Redefining Tech Careers: Data Centers, AI, and Inclusivity with Carrie Goetz, CTO, StategITcom

[00:00:00] Welcome to the Convergence Podcast. I'm your host, Ashok Sivanand.

[00:00:05] Ashok Sivanand, We are 50% of the population. We are not 50% of the tech industry. If we want tech to work for us, we have to be involved as a community and make sure that our voices are heard and that it works for us.

[00:00:19] On this show, we'll deconstruct the best practices, principles, and the underlying philosophies behind the most engaged product teams who ship the most successful products.

[00:00:36] This is what teams are made of. Welcome back, folks. My guest today is Carrie Goetz. Carrie has extensive experience in the field of data centers and has won awards for her work in the field, including from institutions like CIO.

[00:00:51] She is also an Amazon bestselling author for her book on enabling careers in data centers, especially for women, veterans, and folks in the trades.

[00:01:01] I really enjoyed learning from Carrie about the data center space, something that's become a lot more prevalent with many companies evaluating their cloud spend and reinvesting in in-house on-premise data centers as they prepare for the next wave of computing with generative AI.

[00:01:21] We get to hear from Carrie on the episode talking about the importance of diversity and representation in tech as she discusses her passion for helping underrepresented groups like the women, veterans, and tradespeople find careers in data centers.

[00:01:36] We also get to explore the impact of industry, the challenges of the world, the challenges of navigating in-house versus cloud data storage, and the need for critical thinking in today's digital landscape.

[00:01:49] Subscribe to the podcast to get future episodes as soon as they're published.

[00:01:52] If you find this helpful, give the podcast a five-star rating on your podcast app or hit that like button on YouTube.

[00:01:59] Here is Carrie.

[00:02:02] Appreciate you making the time.

[00:02:03] I know that you're extremely passionate about careers in data centers and especially getting women, trades folks, veterans into the data center industry.

[00:02:16] And I'm curious, where did that passion stem from for you?

[00:02:21] So I think for a long time, I was one of the few, you know, three or four women in the industry.

[00:02:28] Very few.

[00:02:29] There were a lot of us.

[00:02:30] And I do believe that the industry would be a much kinder industry.

[00:02:33] The products that we create would be much kinder.

[00:02:36] If we had more representation of all of the people that use those products and not just a certain group of people or a certain group of coders.

[00:02:45] And in order to do that, we really have to have that diversity involved.

[00:02:48] We need the moms in the room.

[00:02:51] We need the dads in the room.

[00:02:52] We need people of all different backgrounds.

[00:02:55] We need the veteran community in the room to help be watchdogs so that we can make sure that these products that we consume and that we put out for other people to consume are not only safe, but also that they provide the services.

[00:03:09] They're going to work for everybody that's going to use them, not just a select few.

[00:03:12] And I think we've seen examples through the years where that has not worked well.

[00:03:16] You've had a career that spans project management, leadership, writing code, running data centers.

[00:03:23] And out of all the books that you could have written, you chose to write books about helping folks get careers in the data center space.

[00:03:32] So tell us a bit about how you've got to where you are.

[00:03:36] I've been in this industry for well over 40 years.

[00:03:41] Well, not well over, but a chunk over 40 years.

[00:03:44] And I've done pretty much everything in this industry.

[00:03:46] I've been in a consultant most of the time.

[00:03:51] I've been in distribution.

[00:03:52] I've been in manufacturing.

[00:03:53] I have commissioned data centers.

[00:03:55] I've been in this industry long enough to decommissioned a couple of commissions that I commissioned prior.

[00:04:01] But I've done coding.

[00:04:02] I've done networking, set up networking divisions for a couple of consulting firms.

[00:04:07] And so, yeah, I've got a pretty broad depth and breadth of experience in this industry.

[00:04:12] Any companies that might be household names or big industry names that we may have heard of?

[00:04:17] Oh, yeah, there's quite a few.

[00:04:18] There's some G500s in there.

[00:04:20] I certainly have had customers in names that you would recognize all over the place.

[00:04:25] And I think that's pretty much anybody in the industry.

[00:04:28] We all sort of end up working for a lot of the big guys in some shape, form, or fashion, right?

[00:04:33] Absolutely.

[00:04:35] And so you've had a career that spans project management, leadership, writing code, running data centers.

[00:04:43] And out of all the books that you could have written, you chose to write books about folks, helping folks get careers in the data center space.

[00:04:52] So tell me a little bit more about where that passion came from for you.

[00:04:56] A, we don't really have curriculum in data centers.

[00:04:59] We have coding classes, which we teach kids about.

[00:05:02] And if a kid doesn't like coding, they write off technology, period.

[00:05:06] Because that's our own exposure, right?

[00:05:08] Totally criminal.

[00:05:10] We have trades that are so absolutely critical to everything we do in society.

[00:05:16] And they all have a place in a data center.

[00:05:18] They have to be built.

[00:05:19] They have to be run.

[00:05:20] They have HVAC technicians.

[00:05:22] We have electricians.

[00:05:23] We have all kinds of trades that support us in the industry.

[00:05:28] And I think for me, one of the big eye-opening moments, I was speaking on a panel in women in tech.

[00:05:33] And the attrition rate for women in tech at that time was 67%.

[00:05:38] Not that leave their job, but that leave an entire industry.

[00:05:42] And I thought, that's so cruel.

[00:05:44] How awful is that?

[00:05:45] If they knew of all the jobs that were around them where that experience would be applicable, then they could have, you know, a better career.

[00:05:53] And they could do much better, feed their family better, all of those kind of things.

[00:05:56] So I thought, you know, what would be really nice is if there was one book that talked about the data center ecosystem, how all the pieces and parts function in a really kind of plain speak, not over technical way.

[00:06:10] And then frame that around the jobs in the industry.

[00:06:13] So you can pick up this book and you can learn why power and cooling matter to compute.

[00:06:18] You can learn why internet connections and sources of power matter in site selection.

[00:06:25] You can figure out what the cabling does and how the fiber optics come into play.

[00:06:30] So how that whole ecosystem fits together.

[00:06:33] And then if you're in a job and say they teach you about one component of the data center and you start selling that or working with that component and say you dislike that particular piece of the industry and you want to go somewhere else.

[00:06:45] Now you know how everything else intertwines with what you've learned about and you've got a better job.

[00:06:50] But we don't teach data centers.

[00:06:52] We don't teach, you know, we teach engineering classes.

[00:06:55] There's a few capstone projects for data centers.

[00:06:57] But despite the fact that every single person has a digital identity in at least one data center, most people don't even know what they are.

[00:07:06] And so I thought that was criminal.

[00:07:08] And then beyond that book, I wanted to make sure educators could add it to their other classes.

[00:07:14] And so that came about because there is an approved curriculum.

[00:07:18] And if you're not on it, it's very difficult to get anything taught outside of homeschooling or some of the private schools.

[00:07:25] But you can certainly add to it.

[00:07:27] You can have books that are in electives.

[00:07:29] So the book was designed to be read in a weekend.

[00:07:33] So if you're thinking about joining the industry, changing your spot in the industry, you can read a book over the weekend that will give you a good idea of what's going on.

[00:07:41] The Educator's Reference allows teachers to add that same book with experiments and different things they can do for their coding classrooms, technical merit badges, shop classes, you know, masonry, carpentry, electrical, HVAC.

[00:07:58] All of those have a place.

[00:07:59] And then the last book kind of came about as a funny story.

[00:08:04] So a friend of mine wanted to go talk to her kid's class about data centers.

[00:08:10] And they wouldn't let her because it was not on the approved curriculum.

[00:08:13] But they told her that she could read the class a book about farts.

[00:08:16] True story.

[00:08:18] And so I thought, well, if we had a children's book about data centers that explained how the Internet works, how things move around, not only can kids figure out, well, I have this device in my hand, but here's a full option of careers that go right along with it.

[00:08:33] And this is how the Internet works.

[00:08:34] Because we don't explain that to them.

[00:08:36] And I think we're getting to a point where power is such an influx.

[00:08:41] We need people to be better consumers and better stewards of what they do in the digital world, too.

[00:08:46] So it was just a way of kind of touching every age group with some information and knowledge about data centers.

[00:08:55] Fostering an engaged product organization and aligning them with the principles around lean, human-centered design, and agile will more than likely lead to successful business outcomes for your organization.

[00:09:08] But getting started or getting unblocked can be hard.

[00:09:11] This podcast is brought to you by the player coaches over at Integral.

[00:09:15] They help ambitious companies like you build amazing product teams and ship products in artificial intelligence, cloud, web, and mobile.

[00:09:26] Listeners to the podcast can head on over to integral.io slash convergence and get a free product success lab.

[00:09:35] During this session, the Integral team will facilitate a problem-solving exercise that gives you clarity and confidence to solve a product design or engineering problem.

[00:09:47] That's integral.io slash convergence.

[00:09:51] Now, back to the show.

[00:09:57] And so the title of your book, Kevin, is a little bit more focused not just on data centers but also specific to helping women, trades, and veterans get into this industry.

[00:10:10] Why those three?

[00:10:12] Why were they super relevant to you?

[00:10:15] So for women, one, I think no women should ever have to count on anybody else for their livelihood.

[00:10:22] They're great-paying jobs.

[00:10:24] We are 50% of the population.

[00:10:26] We are not 50% of the tech industry.

[00:10:28] As a matter of fact, we're not even half of 50%.

[00:10:31] We're in the teens.

[00:10:33] And so if we want tech to work for us, we have to be involved as a community and make sure that our voices are heard and that it works for us.

[00:10:42] It works for our children and our families' needs.

[00:10:44] For the trades, we are nothing without the trades.

[00:10:47] We don't have chairs to sit in.

[00:10:49] We don't have roofs over our heads.

[00:10:51] We don't communicate.

[00:10:52] Nothing happens without trades.

[00:10:55] And I think that they're a very underappreciated group.

[00:10:57] And when you walk in a data center, the trades are already gone by the time that people come in and look at their work and just ooh and ah over the artistry.

[00:11:06] And I think it should be appreciated for sure.

[00:11:08] And it's a pathway that college isn't for everybody and not everybody wants to go that route.

[00:11:12] And I think we should absolutely celebrate the trades.

[00:11:15] For a veteran community, if you serve this country, you deserve every opportunity at the best job possible.

[00:11:22] We have lots of people in the military that are great at construction, for instance.

[00:11:26] We have a lot of people that are great in cybersecurity and tech.

[00:11:29] I do believe that these are very underrepresented and underappreciated groups, you know, and we need more participation from them in this industry.

[00:11:39] There's arguably progress being made on the DEI, DEIB front and also feels like there's, you know, we're in the first inning of getting to where we maybe need to be, where everyone's thriving and we're all firing on all cylinders, right?

[00:11:57] And so I'm curious.

[00:11:57] See, we're not.

[00:11:59] We are and we're not.

[00:11:59] You don't think we're in the first inning either?

[00:12:01] Yeah.

[00:12:01] So tell me what your thoughts are.

[00:12:03] Well, first off, I think a lot of what's happening right now is affirmative action 2.0.

[00:12:08] It didn't work the first time.

[00:12:09] It's not working the second time.

[00:12:10] And until we start viewing people as fingerprints and not the color of their skin or whatever, I mean, if we really want to talk about a balancer, it's money, right?

[00:12:19] Money fixes a lot of things for a lot of people.

[00:12:22] It has nothing to do with the color of your skin, your neighborhood.

[00:12:25] Money would fix all of that, right?

[00:12:27] If you had that money.

[00:12:28] So let's get people careers and places where they can have that money.

[00:12:32] As far as women in the industry, you know, we started making some progress and the Me Too movement hit and it pulled it backwards.

[00:12:39] Honestly, women in the industry and the support for women in the industry that have come by and for women and by the men that support us because they're dads of daughters and they're, you know, uncles of nieces and they're grandfathers.

[00:12:55] And they want their loved ones and they want their loved ones and, you know, the women in their family to have a better chance, better careers and more in the work front.

[00:13:04] That's really what's moved the needle.

[00:13:06] It hasn't been, hey, here's corporate training 101.

[00:13:09] Everybody take this, answer seven questions and your behavior is going to be expected to fall in line with this.

[00:13:15] It's not.

[00:13:15] It's a person to person connection.

[00:13:18] It's very grassroots.

[00:13:19] That honestly is what's made a difference.

[00:13:21] It's 100% grassroots.

[00:13:23] I think you can take all that teaching that you want.

[00:13:26] I know it's a multi-gajillion dollar industry and they make a lot of money teaching that, but that's not what moves the needle.

[00:13:33] What moves the needle is people caring on an individual level.

[00:13:36] And, you know, some companies do a great job at that.

[00:13:38] Some don't.

[00:13:39] And the ones that don't have more trouble finding really quality people.

[00:13:43] And what are some examples of places that are maybe getting along on the right track or making some progress that you find optimism or hope from?

[00:13:55] Well, I think there's a lot of companies that have made, you know, initiatives to say that they're going to be, you know, 50-50 gender by 2030.

[00:14:03] Some of that went up.

[00:14:05] Some of it went down.

[00:14:06] Certainly COVID changed a ton of that.

[00:14:08] A lot of women came out of the workforce during COVID.

[00:14:10] You know, for all of those, they're goals.

[00:14:12] They're lofty goals.

[00:14:13] The problem is we're probably never going to get 50-50 because you're not going to find 50% of women that want to work in tech, for instance.

[00:14:21] That might not even – just because we need the women there doesn't mean women are going to want to work there.

[00:14:26] We have to make these better environments.

[00:14:27] We still have issues with, you know, sexual harassment.

[00:14:31] If you look at why women leave tech and why women leave the industry, some of those problems still exist.

[00:14:36] If all that training was doing that much good, those problems would be gone.

[00:14:40] They're not, right?

[00:14:42] People are people.

[00:14:43] People are going to be good.

[00:14:44] People are going to be bad.

[00:14:45] There's good and bad everywhere.

[00:14:46] And I think instead of highlighting the bad behavior, we really need to start highlighting the good behavior, and then that changes attitudes and it changes actions.

[00:14:54] And so for me, that's why I say I don't think outside of making a lot of people a lot of money, DEI has been all of that helpful because people are still people.

[00:15:05] You know, if we start celebrating good behavior and quit putting up all these barriers and all these blocks to separate people, then I think we have better outcomes.

[00:15:14] I also love that you're providing different windows of career paths.

[00:15:22] That's something that we did at the work in Integral, too, where we'd help folks that were maybe really suited for software engineering and help them get on that path and overcome the fact that education is really expensive and maybe the classroom setting is not the best way that everyone learns compared to maybe a pair programming setting.

[00:15:40] And we created a program called Integrate Detroit to help those folks.

[00:15:44] And then the other thing we would do there is also recognize that we need product folks.

[00:15:49] We need designers.

[00:15:50] We need a lot of folks that can also benefit from the sort of highly lucrative industry that technology is, but not necessarily have to be wired for coding, whether it's sort of how the brain's wired to process math problems the way that engineers tend to do or the ability to sit and focus for as long as they do.

[00:16:08] So working on a single amount of tasks versus something that provides maybe a little bit more diversity in your day, right?

[00:16:15] And so I love that getting folks into the data center world is also an avenue where it is highly lucrative and it is tech adjacent or, you know, tech is built on data centers.

[00:16:27] Let's be fair.

[00:16:28] Exactly.

[00:16:29] And so that higher earning potential, higher profitability kind of will flow into folks working in the data center too.

[00:16:38] If you're kind of, if it's a trades person working on building a data center versus that same person building a building for something that isn't necessarily as high margin as tech is.

[00:16:47] So that is awesome.

[00:16:50] And there's a lot with certifications to your point, right?

[00:16:53] A certification can be a life changing event for a single mom or a single dad or, you know, somebody trying to make ends meet and just figure out what they want to do in life.

[00:17:04] I mean, most kids go to college for six years now.

[00:17:08] It's not four, it's six just to figure out what they want to do.

[00:17:11] Why would you not get them a certification, figure out what they like, let them get started on a pathway, then let the company pay for their college if they feel it's important.

[00:17:20] But, you know, most of our jobs don't really require a degree.

[00:17:24] If you take a drafter and put him with a good engineer for four years, that engineer can do, that drafter can do all that engineering work.

[00:17:31] If that engineer has been kind and showed him, you know, the things to do and because people learn in different ways.

[00:17:38] Nobody says that you have to get a degree to learn.

[00:17:41] Matter of fact, you can go right now today and look up MOOCs, Massive Open Online Courses, and you can take every class that you would take towards a degree for free.

[00:17:51] Now, you don't get credit for it, but you do get a certificate of completion and you can take all of that for your own satisfaction if that's what you want to do.

[00:18:01] But I do have a beef, I think, you know, with the college system in that, you know, we expect people to go through college, companies that, you know, want to pay for those college degrees.

[00:18:11] But let me just task you with this.

[00:18:13] Say you're a company that puts 30 people through a college education program.

[00:18:17] Most of those college education programs are requiring at least five or six electives that have nothing to do with anything that's going to benefit your company.

[00:18:25] They're electives.

[00:18:27] If you took the money for those electives and said, hey, employee, we will pay for the courses that work for our company.

[00:18:34] You pay for the electives.

[00:18:35] We'll meet you there if you want your degree.

[00:18:38] And then we're going to take the money that those electives would have cost us as a company.

[00:18:42] And we're going to put together an apprenticeship program.

[00:18:44] And out of that apprenticeship program, now we can bring in unlimited people through a nationally accredited apprenticeship program, accredited through the Department of Labor.

[00:18:55] And now we have on ramps.

[00:18:57] We don't have to wait four years for a degree.

[00:18:59] We can put somebody to work in 90 days.

[00:19:01] And we can have somebody develop that curriculum for us and help fill that pipeline of talent for us through an apprenticeship.

[00:19:09] And once those people start in those degrees, if they want to go on and take a management class and figure out how to be a manager or something like that, you can do that.

[00:19:17] But you've just funded your own onboarding program out of existing money.

[00:19:22] The number of jobs in the data spinner space are projected to grow faster than the jobs for software engineers in the next two years.

[00:19:33] And I think one of the big reasons why that is is the introduction of generative AI.

[00:19:43] And I think a lot more executives are understanding the power now that they are firsthand users of AI compared to until this latest iteration, it was a data science team that was coming out with insights or machine learning team that was plugging in an API into your app.

[00:20:00] And it was magic.

[00:20:02] But the executives that necessarily kind of provided the inputs or made decisions around storing data and everything else didn't necessarily feel it.

[00:20:10] That's my hunch.

[00:20:10] I got to see it firsthand in mobile, and it feels like it's happening here.

[00:20:14] So what do you think is going to change in the data center world with this increased demand and also more specific demand around LLMs and generative AI?

[00:20:24] For one thing, we're going to grow, right?

[00:20:26] It's not going to happen in a vacuum.

[00:20:28] I do think that we're going to see a lot of older data centers repurposed just for AI.

[00:20:33] But I do think we're, you know, if you talk about a place where we're really at an infancy and a flux point, it's with AI.

[00:20:41] Because right now, we don't know what the legal liabilities are for AI.

[00:20:46] And we don't know how much information we're giving to these AI engines.

[00:20:51] And so what happens when proprietary information starts becoming public information?

[00:21:25] Absolutely.

[00:21:26] What happens when we're doing is moving towards a more AI-native, AI-enabled user base, but also software engineering base?

[00:21:37] Well, I think because we can use AI to flip things so much faster and make things faster, we're definitely going to see more data centers pop up in more places because we do have a latency problem.

[00:21:50] That is one issue that you have, especially with generative AI, is that your input output has to be exceptionally close to your compute.

[00:21:58] So it's not going to be something you do over the cloud.

[00:22:00] It's not going to be you'll get a result setback over the cloud.

[00:22:04] But for companies that are really embracing this and using this, I think they're going to build their own AI facilities.

[00:22:10] And it depends on whether you're doing regular AI or whether you're doing generative AI, what those power models are.

[00:22:17] But we do have a ton of data center space in the U.S. as people move to Colos that was retired or taken out of commission that could be used for that.

[00:22:25] It's a much higher density application, especially if you're doing generative AI where, you know, 20 years ago and even today, there's a lot of data centers that are built out to five kilowatts a cabinet.

[00:22:37] The newest AI on the Blackwell platform is 150 kilowatts in a cabinet.

[00:22:43] That's a lot of increase, right?

[00:22:45] So the power density is different, which means it's a change in your batteries, your UPS systems, your communication systems.

[00:22:52] We are communicating much faster, so we need much lower latency.

[00:22:56] So your input and output has to be similarly located for those generative AI facilities.

[00:23:01] I do think that there's a lot of AI around us, though, and a lot of companies have already accommodated that.

[00:23:07] If we think of the machine learning and the chatbots and some of the intelligence, you know, that we're building into applications, a lot of that exists around us.

[00:23:14] Now, those application providers, their data centers, if you look at a Microsoft and a Google and those guys that are building out AI, their hyperscale data centers are about to be, I don't know, what's hyper or hyperscale?

[00:23:26] What's super hyperscale?

[00:23:27] Hyper super scale?

[00:23:28] I don't know.

[00:23:31] But definitely going to need much more of them, right?

[00:23:33] And I do think that we as consumers have to be more mindful and do a much better job of what we're doing for compute that's responsible and is in need and how much of it is just a silly-nilly want because we are consuming power to make that happen.

[00:23:50] And if you're concerned about sustainability, there's a whole area of sustainability within data centers in and of itself.

[00:23:57] But we have to do a better job of making sure that we compute in a better way.

[00:24:03] But we also, I think, as consumers have to be better stewards with our digital sales.

[00:24:08] So I've got a few questions here of things that I've pondered around how I might make decisions myself as leading a product team or how I might advise an executive at a Fortune 500.

[00:24:20] And I'm curious to know your takes, all right?

[00:24:22] You talked about sustainability, so I'll get to that one first.

[00:24:25] So there's a ton of resources, time, material, money, energy being dedicated towards AI development.

[00:24:33] Various companies, it's an arm race around building models and really new things that we're seeing with Facebook, for example, open sourcing their model and making it way more accessible than the other ones.

[00:24:45] And so sustainability is one that you mentioned.

[00:24:49] The amount of energy that is going into this, to running the data centers, but also a lot of the raw materials.

[00:24:55] And so I'm curious what changes you see coming around the corner that folks outside that specific niche don't necessarily see just as yet.

[00:25:05] So from a sustainability standpoint, a lot of our green resources in the U.S. are not grid connected yet.

[00:25:11] They're out there.

[00:25:12] Several of those projects have been scrapped.

[00:25:14] And I think that we've seen so many of these climate predictions not come true that I think we're starting to see some pushback and some level headedness.

[00:25:25] So, for instance, some of the offshore wind farms have been scrapped now because they're killing the whales.

[00:25:29] And we have issues when the wings break off the wind farms.

[00:25:34] It's a 350 section that has to be buried or disposed of in some way.

[00:25:40] We have problems with the solar farms now.

[00:25:42] We just had the largest water-based solar farm turn into rubbish from a bad weather event.

[00:25:49] We had another one, a solar land-based solar farm that was destroyed in a hailstorm.

[00:25:55] So I don't think we're everywhere that we need to be.

[00:25:57] Right now we can do carbon-captured, carbon-sequestered natural gas.

[00:26:01] We could build a carbon-inert data center with natural gas today.

[00:26:04] That technology exists.

[00:26:06] Why don't we do it?

[00:26:07] Because it's expensive.

[00:26:09] And that's, you know, a big driving force.

[00:26:11] Now, the data centers, most data centers right now, we operate at about 65% stranded power, which is, for those of you who don't know, power that's supplied and, you know, put out there but is never used for whatever reason.

[00:26:26] It's backup power in case primary power goes bad.

[00:26:29] It's power that's allocated to a cabinet but the cabinet doesn't have that many resources in it.

[00:26:34] It's part of the way you balance the floor and the room.

[00:26:37] You know, some of it, there's all different things that we can do.

[00:26:41] Certainly doing on-site generation is better because we're not losing all that power in the transmission as we go across.

[00:26:47] Mini-nuclear has, I think, some play here.

[00:26:51] But I think from a power standpoint alone in the sustainability, that's a big thing.

[00:26:56] Now, if we move to lithium-ion and some of that, the amount of land that we have to destroy to get lithium batteries for backups in the data center is pretty criminal.

[00:27:05] And we have lead-acid, which is almost 100% recyclable.

[00:27:08] And so from that perspective, we're better off with lead-acid batteries than we are lithium-ion batteries.

[00:27:14] We do have other technologies coming up in the battery world too.

[00:27:17] We have sodium-ion, cobalt-based.

[00:27:20] There's other things happening around that.

[00:27:22] So from a sustainability standpoint, for as mature as we think we are, we're very, very immature.

[00:27:29] And I think that we have to do a much better job of not silencing people that bring about contention.

[00:27:36] We have to deal with that contention because, again, that's where that innovation happens.

[00:27:39] If we all think we got it licked, then nobody strives to do anything better.

[00:27:43] And we don't have it licked.

[00:27:45] We simply don't.

[00:27:46] We're a very huge, very big company.

[00:27:48] 60% of our power still comes from natural gas.

[00:27:51] If we did carbon sequestration on that right now, all of that would be carbon inert.

[00:27:58] If carbon is really what we're chasing here.

[00:28:00] The next one that I'm curious about is where this data is stored.

[00:28:04] And I used to work at Pivotal and got to be part of the cloud wave.

[00:28:09] And what I got to see around decisions by people above my pay grade were making is let's move

[00:28:16] everything off the cloud.

[00:28:18] And I certainly had questions about, I wonder why everything is the answer here.

[00:28:22] But, you know, if it's everything is everything, I'm going to put my head down and follow the

[00:28:26] marching orders.

[00:28:27] And now we're seeing this pendulum swing back a little bit.

[00:28:30] Our executives that I'm talking to are saying, hey, the cost of capital is a lot higher with

[00:28:34] the interest rates.

[00:28:35] And that's kind of specific to this specific time.

[00:28:38] And then the cost of data is going up because there's just a lot more data that we're thinking

[00:28:43] about storing, not necessarily knowing what we're going to use it for in the future, but

[00:28:47] knowing that there's a power to extract data from our extract value rather from this data.

[00:28:54] And so maybe let's hoard a little bit more.

[00:28:56] And so this pendulum swinging back and forth around, do we continue using our in-house data

[00:29:01] centers or do we move everything to like the hyperscalers, like you mentioned of AWS,

[00:29:07] Google and Azure?

[00:29:10] So what do you think is going to happen in the next few years?

[00:29:13] And if you were an executive, how might you change your budgets or your priorities?

[00:29:18] So I do fractional CTO services.

[00:29:21] And this conversation is going on in literally every company that I work with right now.

[00:29:26] Every company.

[00:29:28] And, you know, the problem is, is that we see this is a very cyclical industry.

[00:29:33] We go in-sourced, out-sourced, in-sourced, out-sourced.

[00:29:36] We go on-site, off-site, on-site, off-site.

[00:29:38] The real answer is you need to evaluate every application for what it does for your company

[00:29:43] and make that decision accordingly.

[00:29:45] Moving stuff into the cloud was super, you know, it was cheap.

[00:29:48] It was easy.

[00:29:48] They'd send a little device to your place.

[00:29:50] You could, and actually we might start seeing that with AI too,

[00:29:53] where somebody sends a device to capture all your data,

[00:29:56] to turn it into a language model for you and bring it back.

[00:29:58] But the problem with data is moving it around in the cloud is very expensive.

[00:30:05] Egress fees are super expensive.

[00:30:07] That hurt a lot of clouds.

[00:30:08] And people and companies are trying to balance these budgets

[00:30:11] and the difference between CapEx and OpEx.

[00:30:14] But certainly when we talk about repatriating,

[00:30:17] there's a lot of assets that are already there.

[00:30:19] And companies that said, I don't want to be in the data center space.

[00:30:22] I'm going to put everything in the cloud.

[00:30:24] I'm going to get rid of all those people that worked in my data center.

[00:30:27] Now that data center sits there.

[00:30:29] Well, in a lot of cases,

[00:30:30] that company hasn't even fully depreciated those assets yet.

[00:30:34] So those assets are still on the books.

[00:30:36] They're still able to be functioning.

[00:30:37] And you can bring some of that load back.

[00:30:39] But I think what we have to do as consumers,

[00:30:43] because no two companies are alike, right?

[00:30:45] Even in the one data center,

[00:30:46] you might have certain applications that function better in the cloud than other ones.

[00:30:50] Homegrown applications, for instance, don't always function well in the cloud.

[00:30:54] But I think we have to get to the point where

[00:30:56] instead of making these wholesale sweeping things,

[00:30:59] because some CEO read a great article in PC Magazine or whatever.

[00:31:03] No.

[00:31:04] And it caught onto the bandwagon that we figure out what these applications do.

[00:31:08] And I think that that gives us more nimbleness in the future.

[00:31:11] Because then we can say, hey, this application was a great cloud-based application.

[00:31:15] Or, you know, but now we have shrunk and maybe we don't need it in the cloud.

[00:31:19] Or vice versa.

[00:31:20] This was a great application.

[00:31:21] But now we've grown and we've outgrown this application.

[00:31:24] We need to move it somewhere else.

[00:31:25] And I think you need to have that nimbleness and that ability to move things around.

[00:31:29] So whether you go to the cloud, whether you have a data center in a colo,

[00:31:34] whether you have an on-premise data center.

[00:31:36] And I think we're certainly going to see a lot more edge data centers popping up

[00:31:39] just for some of this crazy processing need.

[00:31:42] And because autonomous vehicles have to operate outside city limits, right?

[00:31:47] But I think we're going to see more and more data centers going a lot of different places.

[00:31:50] But I do think that the smartest thing for any company to do is evaluate their stack on an application basis.

[00:31:58] Absolutely.

[00:31:58] And as I'm not saying, I've got this list of call it for argument, say 100 applications,

[00:32:03] and they do different things.

[00:32:04] And ideally, my task is complete when I have a pretty good sense of like,

[00:32:09] which of these should go and continue to sit in a public cloud or in a hyperscaler,

[00:32:14] and which of those maybe we should move back into our in-house data center.

[00:32:18] What are some factors that come to mind that you would use and put different lenses on

[00:32:24] to determine which place you rehouse this application?

[00:32:29] So first off, I would think, how big is the data store?

[00:32:32] What are you moving around the cloud?

[00:32:33] Because movement in the cloud is where you gain most of your expenses.

[00:32:37] How much of that are you pulling out of the cloud?

[00:32:39] How much of that are you currently using?

[00:32:41] I mean, it's never a bad idea to go back three years after you put in an application

[00:32:44] and do, you know, go evaluate all the competitors for that application.

[00:32:49] What features do they have?

[00:32:51] What things do you want to add?

[00:32:52] Because it grows, right, over time.

[00:32:54] A company's needs, what they need out of that application is also going to grow and change.

[00:32:58] So I would look at input-output, look at your communications.

[00:33:02] Where are those going to be?

[00:33:03] Is your workforce predominantly local, where you could use local area networks instead of

[00:33:08] having to use wide area networks and pay egress fees?

[00:33:11] You know, where are those kind of things?

[00:33:14] How much is your equipment cost?

[00:33:16] Do you have the resources in-house?

[00:33:17] Or can you get resources in-house to do the support on the equipment, the software, all

[00:33:24] of that?

[00:33:25] I think you just really have to figure out all of the touch points for that application.

[00:33:29] Because there's, you know, think about some applications like a Salesforce, for instance.

[00:33:33] That's something that really everybody's familiar with, right?

[00:33:35] You can have Salesforce in the cloud.

[00:33:37] You can have your own data stores.

[00:33:40] You can have the two mixed together in a hybrid environment.

[00:33:42] All of that is going to depend on what you do for Salesforce.

[00:33:45] If I'm using Salesforce just to keep up with my customers, I'm probably not using any of

[00:33:50] that functionality.

[00:33:51] So for it to sit out there and everybody to access it remotely is probably fine.

[00:33:55] Now, if I'm building an inventory stack on that and I'm using CPQ and I'm building quoting

[00:34:00] systems into that, I may or may not want to do that on the cloud.

[00:34:04] I may have proprietary information that goes in there that feeds to those that tells me now,

[00:34:08] well, I don't want that in the cloud.

[00:34:10] I want this in my local data center.

[00:34:12] Or maybe just more people locally access that at the corporate office.

[00:34:16] And so it makes more sense to have it there because we're not continuously using up bandwidth

[00:34:20] going out to the cloud.

[00:34:22] So those are the kind of things, you know, does it have a video component with it?

[00:34:25] Does it have a call-based component with it?

[00:34:27] All of those, you know, that affect how that application works is going to matter when you

[00:34:31] make those decisions.

[00:34:33] Yeah, I think that's super helpful in terms of developing some kind of a framework to help

[00:34:38] guide that or a scoring system.

[00:34:41] Because 200 applications is a small number that a lot of CIOs are faced with determining.

[00:34:48] And it's hard to figure out where do you take the first bite to eventually eat the full

[00:34:53] elephant.

[00:34:54] And I think those are pretty helpful ways to start considering to determine.

[00:34:58] So what is an AI application that recently was really impressive for you?

[00:35:07] For me, the most impressive one I've seen in recent times is one called Perplexity.

[00:35:13] And it's very much like ChatGBT.

[00:35:15] So it's a solution set.

[00:35:17] But the one thing that I really, really enjoy about this is that it lists every sighted instance.

[00:35:22] Because, you know, if you in some AIs, if you start checking sighted instances,

[00:35:27] you'll find out some of them are bogus.

[00:35:29] Because if that large language model is picking up bogus information, it's repeating bogus

[00:35:34] information.

[00:35:35] I do like the fact that, especially if I'm using for research on articles and white papers

[00:35:39] and a lot of technical writing that I do, the nice thing is to be able to go check those

[00:35:44] sources.

[00:35:45] And then even the dates and times of those sources, it might have been accurate at one

[00:35:49] point, but it's not necessarily accurate for today.

[00:35:53] So that one, to me, I was so appreciative that it took the time to actually quote those.

[00:35:57] I thought that was great.

[00:35:58] So hallucinations, I think, are going to be a source for just fun stories that we hear

[00:36:05] and probably some level of distress.

[00:36:06] I recently asked ChatGBT to provide hyperlinks to the sites, and it just highlighted the

[00:36:12] same text in blue, but didn't exist.

[00:36:15] When I copied and pasted that into Google, I couldn't find those articles anywhere.

[00:36:18] Well, at least you weren't that attorney that did that for his court briefings.

[00:36:23] Yes.

[00:36:24] Yes.

[00:36:24] That person definitely acted as a scapegoat for the rest of us around what we use it

[00:36:29] for and what we don't.

[00:36:30] So, and what a great reminder that critical thinking is a valued component in our lives.

[00:36:36] 100%.

[00:36:37] 100%.

[00:36:37] And if you had maybe, you know, a niece, nephew, someone in your neighborhood that was

[00:36:44] going into high school and starting to develop their own opinions on stuff, what, how do you

[00:36:49] go about learning critical thinking?

[00:36:51] I got really lucky because one of my electives that you talked about was in critical thinking

[00:36:56] in my engineering school.

[00:36:57] We had very few.

[00:36:58] Full disclosure, I had a crush on a woman that was taking that class, and she told me

[00:37:04] that she should, I should do it too, and that was the reason I chose it.

[00:37:08] And I don't remember the last time I talked to her, but I definitely remember the last

[00:37:13] time I used a lot of the fallacy-based thinking and everything else.

[00:37:17] What is your advice for kids who want to develop stronger critical thinking or parents who want

[00:37:22] that for their kids?

[00:37:23] I feel like critical thinking and speed reading should be two required courses in junior high.

[00:37:28] I took speed reading in junior high.

[00:37:30] It was, I've used it my whole life.

[00:37:32] What a great thing.

[00:37:33] But critical thinking, I think you have to figure out what the adjectives and adverbs

[00:37:38] are and sort out fact and feeling words.

[00:37:41] And when you start getting out the feeling words, then you can start getting more to

[00:37:45] the fact.

[00:37:46] And then you also have to be able to have a paper trail.

[00:37:48] It has to be defensible.

[00:37:49] And just like, you know, for a legal proceeding or anything else, you've got to be able to

[00:37:54] show where that came from.

[00:37:56] In math.

[00:37:57] They make you show your work.

[00:37:58] I think we need to start doing the same thing when we start citing references.

[00:38:02] It was a lot easier when you went to the library and you could sort a library book because that's

[00:38:07] sort of been vetted for us.

[00:38:08] The stuff on the internet gets vetted as it goes.

[00:38:12] And even, you know, there's things that end up coming out five or six years later that

[00:38:16] turn out to be untrue or misleading.

[00:38:19] You know, a lot of this stuff at COVID was a great example.

[00:38:22] Look how much we've learned now that was completely filtered, shied away from us.

[00:38:27] So I think that, you know, you should call BS on every statement and follow your gut.

[00:38:33] Like sometimes your gut will just tell you that something smells a little off on that.

[00:38:39] I love it.

[00:38:39] So it's separating the subjective from the objective when you hear things, trusting your

[00:38:45] gut, finding the sources, finding the paper trail.

[00:38:50] And if I may add a fourth one that's helped me a lot in this, in the, you know, sort of

[00:38:54] capitalistic world that we live in, finding the money trail is another one to sort of understand

[00:39:00] where the incentives are, who's actually paying for this?

[00:39:03] Why does it behoove them to pay for it?

[00:39:06] And that can really inform a lot of further questions to help with that.

[00:39:10] Oh, for sure.

[00:39:11] That's a great one.

[00:39:12] Thank you so much for making time for us today.

[00:39:15] Absolutely.

[00:39:16] How can folks get ahold of you?

[00:39:18] How can we find your books?

[00:39:20] So the books are all on Amazon and Barnes and Noble, or you can go to kerrygetts.com.

[00:39:24] There's a link there to all of them.

[00:39:26] It's also on Audible if you prefer to listen.

[00:39:28] And if you have questions about any of them, please reach out.

[00:39:30] If you need somebody to dial in and talk to your classroom, I do that all the time.

[00:39:34] I'm happy to do that.

[00:39:35] Just explain the industry.

[00:39:37] And yeah.

[00:39:38] Cool.

[00:39:39] And we'll have those in the show notes.

[00:39:40] And if folks are interested in having you speak to them or getting you on as a fractional

[00:39:47] CTO, is kerrygetts.com the best spot for them to start?

[00:39:50] That'll do it.

[00:39:51] Yep.

[00:39:52] Or LinkedIn.

[00:39:52] Either one.

[00:39:53] Sounds good.

[00:39:54] We'll make sure to have links to all of that.

[00:39:56] Thanks a lot.

[00:39:57] And we'll catch everyone next Tuesday on the next episode.

[00:40:06] Thank you for joining me on the Convergence Podcast today.

[00:40:09] Subscribe to the Convergence Podcast on Apple Podcast, Spotify, YouTube, or wherever you

[00:40:16] get your content.

[00:40:17] If you're listening and found this helpful, please give us a five-star review.

[00:40:21] And if you're watching on YouTube, hit that like button and tell me what you think about

[00:40:26] what you heard today.