Inside Intel’s AI ambitions

Note: Transcripts are automatically generated from episode audio, and are not fully corrected for spelling, grammar, and formatting.

PAT GELSINGER: AI was an overnight 40-year success. We think about the role of AI, we think about the role of social networks and online experiences: your finances or your autonomous vehicles, your connectivity and community — everything’s becoming more digital. And everything digital runs on semiconductors. And to me, that’s what the siliconomy is about. Fundamentally, everything on the planet needs silicon going forward. 

Never was there a vote taken in Congress to get rid of the industry, but there were votes taken in China, Taiwan, Korea, Japan, to get this industry. And shifting that, the unique role that Intel played, that’s a really, really big deal.

BOB SAFIAN: That’s Pat Gelsinger, the CEO of Intel. Intel has just announced a new AI Everywhere strategy, including what it calls the world’s first AI PC, in a bid to re-establish its dominance in the tech industry.

I’m Bob Safian, former editor of Fast Company, founder of the Flux Group, and host of Masters of Scale: Rapid Response. I wanted to talk with Pat because the rise in AI offers Intel both opportunity and challenge.

Competitor Nvidia has become an AI darling, with Intel playing catch-up, yet also having its own advantages.

Since returning to Intel as CEO, Pat has channeled the spirit of legendary leader Andy Grove who navigated the company through multiple transitions.

Pat shares why the race for AI’s future is still in early days, and how Intel is preparing itself for that future. Plus, his advice on what every business should be doing now to strategically focus on AI. 

Pat also explains how he was “wrecked” when he was pushed out of Intel a decade ago, but that what he learned in the process has been essential, and how he is applying it today.

Let’s get to it. 

[THEME MUSIC]

SAFIAN: I’m Bob Safian, and I’m here with Intel CEO, Pat Gelsinger. Pat, thanks for joining us.

GELSINGER: My pleasure, Bob.

Copy LinkPat Gelsinger describes Intel’s AI PC

SAFIAN: So, we’re just about a year since the public release of ChatGPT. And yet, it’s felt like we’re in a new era. The attention on generative AI, on OpenAI, it’s been intense. Today, you’re hosting an AI Everywhere event, where among other things, you’re announcing the first ever AI personal computer to be on store shelves this holiday season. That’s an unusually quick timeframe for the semiconductor and electronics industry to turn that fast. What is an AI PC? 

GELSINGER: I always enjoy the Andy Grove quote: “The PC, the ultimate Darwinian machine.” It just keeps evolving. It used to be big, beige, and boring. Then it became a laptop and WiFi — it’s just been evolving over time. And we think of the AI PC as this next major evolutionary step of the PC. And what does it mean? Well, all of a sudden you embrace the AI use cases as the core driver. And I’ve called it a Centrino like moment, right? And if you remember, before Centrino, we had WiFi hotspots, but nobody used ’em. And there were two coffee shops in the nation that provided WiFi, and then Centrino, all of a sudden, made WiFi everywhere. And now, there isn’t a proper kid or college student… I mean, WiFi to them is more important than bread and water itself. 

We think of the AI PC as that same ushering in major new use cases — that it’ll drive the change in form factors of the PC… new applications. In the future, we’ll expect that, hey, my AI PC records the conversation. It summarizes the conversation; it translates the conversation. It gives us new user experiences in how we communicate. It’ll also do work on our behalf. Give me a draft legal brief on this subject. Give me my AI bot to agent for today. When was the last time I talked to Bob? Make me smarter on the things we discussed before and the topics I should raise from the day. All of those will be done on my behalf. My medical experiences, my financial experiences, all of those will be augmented intelligence as a result of the AI PC. Why have we been bound to the keyboard? In the future, I’ll just be able to talk in my natural language more proficiently and effectively to my PC, which will enable new evolutions of the form factor of the PC as well. All of this will be the AI PC generation.

SAFIAN: And the new AI PC that you’re announcing now… So it’s both hardware that enables software that can do some of the things that you’re describing. We’re not all the way there yet. 

GELSINGER: Yeah. This is the beginning. And we’ve been working on these technologies for three, four years. So it isn’t just today. As I like to say, AI was an overnight 40-year success, because the core ideas of AI, and we went through multiple AI winters, they’ve been bumbling along for decades. And then all of a sudden, with major breakthroughs in vision and large language models and ChatGPT, all of a sudden it got good enough. We had enough data, enough compute, the algorithms got mature, and all of a sudden, it’s doing amazing things. But, the next two years, we’re going to have the next ChatGPT, and then two years later, the next one and the next one and the next one. We see a decade or more of continued evolution. We have the next two, three generations already well underway, where we enhance the neural capabilities, the audio-visual capabilities, and bring far more unique capacity for the AI workload that everybody’s PC will enable for them.

Copy LinkBreaking down what Pat Gelsinger calls the “Siliconomy”

SAFIAN: Intel started talking earlier this fall about something you called the siliconomy. There’s nothing about being silly, it’s about silicon, right? Can you explain what the siliconomy is about for you?

GELSINGER: Yeah, and I like to think about it in three different dimensions, Bob. One is the core geopolitics of the planet, where oil reserves have defined the geopolitics of the planet for the last five decades, where economies are built, how nation states position each other. Where silicon and technology supply chains is more important for the next five decades. So that’s one aspect of the siliconomy. Second is the pure economics, where today, when the order of 20% of the entire GDP of the nation, and almost 50% of the growth in the GDP is a direct result of technology. And everything technology, everything digital runs in silicon. So you see it as truly the underlayment of economic growth for the nation. And that became acutely visible through Covid, when all of a sudden the supply chains got disrupted, and a $30,000 car manufacturing plant couldn’t proceed because you didn’t have a $1 semiconductor.

But the third aspect is the future of society itself. And as we think about the role of AI, we think about the role of social networks and online experiences. I often ask audiences, what portion of your life is not becoming more digital? Your healthcare is, your finances or your social networks, your autonomous vehicles, your connectivity and community — everything’s becoming more digital, and everything digital runs on semiconductors. And to me, that’s what the siliconomy is about. It’s maybe an interesting word play, but fundamentally, everything on the planet needs silicon going forward. And that is the essence of what the siliconomy is describing.

Copy LinkInside Intel’s competition with Nvidia

SAFIAN: One of your colleagues said to me, it’s been a crazy time for Intel describing 2023 and all the buzz in the industry anchored around Nvidia, which is kind of a difficult tide to turn or to put yourself into. What’s the difference between Intel and Nvidia?

GELSINGER: Yeah, we think about this role of AI. Hey, obviously, Nvidia has uniquely positioned itself to be a beneficiary in that. And they’ve worked hard in this area. But we think of the AI domain as, this is a long game. And we’re in the first or second inning if we can use a baseball analogy. So we have a long way to go. And we’ve defined our strategy in AI as AI everywhere. We are the volume PC player. We are the one who’s been defining the category. We put AI at the edge — every device, every store, every manufacturing supply chain. That’s what we do.

And we’re going to be competing for those highest end inference and training systems as well. We’re very respectful of what Nvidia has done, but we also say: the world needs open systems around AI, and that’s what we do as a company. The second aspect of our strategy here is very much to be a manufacturer at scale. Nvidia relies on other people to manufacture their chips. We manufacture chips. There’s only two companies in the world, we believe, that can be that leading-edge technology manufacturer at scale. And one of those is in Asia. One of them is an American company who’s done our R&D here in the U.S. We’ve been super involved in things like CHIPS Act. And frankly, I want to manufacture for Nvidia, for AMD, for Google, for Amazon. We want to be that supply chain provider, because the world needs a Western supply chain at scale. And we see the AI surge as driving tremendous interest in Intel as a manufacturer, as a foundry, for those opportunities as well.

SAFIAN: Tell me, what do people misunderstand about the CHIPS Act or not appreciate about it?

GELSINGER: Most critical piece of industrial policy legislation in U.S. and in Europe to rebuild the industry. This is the most critical for the ‘everything going digital’ industry. And hey, never was there a vote taken in Congress to get rid of the industry, but there were votes taken in China, Taiwan, Korea, Japan, to get this industry. And, shifting that in two continents and the unique role that Intel played as part of accomplishing that… That’s a really, really big deal, an important trajectory that we’ve reset. I believe that the industry, as a result, will look back on what we do in the last year and say, that was the turning point for the entirety of the semiconductor industry. And we will be living on that for the next five decades.

SAFIAN: Listening to Pat, you get a sense for how deeply central technology is to our world — and how early we are in AI development. So how should business leaders be acting and planning to stay up with the AI curve? We’ll hear Pat’s strategies and advice, after a quick break.

[AD BREAK]

SAFIAN: Before the break, we heard Intel CEO Pat Gelsinger explain how early we still are in AI development. Now, he shares how he’s adjusting practices inside Intel as a result of AI, and his four-priorities advice for other business leaders to keep themselves on course. Plus, lessons he learned after being pushed out of Intel earlier in his career and how he’s applying them now. 

Copy LinkHow Intel has applied AI to its operations

I’ve spoken to a bunch of CEOs over the past year about AI, and their enthusiasm is often matched by uncertainty about how and what changes to make now. You have customers across all kinds of industries. What kind of conversations are you having with other leaders, and how do you help them think about what this transition to AI means?

GELSINGER: It really is understanding their business practices and where they can maximally apply AI to their business. And for instance, one of the initiatives that I have underway is: how do I use AI inside of Intel? We probably have 120, 130 AI projects inside of Intel, a big company going on. But I’ve said, hey, these are my four things that I do more of than anything else. And I want to know how we’re driving AI into those four things. And my four things are: how we use it in silicon design, how we use it in systems and silicon validation, how I use it in manufacturing, automation, telemetry, maintenance aspects, and how I use it for software development.

Those are my four. And I am personally reviewing those four on a regular basis. But I ask every business leader, what are your four? Where are the things that you do the most of, that are most important to your competitiveness? And how are you putting AI to work in those areas? Because these are dramatic changes, when people can say, Hey, my leading software developers are 10 times more productive when they’re using co-pilot-like techniques. Okay, you better be figuring out how to do that if you’re a software developer. One of the top tier consultant groups are 4x more productive if they’re using AI techniques, well, you better be figuring out how to use AI techniques in your consulting practice. And as we’re launching the AI PC, one of the demonstrations that I’ll be giving as part of it is, okay, this is how we’re now putting a core ultra to work in our manufacturing line. I need the highest end manufacturing PCs that are AI capable for my manufacturing to be AI enabled.

SAFIAN: Often when there are new areas, people want to kind of put that off at the side because they’re worried about it disrupting their current practices. If I’m hearing you, you’re saying, you need this. This is such an accelerant and a competitive advantage that if you don’t have it in your core areas, you’re going to have limitations.

GELSINGER: And the numbers that people are demonstrating in terms of business practice transformation, automation, efficiency, like Boston Consulting Group, one of our partners that we’ve worked with, they were demonstrating 10x productivity improvements on the part of their consultants being able to develop proposals for their clients. 10x, right? You’re not talking about little gains, huge gains. And if you’re not doing that, well, trust me, one of your competitors is, or a startup that wants to replace you will be. So you better be looking at it in the context of your core business practices, and then saying, okay, how do we make this happen?

Copy LinkPat Gelsinger’ reflects on being pushed out of Intel before rejoining the company

SAFIAN: You worked at Intel earlier in your career. You left for a while, and then you came back as CEO. How was Intel different when you returned than when you left? And were there things that you picked up and learned along at your other posts that you’re most trying to apply now?

GELSINGER: When I left Intel, I was pushed out of the company. It wrecked me. I wanted to be the CEO. It was just one of those gut punch kind of periods. But when you go through that kind of experience, you also grow and learn. And most of your growth and learning happens in your disappointment, in your pain, in your failures. So it began a period of great personal growth for me — personal maturity, leadership maturity, but also new skill development as well, learning new cultures of companies. I am the first CEO of Intel that was a CEO before becoming the CEO of Intel. I brought a whole set of things: how to work with board of directors, how to speak to analysts, how to set vision. And I brought a software culture and a systems culture to my leadership capacity.

And I came back to Intel that was mismanaged for over a decade. I wanted to combine those new learnings with many of the things that Intel was traditionally great about. And I often talk about, we’re going to get back to the Grovian culture. And half the company knows exactly what I’m talking about when we say Andy Grove and his manic, paranoid, survive, data-driven, engineering-centric culture. Half of the company has no idea what I’m talking about. We are the company that puts silicon into Silicon Valley. We are the stewards of Moore’s Law. We are not going to let Moore’s Law die until the periodic table is exhausted. And we have a lot of elements yet to go.

This deep passion that we truly are the shapers of chemistry and molecules that truly will change the lives of every human, every person on the planet, in fundamental life-improving ways. That’s the company that we are rebuilding. And it plays the seminal role in the nation in manufacturing and in the economic and security requirements of the nation for the future. And because of that, there’s just this resurgence of enthusiasm around the company. And once again, we’re launching today, the AI PC category. We did it once with the Centrino category. We did it once before that with a PC, and we’re going to do it again. That’s the company that we are rebuilding and restoring, and the world needs us, the world wants us, and we will deliver.

SAFIAN: It does sound like it’s been an intense transition for you all the way through. And in some ways, a good thing that it happened, even though it didn’t feel that way at the time.

GELSINGER: Yeah. And hey, out of the greatest challenges and failures, that’s when some of the greatest growth and opportunities emerge, right? The company is clearly humbler now and more customer centric, and the industry recognizes the unique capacity that we bring. But we also have to then go earn, as I like to say, we earned the industry’s distrust. Now I have to rebuild and earn their trust, into the future. And we’re well on that journey to become the manufacturing technology, the standard bearer, once again. And I think today’s announcement is just a great milestone in accomplishing exactly that.

SAFIAN: Well, Pat, this has been great. Thanks so much for doing it. I really appreciate it.

GELSINGER: Very good.

SAFIAN: Listening to Pat, I’m struck by not just his enthusiasm about Intel and technology, but his enthusiasm about the idea of change. There is so much uncertainty about the direction of our world right now, with debates about AI just one case in point. 

But on matters big and small, we can take more control and guide ourselves toward the future we want, if we approach even the tough moments with passion and positivity.  

I’m Bob Safian, thanks for listening.