Bill Tai's approaching his 30th year in venture capital — but his bets aren't getting any less bold. In fact, his latest involves hiring a large chunk of Cambridge Analytica's staff.

"I'm taking the core data science team out of Cambridge Analytica," he told Protocol. "I've started a new company with them." Tai didn't offer any further details on the plans, other than saying "it has huge potential."

Get daily insights from the Protocol team in your inbox

But other comments made in Protocol's conversation with Tai — about how doing good is now nonnegotiable and data science is causing radical shifts in how commerce and supply chains around the world operate — hint at where his head is currently at.

Tai was speaking ahead of the finals of the Extreme Tech Challenge, a pitch contest for startups doing work that aligns with the U.N.'s sustainable development goals. Tai co-founded XTC as a general startup contest before the organization pivoted to more explicitly focus on social good. In a wide-ranging conversation, Tai discussed the competition, the tension between Silicon Valley and government, and the new world of data science companies.

This interview has been edited for clarity and length.

What prompted XTC to move from a typical startup competition to one focused on social good?

The winners of the past contests had evolved from what were initially mobile app companies, to consumer-facing things, to the winners almost always being judged at the end on whether they did something good for this world. And it was just a natural thing, I think — I think the kinds of judges that we had for the contest and the movement that you feel today, obviously it's one where the next generation really cares about what's going to happen to society and this planet.

The final swing on whether we made a company a winner or not, almost always involved this deep conversation about which one will have the greatest positive impact on this world. And so about a year and a half ago, I got involved with the United Nations on some other things, and I realized what a powerful framework the Sustainable Development Goals could be. And we decided to change what was a fun commercial contest with good winners for social good, to officially something built around finding great solutions for this planet.

Is there any tension between doing social good and the profit motive?

I'd say that there's been a sea change in the general answer to your question. I've been in venture capital now since 1991, so it's coming up on 30 years. And I'd say for the first 25, if you were a company focused on sustainability, you could never raise money. And, generally speaking, you were competing with some commercial solutions that would kick your ass, because they didn't have to care, and they didn't have the overhead expense structure.

But I think there's a confluence of several major social trends and technology trends that are making sustainable companies better than commercial companies that are not focused on sustainability, in many ways. Whether it's raw technology coming down a cost curve, like solar … to digital technologies that unlock a lot of value in things like fintech … to the very pressing need for every brand now to stand for something good to be viable. All of those trends together are creating a real energy shift in society, and waves of commerce, that make the sustainable companies even better investments.

What does the market cap of a Tesla or a Beyond Meat tell you? It tells you that the capital markets are now willing to very much reward brands who stand for something good. You look at the commercial success of a Patagonia. I think that millennials and younger, they will not buy your product if you don't stand for something good. And if you stand for something bad, they will avoid you like the plague.

Is there a risk that consumer sentiment shifts and people stop caring about this stuff?

I don't think so. I believe that people are starting to care now because they're realizing that it really matters. And if they don't care about it, that it could very well mean the end of the planet or society as we know it, given temperature rising and pandemics and things like that, that are of our own doing.

Is there any area that companies in this year's competition seem focused on?

Coincident with the sustainability focus, I'd say there are natural waves in commerce that are very big, where there are giant markets that are being disrupted anyway, whether or not they have a sustainable bent, and the companies with some sustainability stand out a little bit more.

I'd say the fintech category, for one, is being completely remade. Banks today look a little bit like the telecom operators of the late '80s: very high-cost structures, and an international wire transfer looks like a fax now. That whole sector is going to get remade.

There's a lot more material science occurring. There's a little bit more raw chemistry. Regular venture capital became very much about first atoms, silicon in boxes and in networks, and then bits. And now it's molecules.

The change in bits is affecting the digitization of many industries like finance, but there's also a new emphasis on the remaking of molecules for lower pollution, and cleaner air, and better foods, and the monitoring/measurement of health.

The traditional tech VC model has been driven by making something with almost no marginal cost so it can scale and produce big returns. How does that work when you're dealing with physical materials?

It's actually a return to basics. The digital stuff's not going away, but if you look at the kinds of companies that existed in venture from the late '70s through the early 2000s, the replicability and scalability involved hardware. It's only recently that we have these explosive digital giants where you're basically just moving bits around. It's only been a 10-, 15-year period. I'd say for all other times in human history, we've had to build companies out of physical things. And so I'd say it's nothing new. The blip is sort of over, or not over — the blip existed, now we're going back to basics.

What are the kinds of companies you would like to see next year?

There's only three things you need to do in a digital company: You need to lower the friction to usage, you need to make it replicable, and then make the infrastructure scalable. If you do those three things, it always works.

You used to do that with physics: making moving electrons easy, so you turn them into little silver slivers, and you made factories that could produce billions of them. So, lower friction, replicate, scale. The first wave was silicon. The second wave was taking those Legos, and building boxes — computers and communications gear. That produced Cisco and many other things. And then you took those and you stitched them together, and you remade telephone networks, so you replaced the phone companies. So: Legos, boxes, networks.

And then the fourth wave was putting a user interface on it. So how do you move the ones and zeros off the page. So that was Facebook, Google, LinkedIn, Twitter, Zynga — they're all user interfaces, all they do is move ones and zeros on a page.

The fifth wave was the data science that allowed you to know what was moving where, so you could tune and win — because the ones that won in the user interface world were the first movers in really doing A/B testing and using data science. So now you've got this wave of companies that people largely in the consumer world don't understand so well. But whether it's Splunk, or what Treasure Data was, or Snowflake coming up, it's kind of this world of data science companies.

With that, now you have a whole new industry. So if you think about what is an Uber, what is a Walmart? They're basically data science clouds with things on the end of them. I think we're in a world now where that kind of efficiency that was used to route atoms on silicon is now moving assets and value. Some call it blockchain — blockchain's one derivative of it — but you're seeing a remaking of the recognition of the endpoint. Airbnb is a data science cloud with rooms on it.

Whole industries are turning over. You look at Walmart versus JCPenney and Sears: The old way they ran was, you'd send a buyer out to Asia, buy billions of dollars of stuff, stick it in a warehouse, and then hope it sold at Christmas. And then if it didn't you were out of business, and if it did, you did OK. Walmart, because they measure everything that goes through those terminals, they're a liquid inventory mesh that moves where it needs to go, when it needs to go there. They don't have the working capital need.

That's what's happening to the car industry with Uber, that's what's happening to the hotel industry with Airbnb, and that's going to happen to every supply chain in the world over time.

I think you're going to see what was sort of the network effect of connection of nodes on Facebook, abstract itself to companies and the interchange of commerce and the exchange of value, whether it's physical products or dollars.

I want to talk about the relationship between tech and government. Some people take the view that government should be hands-off and that we'll innovate our way out of problems, others think there's a big role for regulation to play. Where do you come down on that?

I'd say that that's a foundational change as well. Silicon Valley really started by ignoring government, in a way. There was a little bit of of help from Defense Department-related spending in some of the advanced technology a long time ago, but I'd say the '70s, '80s, early '90s were really dominated by engineers that didn't care about government, just building things that worked a little bit better, a little bit faster, and selling to each other.

As products became more and more consumer-focused, as opposed to B2B technology plays, they touched more individuals, and therefore introduced a different set of criteria for success and risk factors. You're starting to see that now, even in the data world with the heat coming down on Facebook and Google.

As we move more and more to a world where the exhaust of a business can have a lot of negative ramifications on people and planet, regulation is required — or, the facilitation and support of government — to help bring solutions that are better for people and planet to the forefront. Which is why, in part, we're working with the U.N. Framework on the Sustainable Development Goals. It's to help leverage that entity to reach a broader audience and gain support, not just in the United States but globally.

Update: This article was updated to clarify Bill Tai's remarks on how venture capital evolved over time. Updated July 16.