Most undertaking firms avoided investing in rivals. But GV (Previously Google Ventures) is not the majority of firms. Not only did it come back Loose one sec Google He had a competitive product, now invests in AI firms that go on the strategy to his home company AlphabetOwn efforts in the artificial intelligence race.
GV model – a single LP, independent decisions – allowed him to stay fast and agile during the most explosive moment AI. Partner management Dave Monichiello AND Tom Hulme They support firms in systems, compilers and applications and create each early and late factories. The GV plant is clear: the age of AI rewards those that move quickly, trust the founders and do not shyen in the values in front of the eyes.
Recently, we talked to Monichiello based in the Silicon Valley and Hulme, based in London, which conducted digital investment practice around the world, about the AI GV investment approach.
Monichiello was previously in Kiva Systems, which became Amazon Robotics. Hulme has experience in physics and was a project director in Idea In Europe and the lively investor of the angel before joining the then Ventures in 2014.
Crazy AI gold fever in the West
“It’s absolutely wild time on the west coast in San Francisco. Every weekend there are five or six long walks with founders who are flooded with sheets of AI’s term of office,” said Monichiello. “The founders are filled with other AI founders in their homes. The buzzing is heavy people who are ready by Finishwedding OpenaiS world, at the same time repelling these offers and starting its own, “he said.
Monichiello and Hulme said that they each have medium and highschool children who use technology in their every day lives, write, digest and sum up complex ideas.
From the universal adoption of consumers there is drastic pressure from enterprises to spending this “magic AI” to know this “magic”.
The talent competition is heating up
“We are thinking about a pile as low as tokens and infrastructure, as well as as high as the level of application, and about various industries in which investments at the application level bloom,” said Monichiello.
Early GV investments in artificial intelligence include Latticewhich was purchased by Apple For Siri supply in 2017, Video generation company Synthesiaand data labeling company Snorkel AI. Recently, it was an investor in Thoughtful machine laboratoryA record round of $ 2 billion with a valuation of $ 12 billion.
GV invests no matter Alphabet, his parent and the only investor. This signifies that sometimes it supports firms that might be seen as competitors of products belonging to the alphabet. For example, GV invested in Slack when Google had a competitive internal product, and Lab pondering the machine competes with Gemini Alphabet.
As for why they invested, Monichiello said that that they had heard from their internal recruitment team that the general director and co -founder of the Machine Hunter Mira MuratiEarlier Cto OpenAI, is one of the best recruiters of deep technical talents of AI. “Amazing companies are sometimes at prices that seem very uncomfortable that it was certainly so for thinking machines,” he added.
The competition of research talents AI is intense and the model performance is measurable. This means “for the first time many large companies compete in the same game. So if you look at the finish line, Alphabet or OpenAi, everyone is competing for the most modern models of foundations,” said Hulme.
“We have never seen companies scale as quickly as AI family companies that we see today,” he said. “We have not seen the talents move so often in the foundation model as the amount we see at the moment.”
The team draws attention to where the smartest researchers will operate. “They don’t go to large technology. They go to startups and everyone goes to San Francisco,” said Monichiello.
Application layer
“We wondered to invest in the application layers in companies until we know that there is real adhesion,” said Monichiello. “If you look at Harvey Or OpenVideWe did not lead series A. We led the later round. “
Monichiello said that the query that goals to resolve: “Do these dollars look for use, looking for an application that may be interesting in their company? Can they be closed in production, creating real value for the company?”
To assess whether revenues are experimental or not, the team spends time understanding the use of shoppers – is the product not proof of the concept, is it used every day, or does it save hundreds of thousands of dollars in a table, while creating real value?
In 2024, the company led the C series in AI Legal Tech Startup Harvey with a valuation of $ 1.5 billion. Since then, the company has collected two rounds, and now it has a valuation of $ 5 billion.
GV recently led with Small everlastingSeries B at Openevidence, a platform for artificial intelligence, which helps doctors and other healthcare employees easier to aggregate and process reviewed medical literature with a valuation of $ 3.5 billion.
GV also raised an investment in the amount of $ 40 million from the A series with its registered office in New York Tq ventures On the London online legal platform Lawhive.
“When we look at the companies that come, the revenue ratio is crazy. These companies grow incredibly fast, faster than ever before,” said Monichiello. “And it is very difficult to spend a lot of time looking at the AI application companies, and then go back to looking at other companies.”
“We have companies in our portfolio, such as Bolt.new [a product of StackBlitz] – They built technology, and then turned on monetization and increased from zero to $ 40 million ARR in 12 weeks, “said Hulme,” We think this is a new norm. ” GV was the leader in the initial round in 2022. The company didn’t grow significantly in October 2024, when it launched Bolt.recent to code the text hints in the web browser that began.
Investment strategy at the infrastructure level
In the case of the infrastructure layer, the company’s account is different, said Monichiello. “We start to do a deep technical work and we see that there is a compiler that is universal, and the company needs a lot of money to get there, and really patient capital. We run the first round. We will probably run many rounds over time, but with time it will pay off 100 times, if it is worth a deep investment.”
For this purpose, GV led a round of $ 30 million Modularwhich goals to build a unified computing layer for interface with AI equipment. The company has created a universal compiler, which is a competitor of miracles that permits AMDGPU, which is able to work as fast as these NvidiaMonichiello said. Since then, the modular Batin Bi series in the amount of $ 100 million General catalyst.
Multi -stage, multisctory, multigeo investors
GV has been invested for 15 years and has $ 10 billion in managed assets. The company is based in Silicon Valley with offices in Cambridge, Massachusetts, New York and London. His best outputs in time include UberIN Nest laboratoriesIN Loose AND Gitlabamong others.
CEO and Managing Partner David Krane He manages a team of 21 partners. Together with Krane, Monichiello and Hulme, the fourth managing partner is Krishna Yeshwantwho is based in Cambridge and conducts the practice of natural sciences with a general partner David Schenkin.
Companies meet at least one other partner in the investment process, and then normally transfer the whole partnership. The partners share ideas and thoughts with a potential customer who can then make an investment decision. The team can move quickly and was capable of finish the contract inside a week of the first meeting with the founder.
“A and B are a sweet place,” said Monichiello. “Our work, when we find an area that we are excited, is a look at A and B., and then saying:” Is this the best strategy to invest in this category? “
If the appropriate company is in the D series or grain, the team will make this investment. Team members are not limited by location and can invest in the best firms around the world.
Hulme thinks more in cities than countries. “For every great founder you need great operators, perhaps five or 10 large operators, and the global pockets of them remain, in my opinion, in San Francisco, New York and London. These are the cities about which we are most excited about. But we also invest in places such as tel. Aviv that grow quickly.”
Analog business
“We are a Venture Capital business, but we consider it a business of human capital. We are as good or as bad as people in the GV team. And, interestingly, they are as good, bad as the people in which they invest. This is a very analog business in this way. It is simply super-sprayed technology,” Hulme said.
Monichiello said that the diverse approach and flexibility of the company are crucial.
“GV is a place where we can have 30 people in 30 different directions, investing in 30 different things. And we do not try to do a consensus,” said Monichiello.
Explanation: This story has modified from the original publication to update information about the practice of the company’s natural sciences.