but have shifted toward high-stakes commercialization.
The Laude AI research institute is designed to straddle both nonprofit and commercial worlds, operating as a nonprofit with a public benefit corporation (PBC) arm. Konwinski categorizes Laude’s investments under two themes: “Slingshots,” for early-stage, high-potential research, and “Moonshots,” focused on long-term societal goals like healthcare, education, civic discourse, and re-skilling the workforce.
The institute has already been involved with Stanford’s “terminal-bench,” a benchmark used by Anthropic to evaluate how well AI agents handle real-world terminal-based tasks. This signals a clear interest in independent, meaningful benchmarks, rather than metrics designed to promote a specific model or vendor.
Where Research Meets Venture: The For-Profit Laude Arm
While the institute is nonprofit in nature, Konwinski has also launched a separate for-profit venture fund under the Laude banner. Co-founded with former NEA investor Pete Sonsini, the fund boasts a deep bench of over 50 leading AI researchers as limited partners (LPs).
In 2024, Laude led a $12 million round in AI agent infrastructure startup Arcade, and a spokesperson confirmed the fund has made several other stealth investments. Although Konwinski has personally committed $100 million, Laude is open to outside investment—especially from other successful technologists who believe in its hybrid model.
It’s no mystery how Konwinski is funding these ambitions. Databricks, which he co-founded, was valued at $62 billion after a recent $15.3 billion round. Perplexity, the AI search startup he also helped found, recently hit a $14 billion valuation.
The real question is whether the AI space needs another well-meaning research institute or hybrid venture vehicle. And in a world full of murky “AI for good” initiatives and vendor-backed benchmarks, maybe the better question is whether the Laude model—driven by researcher autonomy, long-term vision, and deep technical roots—can stand out as a credible force in independent AI innovation.
