The Future of Automation and AI with Honeywell CEO Vimal Kapur | At the Money

Pleasure Barry, thanks for hosting me.

Original career plan was to work and get a job. That was a career plan. Yeah, that was a career plan. And you know, first I did two small stints of a job and then I joined Honeywell in early '89. It was a new company in India just set up. So I ended up joining a startup because it was set up as a joint venture between two large companies. There's a large Indian company called Tata Group. They jointly... automobiles, everything. So they invested in this venture. It's big Honeywell with a lot of tech and then they create this joint venture in which you show up and it's basically creating something from scratch. We had no revenue when I started. Our revenue was 0.00. So you learn how to build a company, how you scale, you wear multiple hats like in a startup, you don't have a very defined role. So I think that early experience of high flexibility and you know, growing through a very high pace in a short period of time that laid some very strong foundations, you know, for me.

It turned into what it was planned for because Honeywell did not have its automation business footprint in India at that time. You're talking 40 years back. So they partnered with a local company to scale the business. They already had those products and capabilities in US and they were trying to get into Asia and they formed partnerships in few countries, India being one of them. And the strategy was to penetrate the local market, develop the local capability and we were able to do that quite well. So it's not that we have to change our product strategy but we have to learn as we go through. We had intense local competition, how do you beat that, how do we create our own revenue stream there. So it was a very successful story.

I mean I think it's an advantage in a way to work in a practical business because you have to deal with actual problems which the business deal with. And having worked in different businesses gave me an opportunity to deal with a different customer situation, different end markets, operational issue, commercial issue, product development issue, supply chain. So I would say I mean there's no replacement to formal education. I'm not suggesting that having higher degrees is a disadvantage. But I would say that there is equal amount of advantage to get practical experience. And I was benefiting from variety of experiences I got in my long career in Honeywell.

So process solution business is, you know, it provides automation system in the energy sector. So energy sector think about it, refining, petrochemical plants, other oil and gas facilities, pipeline terminals, even I would say facilities like which may be paper, metals and mining. So these facilities are very complex in terms of their operating procedures and if they're not automated it's nearly impossible to run them. So this business provides a sophisticated automation system to these large companies. So think about Exxon and Shell and BP as kind of a typical customer or Aramco in Middle East and ADNOC. So this serving these customer, this business was very global or is very global even today. The business still is very successful and I became CEO in 2014 of this business. And oil downturn happened within 6 months of I becoming the leader of the business. So you learn through tough experiences. Oil price was from whatever $140, $150 to like a big nosedive and we did a lot of work in that downturn. Learned a lot. But primarily our question, this business is all about sophisticated automation in complex facilities. And then I moved to the building automation business where we still do automation but now in this case buildings of different type, hospitals, airports, schools, university campuses, data centers. And there the business model was very different. Now you serve multiple buildings through variety of channel partners across the world. And so our strength comes through product innovation. Our strength comes through channel management. Very different business model compared to what I did in my process automation days. And then performance material and technology, very interesting business. They build technology, they build energy infrastructure. So if you want to build a, if you are a refiner you buy crude which we all hear a lot about today due to, you know, ongoing Iran conflict. You don't sell crude, you sell product. You sell gasoline, you sell diesel, you sell jet fuel. So they have options to make multiple products and as the input changes or the market needs changes they need to decide what are the options they have to build different offering from their perspective. This business provides technology to energy company to build energy infrastructure because it's a molecule transformation, converting one molecule to another molecule, there's heavy technology involved behind it. So performance material and technology provides technology to the customer to build energy infrastructure. So very high technology or research oriented business. You have a lot of chemical engineers who are going to invent the next best technology and you provide their technology to some very large companies. And that was fascinating to lead that business to see that cycle of innovation and work in that business. So yeah, very diverse experiences in variety of sectors, different business models which I'm benefiting today because now I have experience of dealing with different markets and different situations and that practical experience helps you a lot as you really get into your CEO job.

I mean, I think at that time the biggest challenge that time actually was the chip shortages. And how do we really redesign our products because chips are simply not available. So we really had to learn how do we redesign our products in a much shorter period of time. So think about if we designed a product in one year, we have to do that in two months because there's no other option. If we don't do that, we can't have an alternative source of supply and we can't ship our product. So I used a lot of experiences on dealing with such different scenario in prior jobs and we were able to successfully deal with that. That was also a job I also got exposure to the businesses of Honeywell which I hadn't run before. Aerospace being the biggest one. So that got added into my responsibility. So there was a lot of learning there on how that industry works which is totally different from everything else I had done.

Different animals in the sense of the end markets they serve. Right? There are some commonality of the business models and you know, there's a common denominator but there are differences which really led me to think about whether we are good to be one company or multiple companies when I started as a CEO. And part of it was the differences between them but part of it was the opportunities which is ahead of us that how these businesses independently could shape or scale much differently versus when we are together. Which led us to do a lot of work to think about optionality and pros and cons of each option and which led us to make a decision that we are better off to split into three companies.

Yeah, it evolved a lot. I would say, you know, there was a big change moment in early 2000 when Honeywell and AlliedSignal merged together. Yep. So a little bit of fun fact, AlliedSignal acquired Honeywell and changed its name to Honeywell. Which doesn't happen. The acquirer keeps name because they figured Honeywell brand was so powerful and was more impactful. So they changed their own name. So that was a big moment to your question on cultural assimilation of two large companies. It was kind of merger of equals and it did go through its own motion of ups and downs. And that's when Dave Cote came in as chairman and CEO of Honeywell and Dave did a great job to rebuild the Honeywell culture which was much more one company mindset. We are not two companies, we are one company, we're going to work towards one stock, one Honeywell mindset. Put a lot of operational culture in the organization. So that was one phase under his leadership. Then my predecessor Darius Adamczyk, he became CEO in 2017. He further enhanced our operational excellence skill. He invested a lot of effort to build more digital backbone of the company, simplifying Honeywell in terms of internal systems we have. Darius was very passionate about digital on how to mine data and create more capability for our customers. So he created a culture of more operational excellence, more operational rigor while Dave was much more focused on one Honeywell mindset, cultural integration, not multiple companies. And as my tenure comes in over the last now two plus years, we are now pivoting from the more growth oriented company. And the reason that's important is that over a period of time our margin rates have grown up and we were sub 10% margin company in 2005, 2006. Last year was 23%. So our earnings growth is going to come more from the topline growth versus margin expansion. Not that we won't do margin expansion but we can't get from another 15%. There's no headroom. So growth culture is important which means we have to be more externally focused now. We need to understand our markets, need to understand our customers, what's changing, need to understand our competition. So our company even though name preserves itself as a heritage but it has been constantly evolving itself and that's one of the reason this company has survived 120 years because it has courage to reinvent itself versus being inward looking and always saying that okay we are what we are and we are not going to change.

So I would say we have to go back to where the automation industry started from to better appreciate how will AI impact automation offerings or automation products. Go back to mid-70s when this industry got created somewhere in mid-75 timeframe. When computing was invented, chips were invented, there came the need to say the world has a lot of these expensive assets. Those assets are not running very efficiently. So can we move from the older technologies which were kind of World War I, World War II, war era to more modern digital technologies. And the way automation system was created was that you sense a set of properties on how a particular equipment or a machine or a processor is running and then you have a software program running in a computer which is going to make sure that it gets back to the desired condition what it wants it to be. So it's a logic based predefined system. And the assumption was most of the time this will work in a normal situation. When exception occur human will take a call. So automation systems were always designed with a human in the loop and human was supposed to take care of change in input condition, change in output conditions, maintain the equipment, take care of maintenance requirement down the line. Now you fast forward 50 years before AI and data science came in, the people who are running these equipment or automation system or different facilities in different environment. Think of a pharma manufacturing facility or a data center. They acquired a knowledge on the exceptions which were occurring in those operating conditions. But when they retire or they move on, their knowledge went along with them. So when the next set of people came in, they kind of have the same learning cycle. Maybe some of it was captured in some documents, some manuals, but not a lot. So what AI is solving for is our systems have no intelligence layer on top of the core automation layer. So that when the next human being comes in they're not starting from scratch. They have an advantage of all the learning over the last 25 years all built in. So they get to say when this condition occurred nine out of 10 time this was done it always worked. So you as a human being can say okay I think I will choose this logic. So human still needs to make a decision. So I think it's changing the human and making them more capable at the heart of it. And the reason it becomes even more compelling now is the shortage of skills which are happening in the industrial sector for performing these kind of tasks. So I would say it's a perfect convergence of the situation that more capability is coming into our system because of availability of data science and at the same time situation requires this capability to be there because less people are available to do this work and that's going to create more capability in automation system. So automation system remains, intelligence layer is on top of it. So it makes it automation system better in terms of what it can do by preserving its capability.

The thinking behind that was when I started as a CEO, my incoming thesis was that we have to simplify this company. It's performed extremely well, great return to shareholder, great service to our customer, but what will we do for the next 25 to 30 years? Are we set up for that? And my thesis was that we need to simplify this into few things where we have a scale. But I started the job in middle of 23 as a CEO of the company. Two things happened in the year of 23 which is good to kind of reflect back just 3 years back. That was the first year when aerospace cycle really became very strong. That was the year one where everybody said oh this industry is growing a lot. Let's pay more attention to it. And this was also the first year when something called AI was talked right. So if we were sitting here three years back we wouldn't be talking AI. So it's that recent phenomena. So the question we had to really ask ourselves that if we have to simplify as a company and these two external drivers are occurring simultaneously, a huge demand in our largest business which is aerospace, automation which is core to Honeywell is going to probably redefine itself with AI. Should we do it as one company or should we do it in a different construct? And that question got into a problem solving by early 24 to say let's look at all the scenarios, what's the possibilities, what others are doing. And as we did the work over 2024 we got more and more conviction it's better to separate automation and aerospace into two separate companies. But we ended up making three decisions because specialty chemical is an extremely good business which neither fitted in any one of these two and we said it's compelling to also spin that off as a separate company. So rather than two, they ended up becoming three.

It's a pure play chemicals business. Just straight up chemicals. Chemicals business, they make refrigerant which goes into your car, which goes into your home. They have some other technologies which are related to chemicals. That business is doing extremely well as a standalone company. The automation which you mentioned, building automation or automation of industrial facilities, that's part of the remaining Honeywell which is Honeywell Automation. Now we will not be called Honeywell Automation. We're using it just as a descriptor on what the business will be. We will reimagine our name as we go by in couple of weeks from now and we'll reveal the name what it should be. But for sake of simplicity, a chemicals business, an aerospace business, and an automation business.

So part of it became into advanced material, advanced material solstice. And then part of it is retained within Honeywell. It's split into kind of two.

So think about automation business serves three large end markets: all types of buildings, all types of energy facilities, and all types of industrial facilities. That's what we have kept in the automation. And we also were conscious that we should not make automation business serving so many segments that it becomes confusing again. So we want to narrow down to a few very large and impactful segments. This market size is about $200 billion. We will be just shy of 20 billion of revenue so we have lot of runway to think about creatively what more we can do, how do we grow more. So we're not shortage of runway. Secularly automation is a naturally high growth segment because it's something which is so basic to existence of an industrial facility or on an asset. And then when you add the AI story coming on top of it, it's going to have increasingly more growth momentum. So all things being said, yeah, it's very well positioned for a compelling future.

Right. So we do make aircraft engine for the business jets. Some more midsize, smaller engine. The business jet engines we make, we don't make the big engines. But we are a systems company. We make different component from the nose to tail of the plane. So our components are right in the cockpit. Our components, so we make radars, we make navigation system, we make brakes for the plane, we make environmental controls in the plane. So we are a systems company. We make engines, we make APUs. So our approach is system design for a new platform. So every platform comes in and it could be a commercial plane, could be a business jet, could be a defense platform. We will pitch in different components and systems of Honeywell. Customers will select many of them, some of them, then that will become part of that fleet for decades and decades. So it's a multi-product business not constrained to one particular product line and the business model is more powerful because it's a systems approach and not a component approach. So you're right in the heart of the systems, you understand how the whole mechanics work and really add more value for our customers.

I would say the situation in our case was bit unique because we started doing work to investigate our future optionality early 2024 and did a lot of work and actually even announced the separation of chemicals business in October. Elliott wrote a letter which was in public domain and I got to see it at the same time and everybody else saw it to say we should further split aerospace and rest of Honeywell too. That was their argument. There's a more value to be created. The good news was that we already had done the work and we were convinced that's the right thing to do but we had not announced anything. So we treated them as another shareholder who has a point of view and we have to articulate our strategy. So there was strong convergence on the thinking and I think we worked with them very collaboratively on path forward. And I would say that there's a lot being said on activist shareholder but my experience is that they are like any other shareholder who have a logical argument. If you have a counterpoint you should support these are the facts and data. Or if you support their point then you have to execute it and in that case it just becomes much more of not what to do but how to do it. So our conversation with Elliott like any other shareholder was this is the situation, here are the paths, this is how we are thinking about it. And we benefited from their expertise in capital markets, how the shareholders will react and definitely that helped us to shape our decision in terms of in a certain way which was very constructive.

I think every business model has an era. So I think we have to go back to what created this era of conglomerate or larger companies. It really started from when the world was started becoming more globalized after 2000. China came into WTO, the world became more global and there was much more global trade which became the norm on how companies were growing. All US companies started growing globally. But at the same time they were able to drive a lot of productivity by taking manufacturing into Asia. Lot of manpower productivity by doing work in different virtual way with a lot of IT skills coming in. So there was a case to make bigger companies bigger because they had the unique knowhow to drive a lot of productivity and scale at a global scale because they were already present there. And that cycle persisted for almost 15 years till the time that value was captured. And that value capture became generally known. Therefore the question started asking to say is creating this complex company worth it or simplification or a sector focus is a better way to do it. So I think there was a reason that proposition really worked well and created a lot of value. Take a case of Honeywell, our shareholder value creation from a time of 2000 to 2017, 18, one of the best in class in the entire S&P. So it's not that anything was wrong. We created tremendous shareholder value but now this point of saturation comes in and then it really brings you to the point of specialization. If the markets have scale and you can preserve scale while you're a specialist that's best of both worlds and that's what we are doing now to create a scale aerospace company, a scale automation company. We are still very global, we still have very mature processes but at the same time we are focused on singular segment. So I guess like in everything else you learn through cycles and this...

The cycle is all about having the mix of scale and specialization. This will persist until something else comes in where there's a case to do something else. And I feel good about where we are in our position, and this is going to create much more shareholder value.

I think the situation for each company is very different because separation cannot create value alone by itself. You have to be convicted that the standalone asset has enough growth potential and invest in an asset base which is going to grow, which is going to create value. So I think comparing the example you gave versus Honeywell is absolutely a very different portfolio.

Very, very different. I mean, so I would say that our drivers were more around what I talked about. Our stock price was more static. We did not destroy any shareholder value. So our question was how do we create more shareholder value with external factors coming in, growth of aerospace, growth of AI. Is that inflection point for us to make a different decision? So we did it more from a point of strength versus we have some crisis coming in. So sometimes you use your point of strength to make the right decisions, and we did it fast and we did it right. I think every other company it came from different circumstances.

When we started looking at it, some people believe that we got influenced by the success of GE. I want to remind that GE's success came post our decision. That was a process which was occurring.

We were far along the way in our own analysis by that time. So I think those are parallel things happening. So there's no one thing you can attribute to say that this thing influenced it. It's a combination of the reasons which all come together, and that's what really brings us to where we are today.

So I would say that we as a country have an advantage of being the leader in the space of cloud and data science. And companies like Honeywell have responsibility to take that know-how which the tech sector is creating, be it Microsoft, be it Google, be it Nvidia, and all the very capable tech companies. How do we bring that capability into our sector? Because our customer is not looking to buy a cloud capability or they're not looking to buy an AI LLM. They want to solve a problem. They want to run a business. They want to run an operation. They want to have more uptime. They want to have more profitability. So our job is to take our system to what I mentioned to you before and add this intelligence layer. And what this intelligence layer is all about is taking capability from the tech companies, take large language models from the likes of Google and Nvidia, use the cloud power which is there from Amazon and Microsoft, but really build a purpose-built offering in the industrial sector. And as we are doing that, we are able to create the agentic models for our customers, and that's what they buy from us. The underlying plumbing, what we have, they don't want to know it. They don't want to know how this is built. They say, 'Yeah, so you're automating this piece of my work, that's great, so I'm going to get more productivity for that. How much should I pay you for it?' Right? So I would say we are in the state that this is no more a hypothesis. We are not in the early innings, but we are in the stage of deployment of these capabilities across different customer base. The why it is not taken up at scale is because our customers have to go through a significant change management in their organization because fundamentally the roles of people are changing. Some roles require skills which are less important today and some more new skills are required, and they can't do that overnight just because I created a new set of technology. They have to absorb it, they have to ingest it. But we have some fabulous examples of customers using scale in different sectors like university systems, quick service restaurants. People are using some of our technologies at a very large scale in refineries, etc. So I would say that if I'm sitting with you 8, 12 months back, I would have said very modest deployment. Sitting today, I would say I'm very excited on what opportunity we see. A year from now, I would argue that the penetration will go up substantially because this is real economic value creation from what we are really profiting. And we as a country are leading because we have the core components of this technology. And now we have to take this capability across the world, and our customers are excited. They really like what we are doing.

So, think about it. I mean, when you look at a small fast food dining restaurant, there's not much automation in that, but it consumes energy for sure. I mean, let's take a typical McDonald's restaurant as just an example. There's a kitchen there, there's a fryer, there's refrigeration just keeping a lot of products there. There's of course lights going on. These assets were never thought as a way to improve energy efficiencies by companies like us. We say we should automate a large hospital. Oh, it's massive. There's a lot of opportunity there, or a large building. These assets were never paid attention by us because there was no technology available. But when the cloud technology came in, we are able to connect these assets flawlessly in a matter of hours. And then you're able to use a lot of AI-based rule sets to understand what should be the energy consumption actual versus what it is today and give that tools to the owner to say. An example: we connected a quick service chain in the UK, I think something like 500 plus of their restaurants into a single operating system, and they are observing 30 to 40% energy reduction. Like anything else, the good old management principle: what you inspect is what you get. Once these were running off, I own, nobody paid attention. Even though they desired to do something, there was no mechanism. So we created an easy mechanism to make this available to the customer. So all of a sudden they're able to generate a lot more productivity without adding too much of cost. And that's the power of the new tools which is coming in, which was not possible. And that gives me a lot of excitement that this is going to be much more level of productivity and efficiency which is less talked about. Whenever there's AI dialogue, it's about jobs. It's going to cut jobs. Nobody talks about economic value creation. It is doing a real value for our customer base, making people more productive. That's the story of the industrial side which probably requires more amplification.

Applies to all of these kind of assets. And many people have done this work. So it's not that we have created some new inventions. Some of them have done this kind of discovery, but this effort was not very standardized. It's like a custom-made thing somebody will do because we are a big company, you can afford it. But when you do at a large scale, there are hundreds of these chains, there are hundreds of retail stores. We're also doing similar work around the big retail store chains, very similar example. So these distributed assets are becoming a way of capturing value. That's one end of the equation. On the other end of the equation, when you have retirements coming and our customers are worried about knowledge going out of the door, they're looking at a mechanism of knowledge capture so they can perform their task. That's also penetrating very rapidly. So scenarios are different. Some scenarios are looking at 'I never paid attention and now I can do it.' Some are saying 'I have less people, do something about it.' But the capability is fundamentally the same. It's the same capability which solves both the problem.

Fair. I think there's a scenario in your example because it's a technology displacement of some old method versus a more new method. But the reason I believe more bullish about it is that we are solving a known problem, and the known problem is the world has less people to do a lot of work around skilled labor in the industrial world. That's a real problem. So our solution is not trying to find a problem. We are giving a solution to a known problem. Adoption rates are lower because of the change management issue. But this is a change management of the order of 18 months, 24 months, 30 months. This is not a decade. So I remain very optimistic given my experience in these sectors. The adoption rates are going to be much quicker because the problem is real. We are not inventing the problem. This problem exists, and by the way, this problem is everywhere in the world. This is not a US problem only. Skilled labor. Europe has more population shrinkage than the US. Go to Japan and Korea. They have the same problem. China has population shrinkage. So this is a universal issue. This is not invented here. Now we get excited on the job displacement happening with robots and humanoids. That's a small portion of a manufacturing industry. And that probably is also displacing some tasks which humans are not willing to do, like lifting boxes, right? Yeah. I mean, okay, it's not very interesting. But then there are other jobs with other sectors which we address where a physical AI or intelligence layer is going to create a tremendous amount of economic value.

I think the AI impact will be different in each sector. And I think if we make it too broad, we are losing the bigger picture. But when we are making it specific to a segment, then you're being more precise to say in context of the end markets we serve, the industrial sector, I talked about examples there, it's all about the skill shortage issue which is very different from if we are using AI for better search engine, if I may, using AI for making a summary of our talk which somebody can do. That's a very different use case, and one can argue is it going to add productivity or not, or is it going to take away jobs. That's a different scenario from simply not having people to do work. Very different scenario. And I think that makes our case more compelling. The adoption rates are driven by a near-real need versus we are trying to create a need which is unknown. And that's not being talked about a lot more. A lot more dialogue is around job displacements, but those are more in the jobs which could be automated like finance function or HR function. Maybe to a certain degree it's true, but not to the point, my personal view is that it's going to have the amount of impact which is being talked about.

So I think we have to appreciate the fact that where we are deploying AI, it is substantially different from what we are generally talking about in broader public domain. If you think of applying AI in an industrial system, let's take a case of a hospital and I want to apply AI into an automation system to make it more efficient. The data of that is not in public domain. The data is in Honeywell's system or it's one of our competitor's system. So you cannot go to the internet and train anything because there's nothing to train on. So that makes data friction as a big problem in the industrial sector, which in a way becomes a protection layer for us. But that doesn't mean friction becomes a protection layer. But it doesn't mean we should not do anything about it. It's to say, 'Oh, I'm protected.' It means we should take it seriously to think of potential threats coming in because if the data friction is removed, which is hard to do, but humans are very intelligent. So we have worked very hard to remove the data friction and also use our domain knowledge because interestingly, you cannot solve a horizontal problem in the industrial domain. What I mean by that is you do not have a software application like a CRM system or an HR system. The problems of each sector are very different. If you're a refinery, you're trying to produce more jet fuel and more diesel. If you are a life sciences manufacturing facility, you're trying to produce drug with minimal quality giveaway. But if you're a data center, you want more uptime. Your problems are so different. So we can't create a magic AI application and sell to everybody. We have to be purposeful that where do we use our data and what problem we solve, which only come from years of experience. So those two really become, in a way, a constraint for a generic company to come in because the data friction and lack of understanding of domain, which means companies like us which possess both have to solve this problem. And that's why we are very bullish about it to say we're going to do it. We're going to take all the capabilities from tech companies and build a new set of capabilities to take our industry from a pure-play automation to more towards autonomy. And autonomy doesn't mean humans will disappear. Humans will become more empowered. Humans will become more capable. And to the extent there's some skill shortages, it'll address that point.

Well, it's definitely a challenge for companies. To have more stability is what companies want. So I would say that companies like us have very mature processes to deal with it. So every time this issue occurs, we have some sort of disturbance for, depends, 4 weeks, 8 weeks, 20 weeks, who knows, depending on the situation. So we have learned how to deal with it. But it doesn't come without a cost. You lose some growth in that window. You may have to incur extra costs, like it happened in the case of tariffs, because when tariffs got announced, we have no choice but to pay it. Right. Whether we can recover it or not is a subsequent decision.

We did not file any litigation.

It was not big for us. We were mostly down to the second part of your question. We have been doing manufacturing local for local for multiple years. So we made for US in US, made for Europe in Europe, made for China in China. So we don't move a lot of stuff around. However, what we cannot control is the global nature of the components we buy. Right? If I have to buy everything in the supply chain and raw materials, correct? Because we can't make everything. So if a component is made in Korea, like batteries, we have to buy it from there. And if a component is made in China or somewhere else, we have to buy it from there. So that impact is certainly not under our coverage because we don't have an endless capacity to invest in everything. But our core manufacturing, we have 150 factories, you know, and they are well around the world and they're well distributed across the world. I mean, so we don't have this foundational challenge of reshoring, but we certainly have to deal with changing environment in which we have to think about more local supply-based development aligned to what the expectations are at this point of time.

So defense is a big opportunity for our aerospace business. That's about 40% of the aerospace business. So it's certainly the current changes in geopolitical environment and government spending more money is only positive. So it's going to become an even more growth driver for the business compared to what it had. So when we started this thesis two, two and a half years back, we did not predict this level of demand in defense. But now that's really a reality. Whether it's in the US, whether it's some of our US allies, there's a lot more growth opportunity across the board for different products and services we provide.

I think we have to look ahead versus trying to look back and be skeptical about it. If we look ahead, the problems which the world is in front of us, we clearly know the US lead in AI. So how do we protect the lead? We clearly have a lead in quantum, which is one of the businesses we own. How do we really keep that scale?

So we own a business called Quantinuum in which Honeywell has a majority stake. We spun it off as a separate company in 2021.

Correct. So there are technologies in which the US has an advantage. The US has to rebuild its supply base for some of the critical sectors like semiconductors, like pharmaceuticals, which are mission-critical. And I think that's underway, but we need to have patience that those things take years to happen. There's not a switch to say we want to do it and those things show up. They can take 5, 7, 10 years. So I think it's heading in the right direction. We as a country have all the capabilities. We have the capital, we have the know-how, but we have to refurbish some of our skills which we lost over a couple of years in a few portions of the industrial sector. But let's not forget we have very, very capable companies which created the same sector all over the world. So those have not gone away.

It is more thoughtful in terms of how do you prioritize. All things being equal, reshoring is the right thing to do. But my personal view is we should pick up the top five and say, 'Okay, here are the five we want to really go after and make it successful,' because trying to do everything is going to be just extremely difficult in order of prioritization.

I think the point we discussed earlier, that the automation gets heavily enabled by AI and really creates the intelligence layer, and that opportunity it creates is being underestimated. I think this opportunity is real because of the skill shortage, because of the knowledge gap which has got created over a period of time. So I truly believe that's something which needs more conversation and more emphasis.

My early managers. I mean, I was lucky to have some very good managers who taught me different things. You know, not to be fearful about whom you're talking to, how do you think about value propositions, how to think global scale. So I think in Honeywell you're blessed to have some very strong leaders in different parts of my career in the first 15, 20 years, which really shape you. Because what shapes you is the first 15-ish years of your life, because once those value systems are built in your brain, you kind of live by that. And I was benefiting from some very powerful mentors in different parts of the company.

So books I read a variety, both from leadership to sector-specific. So a recent one of the books I'm reading is 'The Prize' from Daniel Yergin. If anybody is interested about the oil economy, please do read it. Six months back I started reading 'Chip War.' So some of the sector-specific things, but also read about leadership of some of the people I admire. Dave Cote was chair of Honeywell for a long time. He has some very fascinating books, 'Winning Now' and 'Winning.' Indra Nooyi joined our board recently. She has some fascinating leadership books. So I read some of them. I read a lot of books on China. I think it's underestimated the scale of that economy. So I think we just need to. There's a book called 'World's View: China's View of the World.' Very interesting book. It's like we have a view about China. What about their view? Have we ever asked them the question, why do you do what you do? So I kind of have very diverse reading habits, ranging from my business-specific to leadership to some of the country-specific. So yeah, I toggle around a lot on that.

I mean, both are fascinating careers. I would say engineering is a career which gives you a lot of options. So do pursue that because it gives you a wide variety of choices. Management is something that people should do who have more willingness to take risk and have courage to make decisions, because in the end, at some point in your career, you will have to do both. And if you think that's not your sphere, that's something you're not good at, I would rather argue then you choose something you're really good at versus otherwise you're going to get saturated at some point. But management is an excellent career by itself. So both are excellent.

I don't know. I think I'm always excited about learning new technology all the time. You know, I'm still very curious to read things, how they work. I think I will say that staying curious is very important for us as a human being. We should never be satisfied on what we know. We should always ask the question what we do not know, whether it is about a technology or a business process or for that matter any fact of life. And more you are curious, more successful you are because you're open-minded and you're always willing to learn. And that has been my principle all my life. Always learn something new about anything and you feel very fulfilled.