Andreessen Horowitz is starting a new research unit designed to bring academics into tackling some of the hardest problems in Web3 and crypto.
It’s called A16z Crypto Research and it’s modeled after research labs such as OpenAI and Alphabet’s DeepMind. Tim Roughgarden, a professor at Columbia and Stanford who has written extensively on Web3, is its head of research.
“Usually whenever there's a new groundbreaking technology that has general purpose and can be used for all sorts of different things, there's an opportunity for an industrial research lab to bridge the world of academia with the world of industry practice,” said Ali Yahya, a general partner at Andreessen Horowitz who previously worked at Google on Google Brain.
Dan Boneh, a professor at Stanford, is being promoted to senior research adviser and four others are joining as research partners: Joseph Bonneau, who’s written extensively on cryptocurrency; Benedikt Bünz, a Stanford Ph.D. student and chief scientist at Espresso Systems; Scott Kominers, a Harvard Business School professor; and Valeria Nikolaenko, a former research scientist for Novi at Meta.
While many crypto firms have hired research teams, A16z is attempting something bigger: an entire research lab to take on crypto.
The research unit aims to create a pipeline from research to industry, working with a16z’s engineering, regulatory and marketing teams.
As part of the venture firm, a16z Crypto Research will work with the firm’s portfolio companies and investors to help determine what topics to take on. The group will seek to address broad industry challenges, Yahya said.
It will aim to produce work that is open source and that the entire industry can benefit from. This approach is better than a university-based lab, because it is closely aligned with people working in the industry, Yahya said. “All we only, really want from all of the work that comes out of the research lab is for it to be open source,” he said.
Like other research labs such as Bell Labs, SRI and others, the research is designed to bring academic and scientific research into eventual commercial use. While that could happen through deployment by a16z portfolio companies, projects could also be spun out into independent companies, Yahya said.
The researchers could join a portfolio company to work on the ideas or they could potentially start companies using the ideas. “And in those cases, then we might fund them, which would be a success case,” Yahya said.
And as part of the venture firm, everyone involved in the lab will have upside in the work in a way that’s not possible in a traditional academic research lab, he said.
Topics of study could range from technologies to achieve consensus on blockchains or Layer 2 scaling solutions to application layer issues, automated market makers or NFT design, Roughgarden said.
The research will likely initially flow from team members’ interests. Kominers, who is an expert on applications, could work on issues such as the design of NFT marketplaces. The other three focus more on cryptography and security: Bonneau, who wrote a book on crypto; Bünz, who worked on zero-knowledge proofs and verifiable delay functions; and Nikolaenko, an expert on proof of stake. The team will grow by “many multiples,” Roughgarden said.
Computer science often ranges from very specific to general theories, but this new unit will be able to address real-world problems, he added.
“These are the problems that our portfolio companies are struggling with that you see over and over again across different projects,” he said. “That's kind of a telltale sign that this could be your next great academic research paper capturing the essence of where everybody is stuck.”