2024 Trends and CIO Strategic Focus Areas – Infrastructure Matters, Episode 24

Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this webcast. The author does not hold any equity positions with any company mentioned in this webcast.

Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.

Transcript:

Camberely Bates: Well, good morning and welcome to our last one of the year 2023 from Infrastructure Matters. I am here with my co-hosts, Krista Macomber and Steve Dickens. Welcome.

Steven Dickens: Hey, Camberely.

Camberely Bates: Do we all have our holiday shopping done?

Krista Macomber: No.

Steven Dickens: My wife has literally just given me the last couple of things she wants on her list. She’s terrible.

Camberely Bates: Okay, well this is probably going to get published after you finish with whatever that is. We’re probably going to be published in January. Today what we are going to be doing is covering a bit of what we see are the big trends for coming into this next year. What we’ve done here is we’ve got our own perspective of it, so we’ve kind of rallied around them and we may have some agreements and some disagreements, maybe not. I think we’re probably pretty much in sync because we’re working so closely together with so much of the areas we’re covering. What I wanted to start out with is last year we published some areas and I’m going to do another one this year, on the CIO, CIO imperatives, and CIO strategies. I’ll start off with the first two that I want to talk about and then turn it over to you, Steve and Krista.

But the first two that we identified last year, imperatives meaning you’re going to get fired if you don’t do these right. Number one is staffing, and it’s still a huge issue, and it’s going to continue to be a huge issue and especially as we’ve gotten into some of these new technologies like AI, et cetera. The second one is security, the gift that keeps on giving. We had an incredible briefing from Veritas, Walter Angerer. He’s senior VP at Veritas on their red lab, which is this enclosed lab that they have that’s shut off from everybody else that they insert malware into and watch it play out. I think they literally sit there with a camera because they can’t touch the keyboard. And the kind of things that he talked about about what they’ve seen and how things have shifted over the years over the three months, it’s not stopping guys. So with that, I’m going to turn that over to you guys to maybe comment about where we see whatever trends are happening, maybe that’s going to be impacting helping the CIO in their staffing and also addressing some of the security issues.

Steven Dickens: On the staffing piece, I see a couple of specific dynamics in the space that I cover. Mainframe skills still comes up, so I think there’s challenges there. I think it’s getting better because the software companies, and particularly the development space, we’ve done some good work with BMC this year and their DevX portfolio. That’s making it easier for mainframe developers. So that’s kind of reducing the bar from a skills point of view for people to become developers on the mainframe. I think it’s still a concern maybe on the operational side, although we’re seeing some of the AI stuff starting to filter through. I think Krista and I did a briefing and I’ll let Krista to talk about it more than me with Dynatrace yesterday around some of the security stuff. It’s scary out there, some of the hacks so absolutely agree that any CIO probably the first thing they should worry about every morning is what the first dashboard they should log into to check that everything’s okay is probably the overnight security dashboard. So I mean, Krista, your point of view. You were on the same briefing, and I know you’re the expert in this space, so what was your view?

Krista Macomber: Yeah, it’s interesting as I’m covering more areas of the security stack, some of the common themes that are coming up. And one of them, we talked yesterday with Dynatrace even and that was kind of around the alerting and the visibility, but kind of increasing the confidence that some of these ransomware detection capabilities, for example, actually are detecting signs of malicious behavior and malware and things like that. And that goes directly back to what we’re talking about here, which is the staffing side of the equation. We’ve known for quite some time that IT operations has been very, very pressed for headcount. And then when we look at the security operations side of the house, it’s a very similar story in terms of these teams are inundated with alerts. So that’s definitely I think something that I’m going to be keeping an eye on looking into 2024. Also interesting, I spent some time with the HP Inc team this past week looking at some of their endpoint detection and PC and printer device security capabilities. They talked about some of the alerting as well and detection capabilities and building some capabilities in as well. So again, just interesting how no matter where you look in the stack from a security perspective, that really seems to be something that’s playing out.

Steven Dickens: The only other thing I’d add just on staffing, just to wrap on my thoughts before handing it back to you, Camberely, is we’re going to hear about AI engineering next year. As anybody who’s got deep skills in AI, I think the vendors sucking those skills in so far. I think the end users are going to need those skills in 2024. If this is to anybody coming out of college or anybody thinking of planning on doing a career change, go into as many AI courses as you possibly can and get as many certifications. I think it’s the simplest path to earning more money in 2024 is to become good and proficient at AI.

Camberely Bates: And I know you were recently on Swap network I think this last week talking about some key trends and one of the items that you talked about in there, which addresses the staffing issue is the AI driven operations and observability. Talk about that. What is that all about and where does that play out with this?

Steven Dickens: Well, we’ve got a huge relationship with Splunk. We were talking about Dynatrace. I’ve had briefings from Datadog. Starting to see this coming through across all of the operational tools. I think the landscape of what you need to manage from an operations point of view is ridiculously complex. You’ve got your legacy platforms on prem, you’ve got your new platforms on prem that look and feel and do things differently. Then you may be in three or four public clouds. You’ve got applications running as pure SaaS that you need to back up. I know Krista, you’ve been tracking that space. Just because it’s on AWS doesn’t mean it’s backed up. Just because you’ve got Salesforce doesn’t mean it’s backed up. You look at just the whole landscape. It’s more complex than I’d argue it’s ever been. So just from an operations point of view, the ops teams can’t get their heads around that.

We talked about alert fatigue yesterday when we were chatting to Dynatrace. I mean just the sheer operational landscape. You were starting to see platform engineering and Kubernetes become a mainstream deployment platform. That’s a whole different set of things to manage. So I think AI is going to start to come through in all of those toolings just to make it easier to observe what’s going on and then take remedial action to be able to automate and feed through how you manage those environments. So I think where I see AI, it’s not going to be sexy. We’re not talking about generative AI of images and creating new music or it’s going to be down in the plumbing, down in the engine room. How do we make it easier for an ops team that isn’t going to grow in size typically to do more with less resources? So that’s where I see it playing out.

Camberely Bates: Back to our previous question.

Krista Macomber: Yeah, I was just going to say it becomes more important as attackers are also using AI. They’re creating these, they call them polymorphic or kind of shape-shifting attacks almost that evolve over time as they’re in your environment. The phishing is getting more sophisticated with the use of AI. So to keep up with that, it’s going to become important as well.

Camberely Bates: Absolutely. So let me shift to what I would consider is the CIO top strategic areas. And it’s no secret that AI initiatives for many of these CIOs-

Steven Dickens: I’m shocked, Camberely, you think AI is going to be important. Where did you get AI unique idea from?

Camberely Bates: AI initiatives? And when I say that is being thoughtful in how they’re executing, so it’s a company-wide effort addressing where, how, what, where the dollars are going to go, those kinds of things in terms of those areas. So to that end, there’s a whole lot of things in the infrastructure that’s going to be sprinkled out. I know you’ve talked about some of those areas, so if you want to take it away, Steve on that.

Steven Dickens: Well, yeah, I think obviously AI’s captured the zeitgeist and everybody’s attention in 2023, and I think that’s going to continue into 2024. I think it’s going to be AI initiatives at all levels through the stack. You’re going to be thinking about how you can put AI driven chatbots on your website to help with customer experience. I had a fantastic conversation this year with Nick Delis, who’s the SVP over at Five9 talking about that kind of customer experience, how you empower the customer experience with AI. And he’s starting to see that come through in the kind of real front end customer service customer interaction. And then it’s just going to come back from there all the way through the stack. We talked about some of the plumbing and the tools and how the observability and security people are going to be using it. We are going to see innovation at the hardware level. I think the last three months have seen more innovation in silicon than we’ve probably ever seen in the industry. Intel made some announcements with Gaudi and some of the stuff that they’re doing, but we’ve seen NVIDIA, obviously. We’ve seen Intel, we’ve seen AMD, we’ve seen AWS, Microsoft, GCP. Everybody’s been out there with custom silicon design for AI. So that’s just another platform, another set of instances that an ops team’s going to have to get their head around. I read about Lenovo this week, 17 different servers that they launched, all of them AI focused. So more and more and more stuff coming through. And Dell have made announcements, HPE have made announcements. I think that that’s going to continue. The last six months have just been AI driven everything. That’s going to come through more and more into ’24. So I don’t expect that trend to stop.

Camberely Bates: So we’re going to continue to see the rolling thunder on the technologies out. What happened this week is the Intel had their AI Everywhere Conference in New York City talking about their release of their Emerald product, which is their next generation processor for the servers. And they also talked heavily about the AI PC. Gelsinger is basically saying that I think by 2030 they will have shipped a hundred million AI PCs, and that will be the majority. 80% of the PCs that are shipped will be shipped with those kind of chips that have combined the CPU, the GPU, all the processing kind of stuff in a chip, in one single chip. And that we’re going to see that coming down so that the people that are working on AI technologies or whatever can work on their PC. Also link that back to key pieces of software like an Adobe Suite for design rendering, gaming. But even more than that, I believe is we’re going to see also linking back to the large language models and however those pieces that flow is going to happen.

Steven Dickens: I think Daniel said it best, he netted it down. I think it was in one of his LinkedIn or Twitter posts. The AI PC is the start of the new super cycle for PC refresh. And as he does better than anybody, he sort of nets down the news into that one synthesized sentence. If we’re going to be on a super cycle over the next five years of refresh and people looking at that platform, it’s going to be AI driven, I think.

Camberely Bates: To that end Krista, you were at that HP, not HPE, but the PC people, security session this last week. How does that play into all of this that we’re talking about? I’m sure that they mentioned AI somewhere along the line in that day.

Steven Dickens: Maybe 400 times in the keynotes. I mean just …

Krista Macomber: Yes, I think they made it maybe nine or 10 minutes before they mentioned AI, which was actually pretty impressive. But I would say primarily it actually comes down to attack detection. So looking at different indicators of potential compromise on end user devices, whether that be a PC, whether it be a printer, and kind of using either machine learning, but increasingly also artificial intelligence to be able to uncover, again, indicators that there’s been an attack.

Camberely Bates: And it’s interesting that these AI PCs coming into play now and how much more power we’re going to have on our laptops also, that’s probably going to find its way into our cars some time as well. The other-

Steven Dickens: It’s interesting you mentioned power, and that’s probably a great segue into one of the other big trends I see for 2024, and I would’ve predicted that it would’ve been a trend for this year, but I think just the macro environment kind of meant that we went a different way. So sustainability driven infrastructure. I think, as I say, I would’ve predicted it based on a whole bunch of work I did with IBM on one of their LinuxONE servers. And they went hard into this messaging around, and I’ve seen Lenovo talk about this as well. I think just the macro meant that everybody went focused on kind of business operations, making riffs, getting super tactical. I saw a bull running down the subway in New York, and maybe that’s an indicator on the news last night. Maybe that’s an indicator that the bull run’s starting in the markets. That’s the kind of joke of mine.

But I think now that we’re back to a more positive macro environment, the fed’s going to reduce rates. Things are starting to get better. I predict that 2024 is going to be the year we start to see ESG driven infrastructure coming through, people making some real decisions driven by … And I think AI feeds into this bizarrely. Because those GPUs and processors run really hot, we’re going to actually need it more than if AI hadn’t come along. So I think people are going to be starting to think about workload placement based on ESG concerns. The Lenovo team when they briefed just a few weeks back, you can choose which type of PC you want to buy based on it’s ESG credentials. Then you can decide what type of packaging, what type of delivery mechanism you want. I think people are just going to get more and more aware how do we get back to making some of these pledges actually real.

Camberely Bates: I kind of agree with you, but I don’t necessarily, for the United States yes. We’re still in that mode of-

Steven Dickens: Maybe I’m the European on this call.

Camberely Bates: It is the European issue. We definitely know that that’s hitting them and hitting them hard for multiple reasons. But the United States we still are not seeing decisions being driven.

Steven Dickens: So you think that changes in ’24 or starts to?

Camberely Bates: I think there’ll be pockets because some of the states are doing whatever they’re doing, but there’s plenty of other areas that are not. They’re moving ahead with things, but the decision is not … I can’t remember if I mentioned this in one of the other sessions, but when Randy Kearns was over and working with clients over in Europe, we’ve had two clients tell them that they get no more power, zippo, zero power coming into their data center. That causes you to change your behavior. That’s not kind of things that we’re hearing here. And we also didn’t have the spike in the cost raising of power. Their costs went up eight, ten times because of some of the initiatives that have gone on over there. And that causes you to have to stop and take a look at what you’re burning and how you’re doing that.

Steven Dickens: Some of the pledges, the net zero pledges and some of the things, maybe they’re 2030, but if you take a pause last year, were you behind and you want to catch up? I don’t know.

Camberely Bates: I don’t know. But again, I guess what I’m looking at is from an infrastructure matters and the decisions that are being done in the data center, yes, certainly they’re making some decisions about power and that goes into the concern. But is an overhaul change behaviors completely, not necessarily. There is, and I will go back to one of the other CIO initiatives is the operational efficiency piece of it. So that sustainability or energy, power costs and that kind of stuff goes into that. And especially I think as if they are bringing in more GPUs, which are burning power, yes, we’re going to have those kinds of concerns, especially in the large environments. And yes, the CSPs are concerned about it because it’s a big cost factor for them and they are very visible. So it is as much as a brand issue for them as … It’s important. I mean, I’m not trying to denigrate it by saying they’re doing it for the right reasons, but it’s also a brand positioning that they have in that position. So I think that we’re still slow in the United States, but definitely uptick in the worldwide.

Steven Dickens: What are you hearing, Krista? What’s your perspective on the ESG stuff?

Krista Macomber: Yeah, I mean I would definitely agree. I think it’s more of an issue right now over in Europe than it is here in the United States in terms of it being kind of a requirement when we think about RFPs and things like that. But certainly not to say that that wouldn’t change with time. And I think especially maybe as the United States starts to see some of these initiatives being rolled out over in Europe, I wouldn’t be surprised over the next several years if we start to see some of that approach be adopted here in the US as well.

Steven Dickens: Yeah, I’d agree.

Camberely Bates: I think we’ve got a couple of other ones that were in the list here. One of the ones that I had, and this is a carryover from last year, which is the issue around data and the data pipeline and usability. And the kicker there is the usability. It’s still for AI, it still data has to get fed into the information. It’s got to be quality data or you’re going to train it improperly. So all those areas get to play out. So there’s still this effort within the organizations to look at how do you get, what does that data and data pipeline look like? They’re going to be looking at things like lakehouses where you’re bringing in all the information and you’re not doing ETL, you know, extract, transform and load, but you’re doing that on the backside of the house. So we have more flexibility and usability there. You’re also hearing the CSPs announce they’re aggregating, and the vendors like Adele or that kind of stuff, aggregating access to different large language models, different tools to access the data. So there’s more effort going on to say, “Can I make this process a little bit easier to get to artificial training and analysis?” And I know you’ve got a couple of areas in that area as well, Steve and Krista.

Steven Dickens: Yeah, I think this is going to be the big space for me around AI for 2024. So I agree completely with the trend that you’re spotting. I think getting data curated, prepared, structured, none of that sexy work, but it’s going to be absolutely vital. We were at AWS. We saw them talking about no ETL type solutions. You’re going to see vector databases come into the fore. We’ve seen stuff from MongoDB, we’ve seen stuff from Databricks in that space. One of the best slides I saw this year and kudos to Mark Barrenechea, the CEO of OpenText, was a great use case that he talked about. And I think it captures the point here for me. They took FAA data, a million documents around aircraft crashes, ingested that into their system, prepared it, and then created a large language model demo that you could interrogate why crashes happened. And he was definitive to say, “You give us a million documents, you give us six weeks, and we’ll charge you 350 K to build you an AI project.” And that’s the most definitive I’ve seen anybody. And it was kind of one chart amongst 30 in his deck, but I think it was the most powerful one for me. And two or three of the analysts at the event spotted it as well.

And we were talking about it in the break afterwards, that data preparation, data curation, getting it structured and ready that you can then apply a large language model on top of it. And there’s technology up the wazoo that comes through to do that, whether it’s a vector database, whether it’s a lakehouse, whether it’s more classic data preparation, whatever it is, that is going to be the big task I think for 2024. How do you take all of your documents, all of your enterprise data and then be able to create an AI driven interaction model on top of it, whether you present that out to end clients or whether it’s just for regulatory compliance internally or whatever the project is internally. I think that’s going to be a huge task, and a lot of the vendors, database vendors, data pipeline vendors are going to be coming after that space because there’s going to be big dollars spent in that space.

Camberely Bates: Yep.

Steven Dickens: For sure.

Krista Macomber: I think part of that is understanding who has access to these large language models and how they’re being used and understanding if there’s sensitive or classified data that’s being fed into them. As we talk about kind of curating the pipeline, I think that’s going to be an important part of the conversation too.

Steven Dickens: Yeah, we just did some research with a part of IBM with a company they just bought called Databand, sort of that data pipeline observability. So you talk about operational observability, but applying that same principle to pipeline observability, being able to look across the entire stack. Security’s part of that and also just data quality across the pipeline. So that’s maybe something we’ll put in the show notes for people to look at, but it’s going to be crucial as enterprise AI starts to get more widely deployed. It’s going to be crucial.

Camberely Bates: Well, we’ve hit the 25-minute mark guys.

Steven Dickens: Time flies when you’re having fun, Camberely.

Krista Macomber: It really does.

Camberely Bates: My besties at The Futurum Group. There we go.

Steven Dickens: We’re not recording last week, next week. So can we kind spend a bit more time chatting?

Krista Macomber: I know, I know.

Steven Dickens: It’s not going to be the same without you until January.

Krista Macomber: I know.

Camberely Bates: There we go. All right guys, this has been great. We’ve kind of wrapped up and said thank you very much. We hope you all have fabulous holidays and a great 2024. I think I’m looking forward to this year. It’s been a crazy 2023 and maybe really COVID is behind us completely.

Steven Dickens: We can but hope.

Camberely Bates: I feel like we-

Krista Macomber: Fingers crossed.

Camberely Bates: We’re going into the next generation of our years here, so anyway. Thank you very much for watching and please follow us, click like, all that kind of stuff that we appreciate you doing and continue to tune in. So take care guys.

Steven Dickens: Thanks a lot.

Krista Macomber: Thank you all.

Other Insights from The Futurum Group:

TOP AI Trends for 2024

Intel AI Everywhere Event Unveils Strategic Moves

Some thought on the AI Boom., Silicon and the year behind and ahead