It's a difficult decision I make, and with my family's support to do this. And I thought that after the turnaround cadence, I'm done, go back to my venture capital investment. But I just felt that this is an iconic company, and it's so important for the industry, and also very important for the United States. And a lot of my friends, even my competitors, they're all cheering for me to be successful. Somehow you find a motivation to do something, right? And it's not easy. It's a multi-year turnaround, and I'm in the middle of it. But I think I feel called to do this.
How would you describe the Intel corporate culture today compared to 18 months ago?
I think, you know, the culture when I took over — now, when the company is very successful, that is the dilemma successful companies have. Then you're starting to build a lot of layers and a lot of bureaucracy, and then try to protect what you have. And then over the, you know, 50 years, and I think you're starting to lose that. So I think I want to kind of go back to the original, you know, be innovative, be very scrappy, try to do things fast, and then take risk. And then those are the culture that I want to have. And also the culture of listening, and then listening to customer, and then pay attention to customer, pay attention to detail. And then usually in engineering, people love to do the 90%, the cool stuff. The later 10%, that's where they kind of walk away. And that's where very important: the quality, the yield, the reliability, customer trust. Those are the things that I try to bring that to the attention to the company. That's a culture that we take pride, build the trust with the customer and with the vendor. And then also rally the troops, the team that are more innovative, try new things, and rather than the same old stuff, you know, this is historical, we do this. So kind of a first-principle challenge every decision: why we do this way, and then is there a better way to do it? And like for example, right now AI, you can drive a lot of changes in terms of innovation in design, in terms of manufacturing. Those are the things that, not 'this is what we do all the time' and then just shut up, you know. No, you should be more open-minded and look at what the best example, what are the best tools we can use so that we can adjust ourselves to learn the new tricks.
You mentioned AI, and this is a very special moment for the tech industry with the buildout of AI infrastructure in the country, if not around the world. Does it make any sense to you that we hear people say AI demand is slowing down, or we're going to overbuild on AI infrastructure? Do you see any of these concerns when you talk to customers?
Yeah, I think from my humble experience and knowledge, I think AI is here to stay, you know, just like the internet. And so you can do a lot more things with a lot of agents, you can be more productive and do something more cool and faster. So I think AI can do some prediction that you want to have. So I think all in all, I think AI is here to stay. But meanwhile, like any new technology, sometimes people get carried away, you're starting to grow too fast. Some correction is healthy. Then over time, then you're going to turbocharge again, and then you're going to have multiple really leading companies come in. So a couple of things that if there's any correction: one is security, it becomes very important because if you have millions of agents working, if it's not properly guided, it can create more damage. And so security becomes very important. And then secondly is some of this supply constraint. You know, for example, in memory, I mentioned quite a few times, the memory becomes a constraint. Next year going to be even worse.
The prices have gone through the roof. I mean, they've gone through the roof.
Yes, it's six, seven times, and even that, you're willing to pay, you may not get it. So memory is one, substrate is one, raw material is one for packaging, for CPU we need that. And then thirdly, of course, power is an issue. And then lastly, the capital requirement is humongous and in the trillions of dollars. So in terms of the debt financing, structured financing, those become challenging. It's more expensive to do that. So all in all, I think, you know, look at the capex requirement for the AI and how much revenue generated from AI, it's still a very big gap. So I think it just needs time to digest and catching up. I think the money spent is well spent, but the return will be slower to come in. So people have to be more patient rather than you kind of expecting there's going to be a big jump on the revenue, big jump on transformation. It takes time to develop that.
Have we seen a correction, or do you think one's coming, or are we still in that hypergrowth phase?
I think you already see some correction coming in the last few months. I think we're going to see more coming, and of course going to have some casualties come along. So it becomes very important to be very disciplined and safeguard the company, so that make sure that, you know, when you're deploying some of the AI or the equipment, make sure that you really plan it out, make sure that it's a risk you can take. And then the other part, some of this AI is very concentrated for a few players, and so in some way the risk profile for the investor is increased. So in some way you have to plan, you know, thoughtfully, short-term, mid-term, long-term. And that's why being a CEO of Intel, I need to guide the company and lead the company to do the right things going forward.
