International Business Machines Corporation (IBM) Earnings Call Transcript & Summary

Good afternoon, everybody. Thanks for joining day 2 of Citi's Tech Conference. My name is Tyler Radke. I co-head the U.S. software sector here at Citi. We are happy to have IBM. We have the Chief Commercial Officer and Senior Vice President of IBM Software, Rob Thomas. Rob, thanks for joining us. I think it would be great for the audience here, if you could just give us an overview of your role within the IBM organization.

Yes. Today, I run software products, and I also lead our go-to-market, so I wear 2 hats. I did both jobs independently before, now together. So I spend roughly half my time on each of them. On the software side, that means what products are we building? How are we engaging with clients? How are we investing for the future, a decent bit of time on M&A. And on the go-to-market side, it's all about, I'd say, continuing the go-to-market evolution that we started a couple of years building a more technical go-to-market, building out customer success teams and then spend a lot of time with clients around the world.

Got it. Got it.

Maybe give us a sense post-separation on IBM's high-level strategy, what are you doubling down on? And how has that changed over the last couple of years?

On Arvind's first day in the job, the letter that he wrote to employees, which was shared publicly, said IBM is going to be the hybrid cloud in AI company. And I think we've been very consistent in sticking with that. And you can start to see it then pay off in results because focus is a very powerful thing. And that's also led to even more focus around what businesses we're not going to be in. So we divested some of our health care applications. We recently announced our intent to divest the weather company. So we've gotten very focused on what does it take to be the best provider of technology for hybrid cloud and AI. We do that primarily through 2 motions, which is part of how we redesigned our go-to-market around technology. So a team that's focused just on selling technology, bringing very deep skills to clients, helping them implement technology and then consulting, which is being on the front end of how clients are thinking about their transformation with technologies and underlier and being a key partner in that. Lastly, I would say a big part of executing the strategy and the go-to-market is how we've embraced the ecosystem. We've built new partner programs. We've built new strategic partnerships with a bunch of companies that we can talk about as we go. But I think people are getting the sense now that today's IBM is a little bit different than maybe IBM of the past, and we're really optimistic about the direction.

That's great. So you spend a lot of time with customers, obviously, as the Head of IBM sales teams. What are you hearing on the ground? Certainly, there's been a lot of caution out there on the macro front. But are you starting to see new projects get reprioritized? How does the environment compare relative to a year ago?

It makes a big difference where you are when you ask that question. I've been around the world so far this year. I've been Indonesia, Singapore. Coming up, I'm doing Japan and the Middle East and India. I've been to South America. So there's macro factors that we all read about are happening differently in different places. I would say what is dramatically consistent is companies are more optimistic about the role of technology in their evolution or transformation as a company than perhaps ever before. And I think some of that's driven by demographics. I mean, many of the countries that you go to, there's not really enough workers for everything that has to occur as you think about it on a 5- or a 10-year basis. So the only way to deal with that is through automation, through technology. And so I see clients really aggressive about thinking about how to apply technology? How can we go faster? Obviously, at this point, I can't go anywhere without there being a generative AI discussion, as you can imagine. And I'd say it's a little bit different by country, but the one common thread is there's more optimism on technology in the enterprise than I've probably seen in a while.

Yes. And I guess on the generative AI point, and I'm sure we'll go into the specific product areas, but just give us a sense for the, I guess, maturity of those conversations. Is it, "Hey, Rob, what is generative AI?" Or is it what can I procure from IBM? Just give us a sense for how those conversations are evolving in terms of translating into something more than just a conversation.

To give you a little bit of the time line, we started investing aggressively this in 2020. And it was kind of an unpopular investment at that point because nobody really knew quite what it was. It took a lot of CapEx to get started. But we had a thesis that there was something real here that would solve the problems that we all faced in the machine learning era around a lot of work needed on data labeling. It was very manual. It was very time consuming. So that's played out. I think ChatGPT was a wonderful thing to happen because it kind of got everybody's interest. And so we started working with clients in January around watsonx, which we then announced formally in May, and we made available in July. There's tons of experimentation happening. What we found is you have to quickly connect this to use cases. Hello?

I think it's cutting in and out, but yes. [Technical Difficulty]

You have to quickly connect this to use cases. And there are 3 use cases that have kind of jumped to the front of what we're doing. Number one is automating repetitive tasks. It sounds obvious, but there's a lot of jobs in the Fortune 5000, call it, where people are doing repetitive task over and over again. Generative AI can have an immediate impact on that. Next is customer service. Applying large language models to things like watsonx Assistant, you can get incredible containment rates where you don't even have to have humans involved, meaning 90% plus. And then lastly, this one takes a little longer, but application modernization. Clients thinking about how do I use something like watsonx code assistant modernizing my technology infrastructure, moving to cloud native applications.

I think we make -- yes, another mic here.

Moving to cloud-native applications and using generative AI to make that process easier. So things like Code Assistant. So at this point, a lot of experimentation, but as we start to, I'd say, focusing on use cases, you can see how companies will get a pretty quick ROI.

Got it. That's helpful framing. And the other topic that we've heard a lot from other software companies in our coverage is cloud optimizations. And maybe could you share a little bit what you're hearing from clients there? I know it's certainly something that's been in the world at for at least a year. Do we think we're near the end of it? Is it still a couple more quarters to go? What are you hearing on the cloud optimization front?

Well, we bought Apptio, which shows you where our -- where we've placed a bet, which is I don't believe that this is a short-term phenomena or a fad. Companies will always think about, every year, how are you optimizing your investments in technology. And with Apptio, we get 3 pieces. We get ApptioOne, which is about any of your technology that's on-premise; Cloudability, which is cloud optimization; and then a lesser known product called Targetprocess, which is what is my labor spend that's tied to my technology spend so I can look at total cost of ownership? So I don't think this is a one-shot thing. I think this will continue because as companies want to free up budget for generative AI or quantum or whatever they want to do, they're always going to be look at how do I become more efficient, how do I become more productive in how I'm spending? And so we're really bullish on what we're doing with Apptio.

Okay. Makes sense. So I think it's been about 4 years since IBM acquired Red Hat. It was one of, if not the biggest software acquisition at the time, really placing your bets on hybrid cloud, Open Source. Can you talk about the journey so far with Red Hat? Where do you see the biggest growth opportunity going forward as you think about that business?

The thesis when we acquired Red Hat was that the world would be multi-cloud and hybrid cloud. And I'd say for the first few years, nobody really believed that, that was going to be the case. Most of the client discussions I had, everything was about we're choosing one public cloud. We're going to move everything there. I think that completely changed 1.5 years ago, maybe 2 years ago, where now I don't talk to a single client who's not saying, our strategy is hybrid. We're going to be on multi-cloud. And at this point, Red Hat is the best answer for hybrid cloud. You can see it in our numbers. We've talked publicly about how OpenShift is now over $1 billion, growing 30% plus. So we have a lot of momentum with OpenShift. OpenShift also drives a lot of Linux. And I would call it the hidden gem in Red Hat that's maybe not so hidden anymore is Ansible, which is about how do you do automation in what you're doing in DevOps. As part of watsonx, we've announced watsonx Code Assistant for Ansible. So you get code generation for developers that are building in Ansible. And so we're really pleased with Red Hat. I think since we acquired, it's roughly 4x the revenue base. So we've gotten major momentum. I think the world has kind of come around to our original thesis. And now it's just about continuing to accelerate this.

Yes. And I guess staying on the Open Source topic. Obviously, Red Hat is probably the most successful success story in Open Source. How important is Open Source in your conversations with enterprises?

We've been on Open Source as a strategy since '99, I would say, where we first got behind Linux. I know Red Hat gets all the credit, but I don't think Linux would have gotten to the point it did if IBM hadn't endorsed it back in the late '90s. I think that was a pretty seminal moment for both Linux and for Open Source. Then you fast forward through the years, we were early contributor and committer around products -- projects like Hadoop and then Spark. A lot of the work we did in Spark really became fundamental to how we build software. Spark is embedded in many of our products right now. So then you get to Red Hat. Obviously, that was a huge step for us in terms of Open Source. And now we're always thinking about as we are applying Open Source, are we better off being the builder, committer, contributor ourselves or do we partner for it? An example I'd give you on Open Source databases, when we looked at that market, it's like it's hard to be the provider of 5 or 10 different Open Source variants of database. So we decided to partner for that. So we partnered with MongoDB, with DataStax, SingleStore, Cloudera and EnterpriseDB for Postgres. It's great for them because they get distribution. We get product integration, so it drives other IBM products. And for clients, they get one place to -- one company to work with on their data strategy. So Open Source doesn't mean we have to be the lead on the project per se. We also think a lot about partnering around Open Source. And probably one of the things I'm most excited about right now with watsonx.data, that is a Presto engine using Iceberg formats. And we see a lot of momentum happening in kind of these new data architectures.

Got it. Got it. On the database side, could you just maybe talk a little bit more about that partnership? Is this an opportunity for some of these database vendors like MongoDB to run at some point maybe on the IBM Cloud? Or how are you also participating in that initiative, if it's another third-party database?

We do product integration because we want somebody that's using MongoDB or Cloudera, just to use those 2 as an example, to want to use IBM data integration tools or to use watsonx along with the data that they're storing there. So we view it as a technical integration. And then distribution. We help them get to market. And some of them are at, they're at different levels of scale. Some of them, we can help in the U.S. Some get more lift from IBM around the world where they may not have a go-to-market channel. So it's a combination of those 2.

Okay. That makes a lot of sense. I guess sticking on the database theme, IBM has been a leader in the database market for a very long time. I think your core offering still has very high market share. I guess, how are you just seeing the trends in that space? You talked about Open Source. And obviously, there's some newer vendors like Mongo or DataStax that are having a lot of growth. I guess, what type of workloads are you still going after with your own products and where are you kind of relying on partners with other database vendors? Just kind of frame the growth opportunity for IBM database.

Database is a very good business because clients make a choice. They start building applications using databases, and they tend to grow over time because data volumes are growing. So really happy with how things like Db2 have done. The thing I'm probably most excited about is I think we're at the start of what I would call the fourth epic in data warehousing. First one was around OLAP. A lot of people came into the OLAP market, think IBM and Oracle, Teradata were probably 3 of the big ones at the time. Next was appliances. We had a very successful play with Netezza. Then you went to the third chapter, which was separated compute and storage. The fourth chapter, I think, is as different from the others as they all were from each other, meaning this is going to be about open architectures. That will be things like Iceberg format. So easy access. Instead of proprietary data formats and storage, it will be open data formats. It will be unified query engine. So we're investing a lot in the Presto project and also Prestissimo, which is a unified query engine. We do work with Facebook around that. I think this is going to dramatically change warehousing. These things take 3 to 5 years to play out. But the economics of the third chapter, separated computer storage, are not great for clients. It's very expensive. And we think coming in with a new architecture, a unified query engine, we can give you 2 or 3x their performance at half the cost. And we think this will be a growth driver for us.

Yes. And is that kind of the rebranded, a Netezza offering? Or is that a new SKU or just kind of talk about the product there?

So this is watsonx.data.

watsonx, okay.

So that is the Presto engine that we're building. Now what does that mean for Netezza and Db2? We still have a lot of Netezza clients and a lot of Db2 clients. So part of being a unified query engine is we're enabling clients on Netezza and Db2, they can start to offload into watsonx.data for certain applications or certain workloads. We can actually treat that as 1 query engine. So it's giving them the ability to utilize the investment they have but then grow into new workloads using watsonx.data.

Got it. Along the similar lines of the data warehouse theme, I think you recently announced hybrid cloud mesh, which supports kind of some data sharing capabilities, data connectivity. It sounded, at least to me, kind of similar to what Snowflake has been talking about in terms of data sharing, which has been a very popular use case for them. Can you just talk about that offering and some of the use cases you're seeing from customers?

I'd say it's a little different than that. So let's go back to kind of how we got here. As we were thinking through our strategy around hybrid cloud and AI, and I talked about the focus on the business, we start looking at what are the places that we don't participate in that are highly complementary to our strategy? The obvious one to me was software-defined networking. And I think we're at the very early stage of networking going from being hardware-centric to more driven by software. And when you think about a client implementing a hybrid cloud strategy, it's really hard to do that without application connectivity. And so hybrid cloud mesh is organic development that we've done on, I'd call it, application-centric networking, which is you can run the same application on multiple clouds, how will they communicate and talk to each other? You could do that between on-premise and cloud. And I think there's a real opportunity here on networking that has largely been untapped. And it's highly complementary to somebody that's thinking about hybrid cloud or modernizing applications. And those are the 2 places that we want to be a significant player.

Okay. Okay. Got it. I guess going back to the watsonx on the generative AI theme. Could you just talk maybe more broadly across IBM, how you're thinking about generative AI, both from the software perspective, but also as you think about consulting and services?

I would describe kind of a stack for how we think about it. At the top layer, we deliver what we call assistance. This is packaging up generative AI models targeted at a use case. So I mentioned watsonx Orchestrate, which is an assistant for automating repetitive tasks, Watson Assistant for customer service, watsonx Code Assistant. These are all literally what they sound like, they're assistance for a nontechnical user to use generative AI. Next, you have what I'd call a software development kit. We want to work with other software companies around the world that one generative AI to be part of their products, but may not be able to invest at the level that we've been investing. Like I said, we've been at this for 3 years. If that's not your core business, it's hard to invest for 3 years before you have a product. So I think probably the best example of that is SAP, who announced that they've adopted watsonx as their generative AI platform. So they can pick up pieces from the SDK that give them different experiences when they're working with clients. Next, you've got the platform, that's the studio, which we call watsonx.ai, building models, training models, tuning models. We've integrated hugging phase so you can use Open Source models. We announced that we've integrated Llama 2 from Meta. We've got IBM models, which are models that we've trained based on our data. Then we got watsonx.data we talked about and watsonx.governance, which we will release later this year. And I think when that comes out, we'll be the only person that has a governance product for generative AI, governance meanings fairness, transparency, model explainability, life-cycle management of models, data lineage. And so as we think about generation AI, we want to deliver that complete stack from assistance to the software development kit to the platform itself.

Yes. And that's a -- the governance is a huge topic of importance. I guess as you think about the core differentiation of IBM, I mean, it sounds like it's kind of the complete stack. But what do you see as maybe the #1 or 2 biggest differentiators compared to AWS or Azure or Google's approach?

I think one of the advantages to IBM's model is the fact that we have consulting and technology. Because it's pretty clear to me, the early years of this is going to be about clients wanting assistance to get something into production and get an ROI, which is why I think you'll see this monetizing consulting for us first and fastest because clients want help. You think about the use cases I talked about, automating tasks, customer care, they're going to want a consulting partner for that. So I think that's one of the advantages that we've had. We've also made watsonx available to all the global systems integrators that we work with. So we're launching centers of excellence with the likes of HCL, Accenture, Deloitte. And so I think in the early days, there's a big services opportunity here. But we're also going to monetize through the software because we love the economics of the software.

Yes. I guess as you look at the portfolio today, obviously, it's evolved a lot since you've joined IBM. But from an M&A perspective, like what are the biggest areas that are most attractive to IBM? Obviously, you have a lot of opportunities, whether it's Open Source, generative AI. But what kind of makes the most sense just from a category perspective?

If you look at what we've done year-to-date, I think it's illustrative for how we work through our M&A process. We've done some smaller deals, which was acquiring largely technology that would accelerate organic products that we are building. Like for example, we acquired Ahana, which is a great team that has committers to the Presto project to accelerate the delivery that we were doing on watsonx.data. That's in the smaller category. We acquired Polar Security because we have a great data security product called Guardium, but we wanted to take Guardium and make that available for multi-cloud and hybrid cloud. So you could look at data across all clouds, and Polar Security was a great, I'd call it, a technology tuck-in that gave us that capability to get the market faster. Then you go to bigger ones. We acquired Apptio, which to me was, I'd say, rounding out our vision around IT automation, which is you need capabilities around IT observability, resource management. But having financial operations with Apptio, it kind of makes us, I'd say, the only provider in software that can deliver everything you need for IT automation. So that's about being a bigger acquisition that's highly complementary to a segment that we had already made big investments in. So we continue to look at deals of all sizes. Size is not a strategy per se, but we're thinking about how do we accelerate the areas in software that are important to us. We also do a decent bit of acquisitions in consulting as well. But for software, we've had a very good year in M&A.

Okay. Okay. Makes sense. I wanted to shift a little bit to automation. And maybe it would be helpful just to kind of frame how you see the portfolio of IBM automation products. Where is the opportunity? Where are you investing? And just kind of the -- how that might change with generative AI?

I would break automation into 3 categories, and all 3 are important to us. First is what I would call business automation or RPA. How are you automating different tasks? I talked about watsonx Orchestrate. What's interesting about watsonx Orchestrate, I think it's a good example of today's IBM, and that the first customer of watsonx Orchestrate was IBM. And we've talked publicly about the use case around HR that we've implemented, where we took an HR service center. And in a period of a year, we were able to automate 90% of the tasks. So massive labor savings and massive increase in clock speed in terms of how we execute. And we learned a lot about what worked and what didn't work in the product by going through that. And so by the time we got to the point of launching watsonx Orchestrate, we were really confident that this would be a great product for the clients that we're working with. So I'd say that's one of the main focuses in what I call business automation. Next, you have IT automation. We started with an organic product called AI Ops that we built. We acquired Turbonomic for Resource Management, Instana for IT observability, and we can now give you basically full stack observability and action for IT automation with the combination of M&A and products that we've built. And then the third category would be alluding to on financial operations or FinOps. So automation to me is business automation, IT automation, FinOps. Maybe there's another category that develops over time, but those are the areas that we're going to be focused on.

Okay. On the transaction processing side, I think the performance there has been better. It's flipped to a growth driver. Could you just talk about what's underpinning that? And are you expecting that to continue?

It's really interesting. The timing for this is highly linked, I believe, to the timing where I talked about when clients shifted from we're going to be public cloud only to multi-cloud. Because when there's a tops-down mandate of, hey, everything is going to one cloud, something like the mainframe, by definition, would get less attention. The minute that clients kind of flipped in their mindset and said, "Wait, hybrid cloud is actually going to be our strategy." Suddenly, the mainframe became critical again, and they started putting more workload on the mainframe. Actually, your company, Citi, spoke at a conference earlier this year, where the MongoDB implementation that you all have, you started running that on Z Linux. Because it was like, why do we need like 300 servers to do this when we could run it on a single Z Linux partition? Massive floor space savings, massive energy savings. It's kind of like a no-brainer. But that probably wouldn't have happened if you hadn't embraced mainframe as part of the IT strategy again. So I think there's been a big shift where mainframe is now a key part of the strategy. That's different. Second is we started to do more innovation around the software stack in mainframe. I talked about watsonx Code Assistant for Z that we just announced. Before that, products like Data Gate, which gives you a read-only version of your data on the mainframe, so you can do warehousing style workloads. So we've been innovating in that stack. Third is we have taken some pricing action that we talked about. And it's proven that clients see value in the platform and will continue on them.

Yes. Okay. Great. Let's talk a little bit more about go-to-market and specifically around partnerships. I know you addressed some of these partnerships on the database side, but given that you're now leading a much more product and software-focused business, just how is kind of the breadth of partners that you're engaging change?

As we thought through our partnering strategy and what we want to do in ecosystem, I would say it came down to a few categories. First was we said we need to pick some big strategic partners that we develop a technology strategy with and a go-to-market motion. And in that category, I would put AWS; Microsoft Azure; Salesforce; Adobe; SAP; more recently, Palo Alto. These are companies where we can build a big consulting practice, but we also have a level of technology synergy. We've put a lot of our software onto AWS. We've put software on to Azure. We have technical partnerships with SAP around watsonx that I described, same with Adobe. And so these are big partnerships for IBM that I would say are equally important to those companies because it's opportunity for both of us. So that was one category was being very thoughtful and selective about big strategic partners. 3 years ago, we had no relationship with AWS or Azure for consulting or technology. So this is a major change in direction that's resulting in revenue growth. Next category, global systems integrators. Now that the majority of the company revenue is product, you have to work with all global systems integrators. So we've gone out of our way to embrace working with the likes of Deloitte, Accenture, EY, HCL, TCS, Wipro, all of them on helping them build skills, helping them get more familiar with IBM technology. In many cases, our consulting team also works with them because a lot of projects certainly in the government space become multi-vendor or multi-bid. And so we've built partnerships out with them. Thirdly, as we redesigned our resell program, it's called Partner Plus now. The problem to solve there was pretty simple. We had to make it way easier for partners to work with IBM, way easier to use software, use our hardware, make it really easy to understand how partners would make money, so we redesigned that program. Those have been the 3 main focus areas, and I think we've made a lot of progress on ecosystem. I think partners, since that this is a bit of a different IBM that they can work with, and we just want to build on that momentum.

Yes. That's great. And I guess do you have rough targets or frameworks in terms of the percentage of sourced deals that you're looking to engage partners with? Or how do you kind of think about where we are in that journey?

I think we're still early in it. Our goal is to be a platform company for hybrid cloud and AI. You can't be a platform company, unless you're getting 50% plus of your revenue from the ecosystem. So we want to be as aggressive as we can in getting partners onboarded to watsonx, partners working with Red Hat. And so I think we're still very early on this, but we've made more progress in 3 years probably than even I thought we could.

Yes. Yes. And as you look internally at your own direct sales force, what are some of the big changes that you've made? And certainly, you're in a lot of really exciting developer kind of markets. How are you thinking about introducing more self-service or product-led growth initiatives as some of the high-growth areas?

Three big changes in go to market. One was compensation. It sounds basic, but salespeople are motivated by the right incentives. We changed them to make them much more lucrative. That's had a positive impact. Second piece was building a much more technical sales force. We've built 2 new teams, one is called client engineering, which is a presales team. It's literally what it sounds like. It's engineers who go to work for a client even before we've signed a deal. So they're trying to help the client explore technology, build MVPs, build pilots, get familiar, that's had a big impact. And then second team is customer success team, which is post sale. Clients bought a product, how do we get them quickly up and running and consuming the product, showing them new features, showing them the new version. That's all about driving consumption and ARR growth. So that was the second big thing was just changing the complexion of the go-to-market. And third, as you say, is product-led growth. We need to acquire thousands of new customers. That's the next ambition, which is typically going to be done through a digital sales motion, handed off to the ecosystem, so they connect it back to the ecosystem piece, which means products have to be discoverable on the web. There has to be trials for the products. I would encourage you all to go look at them. I think we've got a lot out there that are pretty good. But every day is A/B testing and figure out how do we make it better, but that is the third dimension. And that's the earliest of the 3 strategies.

Got it. Got it. I guess are there specific products, maybe the watsonx series that you expect to see the most success with those product-led growth initiatives? Or kind of still too early or...

For product-led growth, the best products to acquire new clients tend to be SaaS products. It's probably obvious but also the ones that are more, I'd say, familiar to an end user. So I think about things like planning analytics for financial planning. It's a great product. Most CFOs don't really want to talk to a salesperson. So if you can give the finance staff a way to experience that product online, you can get a lot of momentum there. Maximo is another good example. And to your point, watsonx will be a big part of our focus for product-led growth.

Yes. Great. Well, in the closing last few minutes here, I did want to ask you a couple of bigger picture questions. Just as you look out into the technology industry, what do you think is -- are 1 or 2 big driving forces that maybe under-talked about? Obviously, generative AI probably has enough press time, but like what do you think people aren't paying enough attention about that could have more dramatic impacts over the coming years?

I would say first is Quantum Computing. If you look at -- there's a recent IDC report that said the market was $1 billion or so in 2022, going to $8 billion in 2027. So there's a real market forming here, granted it's early. But last year, we released a 400 qubit system. This year, we'll go over 1,000 qubits. We have 20 quantum computers, so more than anybody else in the world. We developed Qiskit, which is a developer toolkit for building applications for Quantum. We think we'll get to 4,000-or-so qubits in 2025. So we have a significant technical lead here, but it's an early market. But we're really excited about it because we think there will be significant change in technology based on Quantum Computing. And we think we're in a good position as that plays out. But again, we're investing for the future here. And so I think that's probably number one. Two is probably something you don't think as much about for IBM today, but core semiconductor process technology. You may be familiar with Rapidus in Japan, which is the company behind the Japanese government initiative on bringing semiconductors onshore. We're the technology provider underneath that. So this is work that we do at the Albany Center here in New York. And for everything that you all see about what's happening in the world of semiconductors, including geopolitics, we think having a strong technology development of semiconductor technology is really important.

Okay. Those are very interesting. And I think on the Quantum side, and my colleague, Rob, over there is hosting a session later this conference on it. So definitely tune into that. I guess in the closing couple of minutes here, Rob, it would be great if you could just end with what are your top 2 or 3 biggest priorities for the company, and then feel free to address anything we didn't hit on.

Number one is keep our momentum in go to market. I think we've really changed the perception of IBM in terms of how we're showing up with the clients, but that job is never done. So it's what do we have to do next? We're doing a lot of thinking about expansion globally. I mentioned I've been to Indonesia this year. I'll be going to the Middle East. We're investing in markets that have large GDP that are putting technology at the forefront of how they're driving the transformation in the country. So that's really about evolving everything that we've been doing in the go-to-market. That's number one. Two is organic innovation. How do we get more capacity of engineers? More innovation in IBM happening? Really pleased with watsonx and how it came to market, but also thinking about what are the next 3 or 4 things that will have a big impact on enterprise technology. Third is M&A, which is complements what we do organically, but spend a lot of time thinking through M&A, what we should be doing next. We'll continue to be opportunistic there as those pop up. And lastly is we just want to be a great partner with other partners. So everything I talked about in ecosystem, how do we bring in more partners to be a part of this journey, help them be successful with clients, I'd say those are the big focus areas.

Awesome. Well, thank you, Rob. This is a great discussion. Thanks, everyone, for joining in the room. And I appreciate everyone coming.

Thank you.

Thanks.

Okay. We'll get going. So I think we've saved the best to the last, but you can be the people, who are putting on that. But thanks for staying with us through the day. I welcome for those people on the webcast as well. It will not go unnoticed if you've been at the conference for last 2 days. Is that just about every meeting you go into talks about generative AI. And it is very exciting. It's a paradigm shift. But there's a potential another paradigm shift happening in quantum computing. And it's something we recently did 150-page report on at Citi, working actually with over 20 experts around the world in different companies, including IBM, where Jay will introduce himself in a moment; Amazon Web Services, Eric will introduce himself; and also evolutionQ and Donna will introduce the stuff as well. For the sharp minded to view, you may see with evolutionQ that we have the CEO, Michele, who was going to be here, but had a family emergency. However, Donna has stepped in and when Donna explains her role, and what her previous roles were, you'll understand that we have someone who's an expert in a really, really important area, which is the security implications of content computing, which is an area that we're going to get on to. So thank you for joining us.

So if I can, I may introduce you briefly, but can you sort of introduce yourself and then your positioning, obviously, in the company, but also what your -- how your company is positioned within the content computing landscape, what you're trying to do. So Jay?

Sure. My name is Jay Gambetta, I'm responsible for IBM Quantum. IBM Quantum, we have a very simple mission. Our mission is to make useful quantum computers as well as make the world quantum safe. So we're going to have a lot to say on both those topics. From quantum computing, we're doing everything from building the hardware, the software and the algorithms. I think today, we have 1 of the largest -- well, we have the largest fleet of quantum computers. And we have many users that are using this and looking at quantum computing and working out how it goes to industries. From the quantum safe side, we've been focused on understanding the algorithms, but now we're on -- we're working on how we actually understand where cryptography is used and how we will need to change the enterprises use of crypto as we go forward. We call this crypto agility. So I look forward to talking about both these topics.

Great. Eric?

Yes. Thank you very much, first of all, for the invitation. And I'm super excited to be here. My name is Eric Kessler. I'm heading the science team in the Amazon Braket service, and I will tell you in a second what that is. At AWS, our strategy in quantum technology is really rests on 4 pillars. So in 2019, we have started entering in this field by launching our public quantum computing service, Amazon Braket, where we provide on-demand access to different types of quantum computing technologies, quantum computers, quantum simulators, development environments, et cetera. The second pillar is the AWS Center for our quantum computing. This is our research facility located in Pasadena at the campus of Caltech, where we're developing our own hardware, based on some connecting qubits and doing research into quantum error correction and also algorithms. The third pillar is what we call the Quantum Solutions lab. This is a cross-functional team of experts in quantum computing, machine learning and high-performance compute to engage with large enterprise customers in the field of quantum technologies to develop new algorithms and research applications for industry, both in quantum computing, but then also, we always use the opportunity to investigate what state-of-the-art or beyond state-of-the-art opportunities we have that we can deliver to customers today through high performance compute, through machine learning. And then the fourth pillar that is the latest addition to our quantum technologies organization is the center for quantum networking as the research center located in Boston, where we are researching hardware for quantum secure communication and quantum cryptography and key distribution.

It sounds like quite a lot. Donna, and not just the evolutionQ, but please explain what you're doing at Nest as well?

Sure. So first of all, thank you very much for having me today, and for all of you, I'm sure you're very disappointed that I am not Michele Mosca, who was supposed to be here, the CEO of evolutionQ. He had a family emergency and couldn't be here so he asked me to come. He is a worldwide expert in both the physics and the math side of Quantum. And I would like to claim that. I always claim to be the dumbest person in the room. I was that when I was Head of the Cybersecurity Program at the National Institute of Standards and Technology, which included the crypto program with the PQC work and prior to that, AES and all the crypto work, the risk management work, identity management, et cetera, that was all under me at the time. And after I retired Adnet, I went to work actually for AWS for a while, had a personal emergency -- like change myself and took a year off. And then connected back with Michele, because I had been working with him for years on quantum safe cryptography and how to look at that both from the PQC side, but also from the QKD side, the Quantum Key Distribution side and how you think about what is it that quantum computing can do to help us keep ourselves safe in the world and in the era of quantum. So looking at both of those. And that's really something that evolutionQ is about. Looking at the risk management side of the house, in terms of the year of quantum and quantum distribution networks, one of my favorite topics and like where I started at is back in the late-1980s, so now I told you my age. But looking at applied key management and what that looks like and what that looks like in the quantum era.

So it is hugely important that I think all corporates and all investors realize the security side of it, okay, but we're going to get to that. Before we get to that, it's also hugely important about the commercial opportunity of this and in terms of the use cases that it can go to. So Jay, -- and by the way, IBM are leading in this and Jay is a leading operation, and it's fantastic the time lines that you put around some of this stuff. But what excites you most about the commercial opportunity? And when do you think we'll cross over into what's called commercial quantum advantage?

Yes. So well, thank you for saying how good IBM is doing. So quantum, it is a different type of math. And what's tricky about it is, that math promises many things, one of them is to break encryption and we'll talk about it. But together quantum computer working with that math is one of the hard challenges. One of the things we achieved this year is finally getting to a size of a machine, we call it Quantum Utility, where we could actually do something with this machine that we couldn't do with classical. We put this result out. It made on some fun experiences online. But it was start -- I call us getting into this sort of period, where we've got this tool. And so our road map, which we've laid out, basically shows that we want to every year, continue to double, I call it, the performance of the machines. So this year, we're working on a machine that will be 5x better in performance, and it's on track to come out as well as another one that is pushing the size to 1,000 qubits. And so as you keep going through and you keep making these machines more and more performant, there's going to be a point, where we can get to what we call Quantum Advantage. To me, Quantum Advantage is, can I do something cheaper, faster or more accurate than I do in classical. I don't mind, however, I achieve that. I am optimistic that, that will happen in the next couple of years, but certain that it will happen before the end of this decade. So put that as some type of ratio. So we've -- so first, we've got machines now that are beyond what we can do classically. We've got understanding of algorithms of how to use those machines and to go further that are coming. We've got larger and larger machines getting built every time that are more performant. And we now at a certain point, when I get a machine, we call it logical qubits of a certain size and certain depth. When we achieve that type of machine, there's pen and paper algorithms where we can formally prove that they can do things, and we can see a path to how we will achieve that. So I'm not answering your question, but I'm sort of saying somewhere between the next couple of years before the end of the decade. And my personal favorite area, the way I think of quantum, it's like advancing math. So CPUs are good at adding numbers. GPUs are good at vectors, and this has been good for AI. Quantum is good at finding correlations or relationships. And so that matters for extending what we can do in AI. We've got ideas of how we can do the correlations that are difficult for a classical computer. It matters for predicting materials, and it matters in some problems in the like financial modeling or optimization. But that is hard. It's where the problem. It's not big data. It's not doing matrix modification faster. It's looking at some type of correlation or some relationship between variables that is very hard for a classical computer to do, and that's the fundamental math that we enable. And it will -- by doing that, it will matter in many, many different spaces.

Yes. So Eric, I think it's fair to say that IBM have been leaving on the qubit count, so you guys have a really interesting position in the world because of web services, and you basically get to interact with so many different types of use cases and everything else. What excites you most in terms of the commercial quantum advantage opportunities?

Yes, absolutely. I mean, Jay, already mentioned that, ultimately, the reason why we all are here, I believe, is the prospect of having a large-scale fault-tolerant quantum computer, meaning a quantum computer that can execute quantum aero correction protocols to stabilize the computation along the way and mitigate the noise sources. And we know on pen and paper, as Jay pointed out that in this -- with these devices we have, I think really transformational opportunities in a variety of industries. And I think everybody is excited and everybody, who enters this field is excited for those reasons, right? It's a big opportunity that can potentially disrupt a variety of industries. But I think for me personally, I'm also very excited about the path there because I think we have a lot of opportunities to explore different avenues in the near term before we have that large-scale fault-tolerant quantum computer, the fantastic results that the team from IBM recently published about Quantum Utility is an example for that. But I'm also fit to this by background, so I am excited about the prospect of using these devices for scientific discovery, right? That is the first use case that we're going to probably see that is not yet the $1 billion market with broad industry applicability, but will have applications in Academia and maybe in a more niche industry segments for material science. And arguably, these kind of devices are already used in Academia to study fundamental quantum effects, mostly in an analog fashion, right, not as the universal computers, but as special purpose devices. And I think that's also in the near term, a very interesting avenue to explore. We recently launched last year a device from QuEra computing that follows this model and enables simulation of certain types of hamiltonians, physical systems with a relatively large scale. And we see a lot of interest from Academia in these types of devices. And that starts then also present opportunities, as I mentioned before, for other industry players. And as it happens, these devices from QuEra also map nicely to certain type of optimization problems. So we just had 2 papers coming out from BMW, who work with our Quantum Solutions lab with a very intriguing results, I think about exploring the more heuristic type of algorithms, where we don't have pen and paper proofs and where -- this is exactly the right way to explore the possibilities. And one of the reasons why we launched Amazon Braket to bring these types of devices to a broad community of international researchers in Academia and industry to try out the algorithms, to develop algorithms and test them on real hardware.

You mentioned Academia a number of times, but this is not just Academia. This is you're the biggest cloud company in the world. You have interactions with thousands of corporates. And they're using your Amazon Web Services to get quantum availability?

Of course, yes. And at this stage of the industry, we are actually seeing a lot of corporates being interested in collaborating with Academia. So a lot of innovation is driven in Academia and in many cases, through 2-way collaborations between industry and academia, they're using the devices on Amazon Braket for this type of research. So I think especially, when we look in the vertical of material science, computational chemistry, these companies are doing very fundamental research at times and looking and investigating very fundamental quantum models, quantum physical systems. And there's a lot of opportunity, whereas I wouldn't even see that as such a hard line between industry and Academia in this stage of the industry.

Okay. Donna, I know your background is obviously more on the security side, but just you'll be aware of some of the opportunities. Anything that jumps out to you as that excites you most?

So, I think when I look at the opportunities in the quantum space, I do look at in the sciences, when I was at Nest, of course, the physics program in Nest has been looking at quantum for quite a long time and looking particularly in the medical community and the pharmaceuticals. Those kinds of space customer world applications as well as some of the financial services. I did have a chance to look at the Citi report, which I thought really categorize things nicely for organizations to step back and understand the value proposition of quantum computing in a number of fields and understand that it's not just a theoretical exercise. When I think back about AI and AI when I first came to Nest, it was like really a big program, and then it kind of dropped off and all of a sudden popped back up in the past 5 years, because all of the computing power and the store to the capability and the data collection came up and presented itself and it was the right time, but people didn't give up on it. They just categorized it as something else in that intervening 25 years, right? If you look at quantum and what it offers, it's gone from a science problem in my mind, to really much more of an engineering problem. And when we get it right, it's going to offer these kinds of capabilities, which I find to be really, really exciting. I think the days of, if versus when? If kind of question has stopped and now it's just more of a when?

And when comes potentially, Jay, with double exponentiality, is that fair? I mean, we're living through an excellent situation that's happening in parameters for AI models. Compute is up 1,000 fold in the last 10 years. Is that what quantum can offer or not?

I know what you're referring to. I don't think it's the right way. I think -- as I tried to say before, it's a different set of problems. So we have a partnership with Cleveland Clinic that you would think health company. Why are they interested? And we installed a quantum computer in Cleveland Clinic Hospital. And when they look at their problems, they have a lot of problems, where they find that the AI doesn't work because of their correlations are difficult for them to look at. So they're asking a different question. And I think when we say double exponential or we've put that it makes the view to everyone that quantum is classical just faster, but it's actually fundamentally something really deep. It's the branching of the field of computing. So you've got traditional computing and we see what it's doing, it's doing these amazing things. And now we're going to branch. So we're going to look back and say, there's now classical and quantum and it's a different set of math. And so, when we try to give these, we don't pull out the essence. The essence is can I look at something with a different set of [ buys ]. And this is, honestly, what we find is -- so we have a very large team that works on industry problems. So I think we've worked now since we kicked off IBM Quantum over 50 different enterprises have worked with us on understanding quantum. And it always starts -- they always come in with a view like that because it double exponentials are doing everything at once, fast or faster. And you usually find that the first thing is look at the problem with a different set of eyes, a different way of understanding it. And then it's about trying to map that to the algorithms and the hardware that's coming. And then it's -- I agree, it's not an if, it's a when? When will we be at a point of that, that we can do something that we can't do simulate classically. We just crossed the point of sort of exactly simulating these machines. That's a big difference between controllably simulating them to do a business of scientific problem, but you can see that horizon coming. And so I don't like the double exponential question.

I know you're the expert. You can tell me off, that's fine. So however, when we -- despite you're not liking it. When we get to that state, one of the things I love about your position, Eric, in terms of Amazon Web Services is you can facilitate this in the cloud. Just explain how -- if a company wants to be able to get GPU that they can do that as part of the continuum of services that you're offering to make actually pretty quick.

Yes, exactly. So when we started thinking about Amazon Braket, we made a very conscious decision to not have it as a stand-alone service, but really, in a sense, just another AWS service. So we from day 1, follow all the security mandates and all the operational procedures at AWS, just like any other AWS service because we wanted to envision what does it look like to bring quantum computing as part of a platform like AWS to customers. And as we also all see in the industry, quantum computers will not most likely be used as stand-alone devices. But as coprocessors are accelerators to classical compute, and we really want to think deeply about how to -- how these interactions will be facilitated. And obviously, we consider fundamental compute technology as part of the overall AWS compute portfolio. And as such, we also want to bring the community together a little bit, right? So we want to provide access to different early-stage quantum computing hardware, but also cater to software start-ups that are working on algorithm development or building software solutions that can build their SaaS offerings on AWS that then directly is integrated and can just hit in Amazon Braket API in the same environment, and that interaction between developers of software, developers of hardware, algorithm researchers and enterprise and academia is what we hope and believe will start driving a flywheel for the evolution of the of this technology, and we hope we can contribute a little bit to accelerate that innovation.

Yes. Jay well aware that, obviously, IBM did the same and also you have a consultancy business, and we can get to that math.

I was going to say, if you want to use the most quantum computers in the cloud, you don't go to AWS.

Yes.

And so -- let's be clear and we have a very large cloud.

100%. It's more that the cloud can facilitate this ecosystem quite quickly compared to a lot of -- and we should talk a little bit about to the challenge in the supply chain before kind of conscious of time. So we've talked about it n number of times, the security side of it. Just -- I'm just to explain the risk, if you don't mind out there in terms of security risk that people need to be aware. I think boards need to be aware, there's all season need to be aware of this. And then maybe what the solutions are for that, and then I'll come to these guys who have the same.

Sure. So we have a couple of pretty big challenges confronting us today. The first one relates to some of the discussion that we had at the very beginning of this conversation, and that is with the power of quantum era, some of our cryptography that we have throughout our entire critical -- throughout our infrastructure everywhere will no longer serve us. It will not be secure. And that is our public key cryptographic algorithms. So those need to be replaced. And Nest working with the international cryptographic community has come up with some new algorithms called PQC, post-quantum crypto algorithms. And there are another -- a different set of asymmetric or public key algorithms as potential replacements. And I want to get back and talk about that in just a minute. But the other big challenge that we have is that bad actors can actually begin to harvest data now, store it and decrypt it as a cryptographically relevant quantum computer comes along. Even if there's only 1 or 2 that are available, they will be able to decrypt data that's been encrypted today. So that's a big challenge for us as well. So we've got multiple challenges that we're dealing with. With the -- we have -- we were talking about looking in the future of quantum computing. But actually, we have to keep one eye from a security standpoint on classical computers because those aren't going away. So we have to remain secure. Our cryptographic infrastructure needs to be secure with classical computers in mind as well as quantum computers. We don't know everything that a quantum computer can do. So we don't know exactly what that's going to look like or what we know right now that are the public key algorithms that we're using today won't hold up. We know we have some replacements. Those have been studied. To some extent, they've been studied worldwide, lots of research, but not the length of time that we often have. So Nest selected more than one to put in place. So we have to really update our cryptographic infrastructure. And then the other piece of this is -- what is it that quantum computing itself can do to help us protect our infrastructure and to protect our data. And we're just starting to take a look at that. I think you've talked a little bit about some of the communication -- and for some quantum communication infrastructure and what that can look like. So we have some opportunities, but we have a lot of unknowns as well. So security people have a tendency to want concrete answers and that risk diversity, how do we look at this? How do we build the defense in depth kind of a program that's resilient for cryptographic infrastructure, when we have all of these unknowns. And I think that's kind of what's scaring people, who are really in the know and people who are not in the know or who aren't thinking about this, are like, let's wait until some of this sort itself out. It sounds expensive. It sounds hard and it sounds like there are too many unknowns, let's just wait. And waiting is really not the right answer because these kinds of changes are slow and hard and they affect our entire supply chain. They affect our legal agreements that we have with our business partners, with our suppliers, with the things that we are required to do up our supply chain, et cetera. So there's a lot of work that needs to be done. The other piece of this is something that's a little bit of a thing for me personally, too, is that we have to have the right workforce in place to be able to handle all of this. And yes, there are some folks, who really understand everything that needs to be done, but there are not a lot of those folks out there. And so how do we have the right skills -- knowledge, skills and abilities for the workforce in order to make all of this happen. So there's a lot of work that needs to be done and a lot of planning. And if people -- if organizations start now, there are going to be in much better shape than when all of a sudden, IBM or another quantum computing company announcement says, guess what? I have the qubits and I have the fault-tolerance that I need and look what I can do. And then you have, Oh my gosh, how do I make this kind of change? And how do I do this in a way that protects my customers and my customers' data?

So Jay, you guys are very involved in this in several developments. How worried should people be and what should they do?

Well, I think Donna said that good, I agree. To crypt, harp is now to crypt later. I would add -- if you're creating products that need to be on the field for some period of time. That is another factor to do. When I -- I think, we should -- we are certain that a machine like this one. So I don't think you should wait around. I don't want to scare everyone, because I do think there is some time. With the way we are focusing on it, yes we are being part of the relationship with Nest in submitting the algorithms. The algorithms are one side. I think that's probably going to be the easy part, changing the algorithms. I'm not going to say it's really easy, but the easy part. The hard part is where do you use cryptography? How do you actually find in your millions or billions of lines of code, where you've used cryptography or cryptography? And in my opinion, if we do this right, we can even change way crypto -- we can make ourselves much more agile. We can look at our stacks and we can build for a future where cryptography is only going to become more and more important. Yes. It's already really important. It's your first line of defense. Advances in AI are going to even come down on classical methods, advances in Quantum, I mean, our breaks RSA. So how do you get crypto agility? And so what we've been working on is trying to think through, one, the algorithmic side and we'll keep working with the standards like Nest to submit them. But like how do we get tools that can help you explore, advise and ultimately remediate and set us up for a future of being much more crypto agile. I think the right time, honestly, is everyone should start now and start looking at where they're using crypto and not go into it in a fair way, but be proactive. Where is my crypto? What codes? Like if you've got lots and lots of code, even searching that understanding what crypto is being used is not a trivial problem. And so, whilst I do think upgrading the algorithms is going to be, let's say, semi difficult. We're going to come up with good algorithms. I'm more concerned that classical methods may break those algorithms, so that we need to keep beating against them.

I think that's true.

And then we need to get the tools that make it easier to continue to update.

And how do we have a key management -- a cryptographic key management program that also has that kind of agility, so that the algorithms are the black boxes. But the keys spin around, spin around, spin around the archive. If you were to look at how many cryptographic keys flow through your operation today, it would really be mind boggling to you? And how do you have a cryptographic key management system that is -- withstands all of these kinds of challenges and changes that are recurring and that has that agility that we're talking about so that if one algorithm has a problem, you can make the kinds of transitions that you need and you keep those keys flowing.

So it's a huge issue. There are several problems with this. But Eric, I just want to bring you in. Is there anything you want to add to the security side or...

Yes. As you know, security is what we at AWS, always refer to as job zero, right? So the first priority that comes before all other priorities. And so we take this challenge very seriously, and I agree with everything that my co-panelist have said. AWS is also working with this together with IBM and other industry leaders to define these standards. And that is really the first step before the actual challenge, like Jay pointed out, because the rollout of a new standard can take above the decade ultimately until you got to deploy it everywhere. While these standards are making their way through the process. AWS has already deployed different key exchange protocols in our AWS services such as AWS Key Management Service, AWS Secrets Manager and Certificate Manager so that customers can already start exploring and experimenting with these new protocols today. But ultimately, the hard part will be to arrive at a single standard. It allows us to communicate across different parties.

Yes. Well, I'm old enough to remember Y2K and what we had to go through for that this is significantly like that. I got one more question and then open up the audience if people have them. But we did report earlier this year talking about cybersecurity, massive issue, but there is a shortage of people. In terms of talent in quantum and for what we've been talking about today, is this something you worry about? Or are there other things in the sort of challenges in supply chain that you worry about more? So whoever wants that? Jay?

Okay. Sure. Hiring good people is, I think, my #1 job and making sure that we continue to get the best talent. Yes, we build technology, but I often say to -- because I'm responsible on leading the team. I'm more in the business of people and then we create the technology through. It's wide from day 1 from the quantum computing, we've had initiatives. I think our [indiscernible], which is our open source project, is now in over 500 universities. And so we're trying to make sure that we create that new talent. I would argue that whilst we will still need the physicist and I can say it because I was a physicist, we need the computer scientists that come into our field. So I think that is a key thing, how do we educate cross things and how do we reskill a lot of the experts. One of the things I've had the opportunity of taking a lot of experts in IBM and getting them up to speed in the quantum side. So I think that is large. I don't think it's different to the other areas. I think it's the same type of thing. So I think talent is always going to be one. And then from the technical side, we've focused a lot on creating the current generation of computers. I am pretty confident that the current supply chain will meet the demands of the next few years. But if you are ambitious and want to do something very large, the cryo infrastructure, the cost, the footprint, the reliability, all of this becomes almost like we have a road map. I feel confident I can get it to 1,000. I feel confident I can get to multiple thousand. Getting beyond that is it different like size that we are going to have to start now talking with multiple different industries on how do you bring down the cost, how do you bring up the reliability, say, if the electronics, the control system or the cryogenics. And so I think there's a supply chain -- I wouldn't call it a supply chain issue, I would say a new supply chain needs to be created and creating a new supply chain has risks on vendors because they have to do that development. So there's a complicated thing that's going to emerge in the next 5 years. But if we keep making progress, I'm confident we'll overcome it, but they're the 2, I worry. So people, yes, always. And then this new supply chain, not the current one because the current one we can leverage a lot of things, but the new one that really pushes us to the next level.

Good afternoon, everybody. Thanks for joining day 2 of Citi's Tech Conference. My name is Tyler Radke. I co-head the U.S. software sector here at Citi. We are happy to have IBM. We have the Chief Commercial Officer and Senior Vice President of IBM Software, Rob Thomas. Rob, thanks for joining us. I think it would be great for the audience here, if you could just give us an overview of your role within the IBM organization.

Yes. Today, I run software products, and I also lead our go-to-market, so I wear 2 hats. I did both jobs independently before, now together. So I spend roughly half my time on each of them. On the software side, that means what products are we building? How are we engaging with clients? How are we investing for the future, a decent bit of time on M&A. And on the go-to-market side, it's all about, I'd say, continuing the go-to-market evolution that we started a couple of years building a more technical go-to-market, building out customer success teams and then spend a lot of time with clients around the world.

I'm sure you guys both have views on the list above. But just quickly, anybody, any questions you wanted to if any desire to ask?

Yes.

Currently, what is currently -- what is known is your algorithm that solves -- well, that reduces the factorization, which used to be a hard problem. That's why RSA is based on it by. I personally know Adi Shamir, who is the S in the RSA and [indiscernible] and all those that started it many years ago, like in the '80s, all those algorithms. Currently, there are also other algorithms that are being implemented for cryptography that are not based on the factorization being the large problem. The fact that you mentioned that even all the documents that were digitally signed by the same algorithms are exposed will be exposed once this will become true. It's true. A few years ago, in the RSA conference, there was a session like that and Adi Shamir was answering and they asked him and he said, "I sleep well at night." A couple of months later, the CIO of Citibank was visiting Israel and he gave a lecture about the black swan in crypto in security and cyber. And he was talking about this -- the quantum computing that comes, or I told him. Adi Shamir said, "he sleeps well at night." So he told me that Adi Shamir, doesn't have to protect the trillions of dollars that Citibank has to protect. So it's kind of a threat and there are already solutions for it. It's those things in between that have to be solved. And I don't know how many algorithms are there today besides show algorithms because that's a unique algorithm. One from computer -- computing will not accelerate the current regular algorithms for sorting or all those things that are mostly used. Only the factorization is currently no...

But I challenge you to go back and read a report that was done by a group of experts, when there was the fifth generation issue at the National Institute of Standards on Technology.

They all -- by. Well...

If you remember -- [indiscernible] says in that report and said to me personally, number one, we should have crypto agility in our systems. And this was before the quantum work really was underway. We should have crypto agility in our systems and we should not rely on just one algorithm. And all algorithm should have good strong security proof. And I have taken that. I took that to heart that day, and I have never forgotten it. And it's in the report. So this is also an opportunity. The era of quantum gives us an opportunity as we rely on cryptography more and more and more to build more and more agility in what we're doing and also to provide more of the defense in that kind of approach in our cryptographic infrastructure that we don't have today where it's much more fragile than it should be.

Well, it's been question . So just -- and I think you just I think we can take the debate afterwards. But your point on sleeping well at night is very important. This is such an important issue for Board or companies for investors. We haven't even talked about the national security issues that go with it as well, but there are another thing to do. Let me just finish on just very quickly, sort of yes and no answer in a way, if you can. But I said at the beginning, generative AI is sort of sweeping through everywhere. I wonder -- I don't know the answer to answer your question. Is this accelerating the opportunity with quantum? Or is it taking away some of the attention and everything else that goes with it? So I don't know if you Eric, do you have a view? Is it accelerating it or not or you have a different answer?

I don't think that this is a zero-sum game. I think that quantum computing opportunity can coexist with a generative AI revolution and wave that we're seeing, which is super exciting. And probably there are new opportunities for new quantum algorithms in that domain as well, although they're probably a little bit further away.

Okay. Not zero-sum, Donna?

I think, I want to just kind of say, they did always -- if you haven't started looking at either 1 of them, it might be time to be starting to look and put resources in both areas.

Yes. Well, no, I think everyone is looking at the generative AI. I think more people need to understand content. Jay, final word from you?

On the same technology is, in my opinion, is the most exciting time for technology. So be it generative AI or quantum, we're only going to see technology advance.

100% agree. We could talk about this so much, but thank you for your time. We're out of time. But if you want more, please, till the end of the day. Thank you very much for our panel and you can catch them all. Thank you.