The Policy Implications of the AI Systems Era

Intro:

Vince: [00:00:00] What most of us think about artificial intelligence today is maybe like a coding assistant or a chat bot. The next kinda aha moment in AI is when you no longer need to send something off to the cloud, but local AI processing is being done on these devices that you just couldn’t imagine that are in the palm of your hand and you’re here and you’re driving around and that sort of thing.

Conversation:

Daniel: What’s up friends? This is the Next Byte podcast where one gentleman and one scholar explore the secret sauce behind Cool Tech and make it easy to understand. Thank you Vince for joining us on the pod. We’re really excited to have you. First, first order of business. Would you mind giving us a brief intro on who you are, what you do, and who Arm is?

Vince: Yeah. Thanks. Thanks guys. Thanks for having me on. My name’s Vince Jesaitis. I’m head of Global Government Affairs for Arm. Arm is a, I would say they’re in a semiconductor space. They’re not a traditional semiconductor company. They provide building blocks that semiconductor companies use to produce complete chips, so central processing units, graphics processing units peripheral technology that companies will then incorporate into a semiconductor chip and have that physically produced TSMC or Intel Foundry Services or GlobalFoundries or any manufacturing facility.

It’s an absolutely prolific technology that Arm provides. I’m guessing probably anybody listening to this is touching Arm technology in some way, either in their earbuds or in their phone that they’re, streaming us through or their laptop if they’re on a, personal computer.

I’m not certain if this is on video or not, but behind me is a poster that says [00:01:00] Arm customers have produced three 25 billion chips since the company was founded. Just last year our customers produced about 30 billion chips. So to give you a perspective on the scale of how many chips that is, it’s about 900 per second being produced somewhere in the world with Arm technology.

So it’s literally everywhere from earbuds, supercomputers, and everything in between. I’ve been with Arm for about eight years. Before that, I worked at a technology trade association here in Washington, DC. And prior to that, I’d spent about 10, 10 years in government working for a member of Congress largely on telecommunications policies, so things like spectrum management and kinda other regulatory issues around telecommunications.

A lot of what I do is educate policy makers on what the company does and not just what the company does, where technology is going. The technology that we’re working in today won’t show up in your hand or in your ear or on supercomputer for four to five years. So we have a lot of visibility into the future direction of technology and, policymakers are really interested in that. They’re obviously trying to understand all this technology. AI is obviously quite a buzzword right now amongst policymakers, but they also wanna know what’s coming next and what they need to be thinking about. Yeah,

Daniel: that’s a little bit on me, a little bit on Arm.

On that note, I think it’s an awesome point to just start and get people really excited and maybe a little bit scared, but I think there’s probably a little bit of both to be like, there’s stuff to be excited about, there’s stuff to be hesitant about. What is it that you guys are working on now that you’ve got visibility to that you think four or five years down the pipeline, how does that translate to real world impacts for folks who are listening to the pod right now?

Vince: Yeah I think what everybody in semi different technology space is thinking about is artificial intelligence. And I think what most of us think about [00:02:00] artificial intelligence today is maybe like a coding assistant or a chat bot or something where you go to a desktop or maybe your phone and you’re interacting with with a website that’s logging ChatGPT or Anthropic or, Microsoft 365 or something like that.

The technology that we’re working on that I think is really exciting is bringing artificial intelligence out to edge devices. Obviously those are constrained environments. They’re limited in the power that they have because a lot of ’em are running on batteries.

They’re limited in the computing capabilities just because more compute sets up more batteries, so they have to be more efficient in how they do processing. We’re working with a lot of companies to actually both can scale down models from their side, increase compute and efficiency on our side.

You can do a lot of these things, whether it’s, translation on your earbuds or have more capability in your ADAS system in your car that can, help keep you safer while you’re driving can maybe help with traffic congestion.

I know you guys talked about on a previous podcast as well. Yeah. I think that’s, what’s gonna be like, the next kinda aha moment in AI is no longer, need to send something off to the cloud. But local AI processing is being done on these like devices that you just couldn’t imagine that are in the palm of your hands and your ear, you’re driving around in that sort of thing.

I gotta ask, this is an interesting shift of focus in terms of compute, at least for me because it feels every week I’m hearing about these massive multi-million dollar contracts to open new data centers and, what new graphics are and video has made to pack more systems in there, more compute in there, and the water crisis support these data centers and the power crisis.

So what is it that’s making Arm be like, actually we’re not gonna focus on data centers, but we’re shift our focus to the

Vince: [00:03:00] edge? Yeah, so I would say it’s it’s not necessarily shift. It’s more about all of the above. There’s still gonna be a lot of AI that happens in the cloud. It’s just, training these massive models, frontier models like needs that amount of compute.

You’re not gonna be able to do that on the edge. I’ve read some reports over the last couple weeks that 50 to 60% of AI computing is actually on inference. Now, so what that means is you’re, you’re using artificial intelligence and you’re using a model that’s already trained.

The more you can distribute that, the more efficiency you’re gonna get. That’s largely taking place in the cloud right now. That’s probably not the most efficient place to do a lot of this. And so if you can move that outta the data center onto local devices, you’re gonna cut down the computational usage in the data centers, create more efficiency.

So the more distributed this can be. Hopefully the more we can scale off that hockey stick curve of resource demands so I, I don’t think it’s gonna be either or. I think we’re gonna continue to see investment in both for the coming years. But it’s one way that that, we may be able to partially address some of the energy demands and water demands that’s being used currently.

Daniel: And then it makes it more, more useful and more exciting that you don’t need a Wi‑Fi connection to the cloud. Get the answer you need it. That there’s interesting economies of scale there where because you’re not sending data back and forth, potentially more efficient, but then also feels like a faster, more exciting user experience as well.

That’s right. That’s right.

Vince: And you’re already starting to see this, a number handsets can now do on device translation. So they don’t need to send, text to the cloud to have it translated into another language. There’s other things. I use artificial intelligence all the time to run through my schedule.

What do I need to be thinking about for the day coming up? How, what do I need to do to prepare for the Next Byte [00:04:00] podcast? But all that information is on your phone. At some point, you’re gonna be able to do just so much more on devices locally and not need to have a connection or, be able to do more efficiently

Farbod: on your laptop as opposed to sending everything to the cloud.

One of the things I was gonna mention is by offloading from the cloud and doing everything locally you probably also get some security benefits and, given your position of interfacing with the government, how much of that is driven by, their needs and or know what kind of impact do you think it’s gonna have on the security find adoption?

Because now you’re feeling better about not offloading yourself to some cloud somewhere.

Vince: Yeah, absolutely. There’s huge privacy benefits, there’s huge security benefits. Obviously as more capability is embedded in the devices that we’re carrying around. With this day to day, the more data’s gonna be collecting on kinda what we’re all doing.

I just as a consumer would prefer that to be on something that I’m holding physically or on my body physically, or maybe I’m driving around in or whatever. The. On the privacy side not sending that to the cloud gets around a lot of the issues that we’ve dealt with historically in privacy, whether it’s in social media or online marketing that sort of thing.

Targeted ad delivery. On the security side obviously there’s bad actors looking for every avenue to gain, information either from individuals, from organizations and, the cloud is a centralized repository for a lot of organizations and individuals these days.

And so the more that can be held outta the cloud, closer to the user the greater security benefits as well.

Daniel: Talking a little bit more about the relationship that you’ve built with government officials and the relationship that Arm has with the government. I saw one example, I think you posted it a couple months ago around creating an [00:05:00] education coalition to try and help advance AI literacy for youth.

Can you kinda highlight maybe that program are the ones that are interesting that people may not know about? Kinda understanding how, I think it’s an interesting paradigm you don’t get to talk about a lot in this podcast is how technology leaders have to interact with and play with policy leaders to make sure that everyone’s on the same page.

We’d love to hear more kind about the work that you’re doing in that space. And again, I keep saying it, but like the, so what for people, sitting home listening to this, how does, how do those interactions impact the way that they might interact with technology or government home? Yeah, no, thanks.

That’s a great question.

Vince: I would say just taking a step back, the role of governments is one, to create an environment where their citizens can succeed can make a living and also create economic value for the citizenry. And if you there’s kinda two aspects to the educational side.

One is one is what is the direct and fit to Arm of semiconductor industry. Obviously I think you guys are probably familiar, I dunno how much familiarity your listeners will have with governments globally focusing on increasing semiconductor manufacturing and design activity. There’s a number of grants and tax incentives that have been implemented across the globe to try to attract more manufacturing and semiconductor design within country borders. Here in the us there’s two or three leading Edge fabs, which is low single digit nanometer chip production facilities that are coming online or have come online within the last 12 months.

A major bottleneck to the US government attracting kinda more of that sort of investment. And those sorts of facilities here in the United States has been a workforce. And it takes a very skilled workforce to one design [00:06:00] semiconductors and to run these massive complex machines and these massive, complex facilities.

And we just as a nation, have not invested enough in ensuring that the pipelines into those fields continue. Whether it’s manufacturing, whether it’s design there’s a real shortage in workers who are trained in these skill sets to, feed into the workforce and, help Arm, help tsmc, help Intel, help Nvidia design these leading edge chips and then physically produce ’em.

So on one side of the coin is kinda our own needs. So we work with governments to ensure that they are working with the schools that they either fund or oversee through the form of education so that, kids are getting interested in these fields, have the appropriate skills as they go through the educational system.

On the other side of that coin is, I think we, we all recognize artificial intelligence is gonna change the way people work. It’s just gonna it’s obviously gonna provide huge productivity benefits. We’re seeing kind of massive layoffs announced week after week.

And, some of ’em may be AI related, some of ’em may be efficiency related decisions by companies. But it’s clear that this technology is gonna have some impact. And if students aren’t aware of how to work alongside this technology or with this technology, they’re, they’re gonna a difficult time they reach the workplace.

So it was a bit of a long-winded answer, but to the program you specifically mentioned coming outta White House it was focused on kind of both sides of that coin. We were focused more on. Training, the technical workforce is where our commitments were around. The US government can provide massive scale to that, and so more than any one company could do.

In partnership with them, we also were working with Purdue and a couple other [00:07:00] nonprofits that can help distribute educational materials at scale across the United States. And then some company’s commitments into that. A, that White House AI pledge was on the other side of that, which is like, how do we ensure that students are gonna be prepared to, to work alongside or with AI, even if they’re not in traditionally

Farbod: technical roles.

So that’s great to hear. I remember I was listening to one of your interviews it was about the internet of things. And you were talking about how in the early days of IoT there was like this big security and how that started to change and that government had the ability to accelerate some of those changes, but they hadn’t done enough.

I was wondering if are, if you’re seeing any parallels with AI technology, either on the security front or anything else, or if you, there’s been lessons learned that are now applying to emerging technologies. Yeah, no, that’s a good question as well. I would say

Vince: on kinda IoT security I think, I remember I was doing several interviews on that a couple years ago.

We have seen a lot of progress in, in that area, both in the US and UK and Europe elsewhere across Southeast Asia. So there, there have been regulations or at least best practices, but out there by governments to guide industry. And you’re seeing a lot of progress through Project Matter and others who have focused on both connectivity, but the security side of technology as well.

With AI it’s really kinda a mixed bag. You’re seeing different things out of Washington, DC you’re seeing different things out of like state governments. You’re seeing different things outta local governments to some degree, particularly when it comes to deployment of AI infrastructure in these massive data centers.

And then obviously Europe, the uk, China, the Middle East, they’re all doing their own thing. I, I would [00:08:00] say from US federal government perspective, I think that, if I were to summarize it into a sentence, I would say the biggest priority right now is technology acceleration.

I think the federal government does not want to lose this race, whatever that means. It does not wanna be dependent on technology from other nations to, continue to run AI and develop AI. I would say that’s their first order of business right now is to ensure that the private sector can innovate at pace.

So a lot of those safety considerations are frankly falling on companies. And you’ve seen different companies kinda take different steps. That’s not to say there’s a complete vacuum. The National Institute of Standards of Technology, which is a real technical body within the US government does have an AI safety Institute.

And they’re certainly doing some work. They’re engaging with industry. Right now I think the mindset of most of the key policy makers in the US federal government is we need to get this technology out there. And, the productivity benefits, the economic benefits, far outweigh the potential safety applications that are coming along.

I’ll say like the EU is looking at this kinda the other way. They released the European AI Act, which actually has a lot of more safety protocols that companies need to go through before they deploy technology in European states. At the same time I’ve had conversations with a number of European officials who do also understand the need to deploy this technology

Farbod: at pace and ensure that innovate at pace.

So what does that mean for a global company like Arm, right? Like you have different governments with different priorities. How does that shape that, the kind of product roadmap that you guys have? If at all. Yeah, it’s, it, it certainly can be difficult.

Vince: [00:09:00] Obviously there’s efficiency in developing one product and delivering globally, whether it’s Arm or anybody else.

And. We what it takes, like why partner has people like me and why, virtually every company has people like me is because you try to push these different approaches as close together as possible, knowing you’re never gonna get perfect alignment. And sometimes what it results in is developing a product that, meets the kind most stringent requirements and then that just becomes your role product.

A good example of how these things proliferate out, like across borders is, that you use GDPR, global Data Protection Regime, right? It was I think it’s probably eight or 10 years old now, but it was the first kinda comprehensive privacy law that, really restricted what companies could do with data, how you had to protect data that you held.

There was really kinda nothing like that anywhere in the world. And what you’ve seen is that companies that wanted to operate in Europe had to meet that, and it’s just kinda become the defacto way that they operate across the globe because it’s the most rigorous of any of the data protection machines that are out there.

So it’s, it certainly it certainly complicates things from a business perspective. It from a consumer protection perspective, particularly with the GDPR, there, there can be benefits there and obviously companies have found a way to make it work and still be profitable.

So

Daniel: I’ve got an interesting question for you that I think I’m gonna ask it two ways and ask for two answers. One of those is, you kinda sit in this interesting seat between giant force in technology. Like I’m reading the banner behind you and say, your chip are 70 plus percent of the world population, and so you’ve got the giant force in technology.

What’s your wishlist for governments to make it easier for you guys to do the innovation you need to do? And on the flip side of that is, internally [00:10:00] speaking, what’s your wishlist to the technologists on how they can make it more palatable for governments to kinda accept the work that you’re doing?

Is there, do you have a, maybe a sense or two on each side how folks can adapt the way they work better to collaborate, to drive? Yeah, absolutely.

Vince: There’s no silver bullet from a government action perspective. I think, I worked in government and it’s an incredibly difficult job.

Unless you’re a specialist at an agency like nist, which I mentioned earlier, you have to be a generalist. You have to go from a meeting where you’re talking about healthcare to meeting, where you’re talking about tariffs to a meeting where you’re talking about defense budget to meeting, where you’re talking about artificial intelligence, right?

And being able to be an expert in all those areas is just virtually impossible. You know what? What I would love to see, and we’ve talked to other companies and government officials about this is like ways to bring in technologists, even if, just for a short period of time. So you can have some level of expertise.

I’ll say there was a great example of that in, in the US government recently. The CHIPS program, which I mentioned earlier, with the past maybe four years ago, I think and it was a massive sum of money. It was around $40 billion that was gonna go specifically to subsidize reinvestment leading into manufacturing capacity in the United States.

When that passed you, you literally saw, dozens if not hundreds of engineers, dozens if not hundreds of people from Wall Street come in. So you had this real brain trust that understood the technology, understood the financial side, and they were the ones like working out these deals with TSMC and Intel found sources and otherwise that resulted in a really strong lasting program that, brought three leading edge manufacturing facilities back to the United States and probably wouldn’t have happened otherwise.

From the government side, [00:11:00] having that level of expertise with a, in their email is just, I think, invaluable. So the more ways that the government can find to do that, the better from the company perspective. It, it would be great if engineers were more willing to impart their wisdom to government officials.

I think a lot of times, and, I get it. These guys are doing incredibly complex work. They’re just trying to make, nice CPU work, let alone try to like, explain that to a government official. But certainly the space we’re in semiconductors, the technology space is becoming more and more of the focus of governance across the globe.

And, our engineers spending a little more time kinda educating policy makers. Again, this is what I try to do day to day, but they can go a little deeper in explaining, where technology’s gonna go and kinda what the policy considerations are. You know what the positives and negatives of any policy

Daniel: action are gonna be on the development of technology would be a very valuable thing.

Sounds like we need more crosspollination, but not just ideas, but also people, just to sit in the right offices and help people understand what’s going on and, yep. Exactly.

Farbod: I was gonna ask a question, kinda throwing it back to the sustainability topic. We were talking about the edge and reducing the water consumption or energy consumption in general.

I was just curious if you could explain to the audience what it is about Arm that makes you uniquely positioned to spearhead that initiative. Yeah, no, thanks for the opportunity to discuss this. We as a

Vince: company start primarily in mobile devices. So from the outset of, the company developing technology, it was always envisioned that it would operate in an energy constrained environment because it had to run on a battery.

It didn’t have a, dedicated plug into a wall. And because of that the type of technology that we [00:12:00] developed is unique and distinct from a company like Intel or AMD, which primarily started in the PC server. I guess mainframe space back in the day. And I would say for the first 20 years of arms’ existence we were having to operate and, push as much computing capability as possible into these things that, that are running on batteries.

And so because of that we were designing more efficient technology. We still branch out from that and we developed CPUs for data center, for automotive space, for network infrastructure. But it’s all built on that foundation of having similar computing instruction set. When when companies take ours and develop a shift for the data center and all the aggregators doing now it’s aws, Microsoft Google they’re.

They’re starting with a piece of technology that is inherently more, more efficient. And so what you’re seeing is the chips that they develop are generally more efficient delivering comparable computing power to others on the market who didn’t, didn’t start with kind of that origin.

Farbod: Does that answer your question? Yeah, that’s perfect answer. Yeah. Yeah. And, as the companies use cases have evolved, like we were saying, now being in the service space and stuff, have some of those principles changed or is sustainability still the core goal? Efficiency is absolutely still like a selling point

Vince: For Arm.

It’s, there, there are kinda a couple things that drive our technology development. It’s it’s, maximizing performance per watts. So that’s how efficient do you get performance from a chip security and we do a lot at the hardware level. I think a lot of people think about cybersecurity happening at the software level, [00:13:00] but we actually.

Two, quite a lot at the hardware level that we’ll, we will, provide additional security features as as our technology is incorporated into chips, and that’s incorporated into in devices efficiency and security. And then the third is just this kind of computing platform. So it’s, it’s the same architecture that is used in your earbuds that’s using data centers,

Daniel: common computing platform can develop on top of, we have a, we’ve had an opportunity on the podcast just reaching back to you saying secure hardware design is super important.

We’ve had a conversation on the podcast before. Gentleman named Joseph Robert Cron, who’s a lecturer at MIT, computer Science and Artificial Intelligence Laboratory. And, he’s been studying cybersecurity and physical security for a long time, and I think that’s, he’s doing PhD in now. He’s told us why we were there visiting with him.

Cybersecurity is one domain, but he’s the one I’m most interested in is physical security of the hardware. That’s where a lot of the exploits will be in the future. So it’s interesting to hear these thread you guys at technology level and also folks at the academic level kind of echoing the same feature trend is let’s make sure the hardware super secure.

In addition to this, this large emphasis we’ve for the last several decades on cybersecurity as well.

Vince: Yep.

Daniel: Yep.

Vince: And I would add the OEMs in there as well. We, one, one of our biggest markets is smartphones. And we have ongoing work work streams with, handset makers. The same with other device makers, but handsets, PCs, data providers we work incredibly closely with them.

They’re able to one, give us feedback on additional capabilities that they would need to see in our hardware so that we can start to build that in. And, it permeates the supply chain that way. But the, I would say the [00:14:00] other thing is that those features that we build in doesn’t just benefit, maybe the Googles or the Apples or the AWSs of the world it benefits anybody that uses our technology. It’s certainly we work a lot with universities. They do some really novel things looking at security vulnerabilities and hardware.

But, it’s certainly a collective effort between industry but on semiconductor side and the product side, on the software side and academia and frankly government a lot as well. We maybe like getting too in the we for some of your listeners, but two, three years ago the administration was looking at.

One of the largest class of software attacks, which is memory vulnerability attacks. And we worked with ’em. We actually have a piece of technology in our hardware that can protect against that entire class of attacks. So certainly the government highlighting that and pushing industry to take advantage of these things is like another great example of the government

Farbod: assisting industry to, to protect against hardware attacks or software attacks.

You told us early on that there’s a lot of interesting things happening you’re working in arms, so you’re trying to make sure that government officials know about it. Looking at the next, let’s say, 12 to 18 months, what’s coming down the pipeline that you think is interesting that the audience should know about?

Yeah there’s a couple things.

Vince: I would say in the US so much of industry’s focus has been twofold. One, on ensuring that. The regulatory environment is beneficial to artificial intelligence deployment. So domestically the second is trying to proliferate that technology around, around the globe.

And kinda on that first piece we’re seeing a lot of work being done, no actual policy yet, but I think there’s gonna be kinda comprehensive rollout of policy [00:15:00] actions that help streamline deploying data centers or manufacturing facilities for chips or kinda any, anything that’s involved in AI infrastructure deployment.

On a second on, on the second one on proliferating technology around the US government actually has an open consultation right now. Asking companies how they should help facilitate exports of us technology around the globe. Some of your readers may have seen earlier this year, there were a whole bunch of big artificial intelligence announcements coming outta the Middle East when the presidential administration was over there.

I you talking tens of hundreds of billions of dollars deals. And that was great. I think a lot of us were done kinda a one off way. So the government to figure out how they can can create a structure where they can help deploy us technology in a much more organized and orchestrated way.

We along with all of our customers, all of our customers are all feeding into that right now. Another I would say really interesting thing is taking place in Europe. And not sure how much of your listeners have followed what’s been going on over there, but I mentioned US CHIPS Act earlier that the EU has had a CHIPS act as well, which has tried to promote more investment sector across Europe.

The first piece of the CHIPS Act or CHIPS Act 1.0 outta Europe was very focused on manufacturing and not even kinda leading edge manufacturing. They’re in the process of developing a CHIPS Act 2.0 and they’re trying to facilitate more investment in advanced design and advanced production.

Actually trying to grow the capabilities on the compute side that we feed outta these these leading edge massive AI system. Something else we’re,

Daniel: Tracking and engaging in is appropriately there. One one question we like to ask a lot of folks that come to the podcast, and [00:16:00] Vince, I’m particularly interested in.

Your opinion on this is there’s a lot of overarching, let’s say like understandings of where technology trends are headed. But a lot of times folks who are in a specific space in that industry can kinda call that out and say, oh, that’s not gonna happen. Or it might happen, but it look different. It looked like this.

So we call this our hot takes section. We’d be interested if you have any hot takes on where your field is headed and specifically if it’s a little bit countercultural or maybe some people have misunderstanding of where you think things are headed. This isn’t representation of arms official position, but just your own opinion, Vince, but appreciate some hot takes on where you think the field is headed.

And maybe people are people are looking in one direction. You think it’s down the other.

Vince: Yeah, no, this I appreciate this question ’cause I was still in government when IoT really kinda started to become a thing and I think technology companies were coming in and IoT gonna, cook for you.

It’s gonna control, like all this sort of thing. And, I think that never really came about. I think a lot of what companies were presenting to us at the time when I was in government really needed kind of the technology that we’re just now on, on the cusp of.

Back to my point earlier I really think that artificial intelligence moving out to the edge is just gonna I think we’re just gonna be amazed at some of the innovations that we see. I nerd out and, an early adopter and I had the, I had the meta glasses before, virtually anybody else.

I would like I would just like walk around the house and see what it could do. My wife had a painting on the wall and I would, ask meta who’s this painting by? And I was freaking amazed like it could tell me it was a vocal artist and other works that she’d done.

Or I’ll open my fridge and be like, what can I make like with the ingredients in here? And it’ll gimme a recipe. So it’s I just think we’re gonna see more and more of that. I think it’s those are just some examples of and how I’ve used it and have been amazed. But that, that was technology that was in [00:17:00] market a year or two ago.

I think, as more and more computing is baked into these devices that, we’re u utilizing every day and interacting with every day as models get more and more efficient I think just kinda the, the use cases and the all factor of,

Daniel: What these things can do is gonna blow away.

I like the way you’re thinking on this is I think for a long time we’ve called things smart when they had just a connection to the internet or a screen on it or something like that. But I think we’ll actually start to see things become smart, like with intelligence. Yeah. I don’t know that I really need like a smart toaster as an example, but there are interesting parts of where the way we interact with the appliances and devices and the world around us, when those have intelligence too, it can really change the way navigate world.

Yeah I’ll ask you that next time you get a piece of per toast, but yeah. Maybe I do. One, I guess the penultimate thing we’ll ask you before we wrap up, Vincent’s something we also ask you, we on the podcast is looking back through your career, the trajectory you’ve had, the kinda different places you’ve had, thinking specifically about the contingent of our audience or our students who have yet to start their career, who appreciate if you can kinda look back and think about any I wanna be so cheesy to say life lessons, but some of the things you learned through your career that you wish you knew when you were going back, starting your career, coming outta school.

There’s an awesome portion of our audience that is you kinda sitting in that position. So I, if you can kinda reflect on your career and share, share some lessons for those folks. Listen. Yeah, no.

Vince: Happy to and obviously one, one that comes to mind immediately is be curious. I think you, one, one thing I regret is, I went to college, I got a political science and economics degree and, now working at Arm, even when I was in the government, I was like, I wish I would’ve taken more courses, explored more technical fields.

Like I wish I’d just have [00:18:00] built a broader understanding. Of the issues I was gonna be working with. And whether you’re kinda in the same trap as me in a non-technical field, I would say take course coursework, explore what you can in all these areas because obviously technology’s gonna impact us, more and more, year after year if not day by day.

And vice versa, if you’re in a technology space like, think about some sort of social coursework that like, gives you perspective on how this technology’s gonna impact society or, impact your neighbors. Yeah, I would say that’s one, one thing that I always looked back was like, man, I wish, really wish I would’ve taken an engineering class or two.

I probably would’ve brought my GPA down,

Daniel: but,

Vince: That’s

Daniel: alright. I appreciate the perspective there. The last thing from my end and for if you have anything else, but this is our shameless plugs, so Vince. Folks who made this far in the episode, listen to everything you said so far.

There’s a high likelihood that they wanna see what you’re working on or connect with you personally or see, what’s gonna happen next For Arm. We’re gonna have an opportunity for you to make a couple things. We put in show notes where, what’s your call to action? Where should people head next if they wanna learn more?

Vince: Yeah, I think the reason we first started talking to you guys was ’cause Arm had released an AI readiness index, which is again, a lot of this forward looking analysis of where technology’s going. I did a chapter in there on on AI regulation and we’re seeing across the globe that would be one thing.

Arm has a public policy page, so talk about the kinda issues that we are working with governments on. And then the third one there’s a group that’s, it was actually founded by Eric Schnitt, former CEO of Google, called the Special Competitive Studies Project (SCSP). We just released a white paper with them last week on edge AI and things that policymakers should be considering around that.

So those are three. I would I’ll happy to send a link and provide em show notes, but I would [00:19:00] recommend people take a look at it. They’re interested

Farbod: in this. Awesome. Thank you. Vince, I just wanted to thank you for joining us. It’s always a treat when we get someone that’s working in the interface of technology and some other entity like the government.

Selfishly I really enjoyed hearing your thoughts and those, hearing how those interactions go down and what’s happening on that front. And I’m pretty sure our audience enjoy as well. And I’m hoping you enjoyed the conversation too. So once again, thanks for being here. Yeah, I appreciate, you guys are great interviewers and I’m glad to actually talk with you.

Farbod: I to just listen to you, but I’ll continue to do that as well. So sweet. You heard here for first folks, Vince is a listener to the next by podcast. Next by exclusive. Awesome. Thank you so much for your time, Vince.