A version of this article first appeared in CNBC’s Inside Wealth newsletter with Robert Frank, a weekly guide to the high-net-worth investor and consumer. Sign up to receive future editions, straight to your inbox.

Family offices are built to be patient with their portfolios, investing with generational wealth in mind. Giorgos Tsetis is not. 

Through his family office, Great Things, the co-founder of Nutrafol backs high-flying startups at a quick pace and allocates at least 20% of annual net realized profits to philanthropy. Over the past 18 months, he has invested nearly $40 million and committed about $7 million to nonprofits between gifts and pledges, according to Tsetis.

He told CNBC that he hopes the model becomes a blueprint for other wealthy families to give back now rather than as an afterthought. The artificial intelligence boom has allowed him to realize profits quickly, including a seven-times return on Anthropic in 18 months through a secondary exit, he said.

“I’ve got kids, and obviously I think about the future. But what I think more about is we need to solve problems together as we speak,” said Tsetis, 41. “As innovation is creating this extraordinary amount of wealth, what we’re designing is a model to share those windfalls. And there’s restlessness with that. It’s like we need to do it now.”

He formally launched Great Things nearly a year ago after selling his remaining stake in his hair-growth supplement business to Unilever at a $3.5 billion valuation. Tsetis knew he wanted his second act to involve a substantial amount of philanthropy. 

The Great Things formula and 20% giving minimum were inspired by the economics of venture capital and private equity, said Gabriel Cooperman, Tsetis’ financial advisor and a managing director at UBS Wealth Management who helped structure the family office.

“Basically what he’s done is just turned the profit-sharing interest into a charitable-sharing interest,” Cooperman said. “We know it works. We know it’s very sustainable.”

Tsetis’ donor-advised fund serves as a buffer if investment profits in a given year don’t cover the firm’s charitable commitments. Great Things typically makes three- to five-year pledges, supporting nonprofits such as an after-school boxing academy in the Bronx and Every Cure, which repurposes existing drugs to treat rare disease.

If Great Things continues to invest at its current rate, Tsetis expects to deploy another $60 million within the next two years. The firm can move quickly without outside investors because investment decisions come down to Tsetis and one partner, Roman Kalantari.

That said, they are far less bullish on AI startups than they used to be, Tsetis said. The firm is moving with more caution, such as focusing on late-stage rounds to prioritize liquidity, he said.

The AI boom that has reaped quick returns for Great Things is bound to wane, according to Kalantari, the former chief experience and technology officer at Nutrafol who started his career during the dot-com bubble.

“Anyone who tells you there’s not going to be a slowdown or a correction of some kind has really bought into the hype machine,” he said. “When I look at these AI companies, I really try to think about who’s going to survive that correction.”

Great Things is moving away from pure AI startups and focusing on ones that have a durable value proposition and are built on their own tech rather than that of OpenAI or Anthropic, Kalantari said. Lila Sciences, which Great Things recently reinvested in, checks both boxes. The 3-year-old startup has its own AI model and builds automated robotic labs that make scientific research faster and cheaper.

Tsetis and Kalantari are still working out one part of the Great Things formula: how to balance investments in what they see as winning technologies with their broader commitment to impact. For instance, the firm’s portfolio includes Polymarket, the controversial prediction-market startup. 

“This was a conscious decision to participate and see if we can generate significant returns and do what we believe is right with those returns,” Tsetis said, “and at the same time, continue to monitor the situation and see how it evolves.”

Since Great Things is not looking to hold for the long term and Polymarket ownership is a hot commodity, the firm can exit via the secondary market relatively quickly, he added.

Further, adding a traditional impact investing lens may make it harder for the Great Things model to scale.

“We’re just trying to do what’s right for us so we can make the model sustainable and make it work in an extraordinary way for others as well,” Tsetis said.