Meet Silicon Valley’s AI-obsessed mayor

This episode is brought to you by State Farm. Listening to this podcast, smart move. Being financially savvy? Smart move. Another smart move? Having State Farm help you create a competitive price when you choose to bundle home and auto.

bundling just another way to save with a personal price plan like a good neighbor state farm is there prices are based on rating plans that vary by state coverage options are selected by the customer availability amount of discounts and savings and eligibility vary by state Hey, welcome back to Politico Tech. I'm Stephen Overly, and on this show, I break down tech, politics, and policy with the people shaping our digital future.

On last week's episode, I talked about the growing number of cities using AI. And this week, I'm talking to the mayor trying to outpace all of the others. Matt Mahan is the mayor of San Jose, California. the largest city in Silicon Valley. Companies like Broadcom and Adobe are based there, and Mahan himself is a former tech entrepreneur.

Now as mayor, he's pushing for AI to be adopted across city government, transforming everything from bus schedules to business permits. And he's also warning fellow Democrats in California not to overregulate. the companies in his backyard. On the show today, Mahan and I talk about his ambitions for an AI-enabled city, his stance on California's new AI safety law, and whether his recent criticism of Governor Gavin Newsom extends to tech policy. Here's our conversation.

Mayor Mahan, welcome to Politico Tech. Thank you. Great to be here. I've seen you say that you want San Jose to be the most AI-enabled city hall in the country. Why is that the goal? Well, the goal in my mind is to better serve our residents in a world in which we have constrained resources, a lot of uncertainty. We live in a very dynamic world and our residents have a lot of needs. And when I look at the use of technology and gains in productivity over the last 50 to 100 years.

Most of those gains have been had in the private sector. We haven't seen government productivity generally increase all that much. We tend to throw more money and more. people power, more headcount, more staff at issues. And some areas really require that. But I also think we should be adopting new methods, new tools, and trying to automate things where we can and make government.

more productive so we can stretch our limited tax dollars to have more impact. So that's really my guiding light when it comes to adoption of technology. How do we enhance our existing labor force and do more, deliver better outcomes for our residents?

And I know there's a lot of ways in which you're using AI today. What stands out to you as sort of the shining example of how San Jose is using AI and revolutionizing local government? Well, you're right that we're running a lot of different AI pilots, one that is not just in production in the wild, but has now gotten to citywide scale. is our signal synchronization for bus routes. This was a pilot we ran last year. We tested

two of our rapid bus routes. And the way that the tool works, we partnered with a local AI company called Light AI. It connects the buses to the red light signals and dynamically measures the velocity of the bus and traffic conditions and then is dynamically changing.

the duration of red lights to reduce the probability that a bus hits a red light. What we found was this dynamic system could reduce the likelihood of a bus hitting a red light by a little over 50% and speed up overall bus routes by over 20%. which is pretty neat. Everybody's commutes just got shorter and they didn't even realize it. And we didn't really have to impact cross traffic at all either. Got it. And no one's going to complain about a shorter commute and less traffic congestion.

The question I have, obviously, then is, you know, that's an example of the wins. Everyone loves to talk about the wins. What are the challenges? You know, you've done these pilots, you know, you've probably learned more than most about. the challenges of implementing AI, what have you found there? Yeah, you know, these tools take a lot of workforce education and training, and they're not always...

they haven't necessarily been applied to specific use cases in a holistic way. So we found there are some off-the-shelf tools like light synchronization. Or language translation is another good example, where the general use case is directly applicable to meaningful use cases at the city. And we've been able to deploy those. You take another case that's more specialized, like permitting or procurement, where we have a lot of rules, a lot of edge cases, a lot of...

different levels of government involved, state regulations, waterboard regulation, city level. And while I firmly believe that AI should dramatically speed up permitting, to use that as an example. There is no off-the-shelf AI tool that can just be directly applied to our permitting processes. We issue 25,000-plus permits a year, hundreds of different permit types.

Lots of nuance. And so I'm eager to get to that point where it's adaptable enough and dynamic enough to kind of know the right answer or know the right checklist or create the right process around each application. to take more of the discretion out of it. I really think these permits for most buildings or most investments. shouldn't be one-off bespoke situations. And yet out in the built environment, there's a lot of variation. And so that's an area where AI is just going to take longer to.

be fully adopted and integrated in a way where we see real efficiency gains. And so it's not necessarily a failure so much as it's not easy to adopt. And I think it's going to take a lot of time and work and more effort than people realize.

Well, I was going to ask because you're in sort of a unique position, obviously not being only the mayor of a major city, but kind of the mayor of Silicon Valley in many ways, right? A lot of the companies developing this technology are... if not based in your jurisdiction, certainly right in your backyard. And one of the questions I guess I have is, what do you think, being so close to these companies, that other government leaders are missing about?

them and missing about AI? I think part of what is missed often, and I see this even just in Sacramento, which isn't that far away, is a an impulse to want to regulate and guard against the downside. And there's a good rationale there. I mean, clearly the public sector has to protect.

public safety and privacy. And we do have a responsibility to think about those worst case scenarios and guard against them to protect the public. So I don't mean to be dismissive of that, but you hear a lot of the policymakers. come from a regulatory and a risk mitigation perspective. Many are lawyers themselves. And think first and foremost about how do we constrain the tool.

Whereas being in Silicon Valley, you kind of pick up on some of the optimism of, hey, we have all these new capabilities. Why don't we go ask the question of what could we make better? How could we apply this to make something more efficient or more responsive or more personalized? How can we use these new capabilities?

in ways that benefit our community while acknowledging that it'll be messy and we'll make mistakes and there'll be an evolution to it. But there's that sense of... optimism and a willingness to experiment and an acknowledgement that failure, if done right, is actually a method, a means of learning.

and that that's ultimately going to lead us to a better future. And that's something that's palpable in Silicon Valley that I think is often not as obvious in other parts of the country. It's an interesting point, and I do want to come back. to regulation. It raises the question for me, though, I mean, with the Silicon Valley optimism, do you have any reservations or concerns about

cities like San Jose being dependent on tech companies and on their AI models for city services. I mean, this technology is very powerful. We also know it has some flaws that need to be worked out, too. Yeah, we haven't seen a lot of vulnerability in that regard yet. In fact, with many of the pilots we're running, there are multiple potential vendors. I'll give you another example of one that we're really excited about.

owned or contracted vehicles that cover every mile of every road every week in the city, we ought to have sensors, essentially cameras, that can better monitor what's happening out on the streetscape. And then when relevant, generate a service request and a response. Have a team go out and fill the pothole or fix the streetlight or pick up the pile of trash or cover the graffiti.

And thus far, we've found there are many companies working on these solutions. Now, there's always a risk of eventually, if it's a winner-take-all market and there's truly only one company that offers the service, that's risky. One thing we have always looked at in our procurement is ensuring that we own the data and we keep control over things like our data retention policy, like having the data deleted on a regular basis. We don't want to.

allow data that's collected in the city on our public space using public dollars to become controlled by a private entity or shared with third parties. So there are things like that that we bake into our procurement process. But thus far, to be honest, there's kind of this wide world of opportunity and experimentation, and we don't quite know where it's going to go. We're really still in the early stages of using these new capabilities to try to improve city services.

It's interesting to hear that you sort of build some of those safety measures into your procurements, especially when it comes to things like data privacy. I don't have to tell you that Governor Gavin Newsom just recently signed this first-of-its-kind AI safety law there in California that effectively requires frontier AI developers to report major safety concerns or issues with their platforms.

Now, this is something that some in the tick industry were opposed to. Do you think this is the right move for California? You know, I will say this bill, SB 53, is... much better and more balanced than the version we had last year that the governor vetoed and I opposed, as did many. It's better. I will say I always am fearful. of this impulse to overregulate and kind of preemptively prevent all possible risk and downside. I think there is this reflex in government, particularly in California.

to over-regulate. And we have this abundance movement now growing that I think is trying to offer a counterpoint and saying sometimes we overly constrain. the state's capacity to do things. And in the long run, we all are less well off. And maybe we prevented that little bit of downside or that edge case.

But in the long run, we've actually held back progress for society as a whole. And these are tough calls. I mean, these are tradeoff issues. They play out over many years, if not decades. It's hard to know what the optimal. level of regulation is. This particular bill I'm more comfortable with because it's limited to the largest companies, 500 million a year in revenue or more. The requirements around reporting and whistleblower protections and incident response reporting specifically.

are pretty reasonable in my view. And I do think we need government to be in a position of understanding what's actually happening in the sector. I'm a little allergic to just this impulse to just add a bunch of AI regulations because it's a new thing that some folks are scared of. I think it's a pretty well-balanced piece of legislation, and I'm hopeful that it won't have huge...

And I'm also hopeful that if the industry is telling us, hey, there are elements of this that are creating these unintended consequences, that the legislature will take that seriously and, if necessary, make tweaks to the regulatory framework over time. This episode is brought to you by White Claw Surge. Nice choice hitting up this podcast.

No surprises, you're all about diving into tastes everyone in the room can enjoy, just like White Claw Surge. It's for celebrating those moments when connections have been made and the night's just begun. With bold flavors and 8% alcohol by volume, unleash the night. Unleash White Claw Surge. Please drink responsibly. Hard Seltzer with flavors. 8% alcohol by volume. White Claw Seltzer Works. Chicago, Illinois.

From other Democrats I talked to, it seems like that position would sort of put you at odds with some in your party who feel like tech is not regulated enough, you know, that these companies are so big, so powerful, that there do need to be more guardrails on them. Is there tech regulation that you think would be good or accountability for the industry that you think would be good? Or is all of it harmful to innovation, if you will? Well, I think there are.

regulatory frameworks related to privacy that make a lot of sense. Obviously, the European version of it is stronger than in the US, but I think we've seen a strengthening of consumer protections related to people's personal data on the internet. And I think that's a good thing. But I do think that in California, we have seen a failure of regulatory excess in many sectors without recognizing the risk of pushing innovation out of the state.

or even out of the country to places that will have potentially significantly lower standards. And so it's always just a question of what is the right balance. To use a different sector, I think many of us would agree at this point that in hindsight, California's incredible level of regulatory burden process and trying to, you know, assess a.

a fee to cover every externality of building housing has led to a world in which we have a dearth of workforce housing. We have this huge housing shortage in California that was largely... created by overzealous regulation and process. So it's just a cautionary tale. It's not to say that we shouldn't regulate the tech sector, but we should also recognize that these are the companies that are creating.

New innovations, new tools, new approaches, new capabilities that offer the greatest promise to enhancing our lives, whether it's in robots that can perform surgery better than any human, or it's autonomous. and electric vehicles that provide cleaner air and have fewer crashes. I mean, the promise of improvement largely comes from these innovative companies. And they're often generating the tax base and economic growth that creates

opportunities for employment and in the tax revenue that we use to fund services. So I just think California has to be careful not to kill the golden goose, as the proverb goes, and make sure we're also working with and listening. to those in the industry and not just assuming that because a company is big, it's therefore bad.

You know, that's a really interesting point. And I want to ask because I know that you and Governor Newsom don't always see eye to eye. And I think my political colleagues have dubbed you one of his home state haters, if I'm quoting that correctly. When it comes to AI, when it comes to tech, I mean, are there things you'd like to see Hamby do differently from Sacramento?

Well, I do want to just set the record straight that I'm not a hater. I have praised the governor many times. I would say he and I, I understand the headline is good for getting interesting clicks and all. I understand how the internet works, but I've praised the governor. Many times for I think our disagreements are really not not not political. They're certainly not personal. I think it's just.

Often on the substance of specific issues, I have been willing to publicly disagree. And I think that's healthy. I've argued that in California, we've gotten complacent. We've had essentially. one-party rule, and I happen to be a member of that party, the Democratic Party, but I serve in a nonpartisan office. I'm a local elected official leading one of the state's largest cities. I try to call it like I see it, and I try to be outcome-focused. And so on public safety,

safety, addiction, mental health issues. I have been critical of certain policies or lack of policy action from the governor on AI and this new economy. I actually think the governor's largely gotten it right. An area where I would like to see him do more, and it isn't just him, it's Sacramento, it's all of us as leaders in the state, is to recognize... the immense amount of capital flowing into this market right now and how much of it is choosing not to flow into California.

because we either don't have the energy capacity, the grid isn't sufficiently upgraded, or we have land use restrictions and other regulatory requirements that just make it so expensive and so slow. to deploy capital in California that we're at risk of seeing much of this economy, this AI powered economy, just relocate to. places like Austin and Northern Virginia. And I do worry about that. I'm not sure we have been as laser focused or as aggressive in addressing.

the barriers to deploying investment in the grid, in new energy production, and the actual construction of R&D labs, manufacturing data centers. And in San Jose, we're lucky. We have two- gigawatts of power coming online over the next four years more than any other city in the state. But we're pretty unique in that. And the demand in this space is in the tens, if not.

100 gigawatt scale at this point nationally. And we just don't want to see 95% of that go to the rest of the country when we should be competing for it right here. The race is on already, as you know, for the next governor of California. I'm curious when you look at...

the candidates who are out there in kind of the landscape, is there anyone who you think gets that, who you think is the right candidate for Silicon Valley for this new tech economy? You know, I'll say I haven't seen it yet. But it's still early. I'm listening. I'm aware of who's jumped in. I'm paying attention, curious if anyone else will jump in. I'm trying to listen to all the candidates. They, of course, refine their...

The way they describe the issues, the issues they're focused on, they hone that in over time as they're out getting feedback. from voters across the state. So I expect there to be some evolution in how they're thinking and talking about issues. I'm excited to invite all of them to come through San Jose. We think we have a story to tell.

And I hope a positive example that the gubernatorial candidates can look to and say, maybe the rest of the state can do more of this doesn't mean we're perfect. I would say an area where we've lagged in full transparency is home building. It is too slow and too expensive.

to build housing in all of the Bay Area. And San Jose is certainly not an exception yet, though we're working on using AI and permitting. We're reducing fees. We're moving to more by right or ministerial approvals. We're trying to take seriously. are policy failures on that front. So it's an evolving conversation. I'm interested to see as the field kind of fills out if any others join in the next couple of months.

And I'll be really pushing each of the candidates and listening for them to articulate where the state's current approach has failed to deliver the outcomes we need. I see a lot of those examples. unsheltered homelessness, untreated addiction and mental illness, high cost of living.

You know, us losing investment to other states and data centers manufacturing R&D labs. I don't think we're getting the outcomes we need in these areas. And I think in many cases, there are direct policy decisions, particularly at the state level.

level that have been made that are making it harder for us to deliver those outcomes. And so I hope somebody's going to come in and say, the status quo really isn't working. Here are the five to 10 things I'd do differently to try to deliver better outcomes in those. critical areas. So that's the kind of conversation I'm hoping we can have. Got it. I'm obligated as a political reporter to ask if you might be someone who comes into the race. You sound like someone who might be considering it.

You know, I'm really trying to run San Jose. I appreciate the question. I have gotten a lot of feedback or a lot of outreach over the last six months or so. I think people are seeing. that our back to basics approach in San Jose is working on homelessness, public safety, removing barriers to home buildings, spurring economic development. But I want to see through what we're doing here. I didn't run for mayor of San Jose to be.

stepping stone to something else. My wife and I are raising two little kids here. I'm just two and a half years into this role. And because of a realignment in the timing of the mayor's race, have the potential, it's up to the voters, but the potential to serve for 10 years in the role. And I'd like to see through what we're doing here. So I'm not...

I'm not planning on jumping into this race, but I'm watching, I'm listening to the candidates. I'm hoping that we're going to have a strong change agent emerge from the field with a clear vision. for doing things differently to deliver better outcomes. And I'll be excited to support that person if and when they come along. And in the meantime, I'm going to focus on running San Jose and making it a great...

As good of a model as it can possibly be for the rest of the state by delivering a higher quality of life and greater opportunity for San Jose residents. With AI, it sounds like getting you part of the way there. Mayor Mahan, thanks for being here on Politico Tech. Thanks for having me. I really appreciate it. Digital Future Daily, and Morning Tech. Our producer is Normal Molechel. Pran Bandy made our theme music. I'm Stephen Overly. See you back here next week.