Navigating AI Adoption: Insights, Regulations, and Data Center Expansion

[00:00:13] Let's kick off the year with AI. Where is adoption? What regulations are coming? And what is the impact of data center expansion? Esther Shethu and TC Doyle join me, plus your questions, as we consider how to guide customers. Welcome to the Business of Tech Lounge, the live version of the Business of Tech Podcast. It's Wednesday, January 8th, 2025, and I'm Dave Sobel.

[00:00:39] I'll be taking questions and comments throughout the show, so make sure to put them in chat. If you have a question, we will happily respond to it.

[00:00:49] Now I want to thank Sales Builder, our Patreon sponsor whose support makes this show possible. Focus on your IT sales workflow with the power of automation and visit them at salesbuilder.com. That's B-U-I-L-D-R dot com.

[00:01:03] Now, reminder, I am watching that chat and we'll keep an eye out for your questions and answer them over the course of the show.

[00:01:11] Now Esther Shethu is a news writer at TechTarget focusing on artificial intelligence. She's contributed to publications such as the New York Daily News and the Brooklyn Daily Eagle.

[00:01:22] She holds dual degrees in English creative writing and film studies from Trinity College and a journalism degree from the SUNY Graduate School of Journalism.

[00:01:32] Welcome to the show, Esther.

[00:01:35] Thank you so much for having me.

[00:01:37] I'm excited to talk to you and also joining us today, TC Doyle is a seasoned journalist and editor with expansive experience covering the IT services and managed services provider sectors.

[00:01:48] He's held several senior editorial positions at various industry publications, providing insights into channel strategies, market trends, and technological advancements.

[00:01:58] TC, welcome back to the show.

[00:02:00] Great to be here, Esther.

[00:02:02] Lovely to see you.

[00:02:03] Hi, TC.

[00:02:04] So I am super excited to have you both here as we sort of kick off the year.

[00:02:08] And in a way, I want to give this as a bit of a baseline of where we're at and where we think some of the impact might be.

[00:02:13] First, I wanted to talk a little bit about where we are in terms of enterprise and business adoption.

[00:02:19] Now, Esther, you literally just covered this on your beat.

[00:02:22] And I wanted to get a little bit of your insight in what your reporting has given us to where we are on business and enterprise adoption.

[00:02:31] Yeah, I think that we are kind of past that stage where, you know, where there was that much excitement. Right.

[00:02:40] We are now in the stage where businesses are trying to figure out what to do with this.

[00:02:44] We have some vendors who are refining their products.

[00:02:49] And then we are moving into a lot of not new technologies with things like agents, but like, you know, there's now excitement around agents as well.

[00:02:59] And I think that's the buzz that we're all looking for.

[00:03:02] In fact, I was just working on stuff for today's news podcast and the U.S.

[00:03:06] government's latest audit of their projects in flight showed more than seventeen hundred different different implementations at various stages across the federal government.

[00:03:16] Now, TC, one of the things that I've been thinking about a lot is a quote from Forrester.

[00:03:21] I kind of wanted to get your opinion on because Forrester was analyzing that roughly half of the spend that they're expecting around A.I. and generative A.I. spend comes on the consumer side.

[00:03:33] And the other half is API, which would lean into either product development or enterprise A.I.

[00:03:41] Like, am I how am I reading that?

[00:03:44] Well, and what does that speak to kind of the solution provider opportunity?

[00:03:48] Because theoretically might be on both sides of that.

[00:03:51] So a couple of things to unpack.

[00:03:55] Everyone's talking about A.I.

[00:03:57] as if it's a monolithic block.

[00:03:59] We have different forms of A.I.

[00:04:02] So let's go to your point.

[00:04:06] Originally, getting started with A.I. can be daunting.

[00:04:09] In fact, 26% of organizations that have developed the capacity to go beyond proof of concepts and generate tangible value for their organization,

[00:04:22] they have no measurable impact.

[00:04:26] And just 4% of organizations say that they're getting benefits from cutting edge A.I.

[00:04:32] Well, we know that that's nonsense.

[00:04:35] Individual organizations and entities within corporations, within silos, are giving immense benefit from A.I.

[00:04:44] I think to answer your question, though, directly is you have to measure where they're applying it, how they're applying it, and what they want to do with it.

[00:04:54] If you can track that, then you really see how A.I. can improve organizational impact.

[00:05:01] And that's imperative.

[00:05:03] Now, Esther, when you're talking to these very, you know, particularly when you're talking to enterprise customers, mid-market customers or small businesses,

[00:05:11] give me a little bit of sense of where the use cases are that are coming to the forefront most of the time.

[00:05:17] Yeah, I would say, like, we're still seeing the use case be like maybe customer service, you know, the call center.

[00:05:23] That's where the use cases is.

[00:05:25] Content generation is where, especially when you're talking about, you know, generative A.I., that's where the use cases is.

[00:05:31] I think that we are trying to get beyond those general use cases because at the beginning, right, we were promises everything in the kitchen sink.

[00:05:40] Where it's like, you know, we're going to have personal A.I. assistance doing this and that for you.

[00:05:45] We haven't quite gotten there yet, but we're still the general use cases, but we're kind of starting to see it, especially if you watch NVIDIA keynote at CES.

[00:05:55] As it seems like we are pushing there, just not quite there with, they're still promising a lot is what I would say, but hope to see the implementation soon.

[00:06:06] Well, let's pause there a little bit.

[00:06:08] So by the way, for my listeners, I'll be doing a whole CES breakdown on tomorrow's show.

[00:06:12] I have been taking the tact of letting them do the announcements and then that way I can sink in and think about it and make sure they make sense.

[00:06:19] But one of the things that I think is important for us to talk about is we don't want to rush through this stage based on promises of the future.

[00:06:28] If what they're talking about now isn't delivering, then that's important to highlight.

[00:06:33] But at the same time, I think there's some value here.

[00:06:35] And what I'm trying to unpack is, you know, what are the value that we're seeing based on the past year of implementation?

[00:06:41] What is the tangible business outcomes that people are seeing?

[00:06:46] I know in some of the conversations that I've had with customers that there is value in summarization capability, in the ability to process large information.

[00:06:57] And one of the things that I'm seeing useful implementations and use cases are is around what I call the reporting problem.

[00:07:04] And what I mean by that is, is we've oftentimes had difficulty extracting information out of large databases.

[00:07:11] Either they have to build, they get locked into pre-canned reports that don't necessarily give them the information they want,

[00:07:18] or they have to then use a complicated report builder to learn a new reporting language to get information out of it.

[00:07:26] But the ability to apply generative AI here actually does allow for that.

[00:07:31] But I want to make sure that we're actually getting enough value out of this before moving on to the next hype cycle.

[00:07:36] TC, how are you looking at this in terms of measuring whether or not we're actually getting good ROI on the stuff that's in the field now?

[00:07:45] The look at the use cases that Esther talked about, you've talked about, those are imperative.

[00:07:50] So what we're finding is AI is evolving and we're getting past generative AI, which requires a lot of human input every time to exact.

[00:08:03] But we're going to get to agentic AI, which is the big buzzword for 2025.

[00:08:07] And that's going to open up a whole new class of use cases.

[00:08:13] What do I mean by agentic AI?

[00:08:15] Agenic AI can use large language models to act autonomously on behalf of humans to exact specific things.

[00:08:24] Think about go-to-market strategies, sales strategies in particular, marketing strategy.

[00:08:29] It can conduct campaigns.

[00:08:31] It can isolate and identify user profiles.

[00:08:36] It can fire up email systems.

[00:08:40] It can warm up emails so they don't get blocked.

[00:08:43] All of this can be done in background, autonomously, monitored by humans.

[00:08:48] But when you start thinking about the use cases, whether it's an interactive, you know, bot that helps you, you know, book an airplane or book a trip, or whether it's something that helps you launch a campaign.

[00:09:02] We are starting to see the use cases explode in terms of these new agents.

[00:09:09] Microsoft has them.

[00:09:11] Microsoft has them.

[00:09:11] Salesforce has them.

[00:09:13] Landbase has them.

[00:09:15] Google has them.

[00:09:17] This is a whole new world.

[00:09:19] And this, I think, will take root in 2025.

[00:09:23] So I'm going to bounce that off of you and say, like, in terms of reporting, how much of what you're seeing out there in agentic AI is happening versus is being promised, particularly thinking about in the context of NVIDIA and their keynotes recently and Salesforce.

[00:09:39] Yeah, I think that in terms of agentic AI, I would say I don't think a lot of it is being delivered quite yet.

[00:09:48] I think it's nice that you can maybe have an AI agent do coding for you or all of this.

[00:09:54] And I still think that part of it is that there's confusion about what exactly is AI agents.

[00:09:59] That's why whenever I speak to like analysts or even vendors, I'm always asking, what exactly is an AI agent to you?

[00:10:07] Right. Because at first we had this confusion with Microsoft Co-Pilot.

[00:10:10] What is it? A Co-Pilot?

[00:10:11] Is it not a Co-Pilot?

[00:10:13] Now we have AI agents.

[00:10:16] And what I would say on the one hand of it is what is the very much like having robots do stuff for you.

[00:10:24] And right now we're just now seeing like, oh, we are having large language models drive these agents to perform tax.

[00:10:32] Right. We're still at the very beginning stages in my opinion.

[00:10:35] I want to watch for this.

[00:10:37] I think it's interesting.

[00:10:38] Sometimes we're not always aware of the use cases that are happening in the field because that's their competitive advantage.

[00:10:43] But at the same time, when they're having success, a lot of companies want to brag about it because they want more of that.

[00:10:47] Now, I want to want to pivot us particularly because you've done some recorded reporting around some of the regulation that's coming here, because, you know, this is a space that is moving quickly.

[00:10:57] But at the same time, you know, regulators are starting to be concerned about some of the potential unexpected outcomes.

[00:11:05] On top of that, I think regulators are also looking at how they can apply existing laws and existing frameworks to this existing piece.

[00:11:15] Give me a little bit of the breakdown of what you've been reporting on this so far.

[00:11:17] Yeah. When I talk, I don't really do a much reporting on regulation.

[00:11:21] We have Mackenzie Holland, who does a lot of that.

[00:11:24] But when it comes to AI, the I guess the main thing is the fact that there isn't that much regulation.

[00:11:32] And we don't really expect much regulation in this year following the shift in government.

[00:11:39] But yeah, a lot of people always say that regulation takes us take slowly to follow technology.

[00:11:46] And that's what we continue to see with AI.

[00:11:49] Obviously, we have the EU AI Act.

[00:11:54] But other than that, it's just continuing to be slow.

[00:11:58] Obviously, we also have states taking things into their own hands.

[00:12:02] And a lot of people expect that that will be what happens in this year.

[00:12:07] So TC, I'm dying to get a little bit of a thought on the business opportunity here,

[00:12:10] because one of the things I find interesting is, of course, in compliance is the fact that you can create entire businesses out of the fact that there is complexity in regulation.

[00:12:21] If that isn't going to happen in AI in a way, are we potentially holding back a portion of the service provider market?

[00:12:27] Or is an open free-for-all kind of a good thing to look for in 25?

[00:12:33] I'm going to reframe that, Dave, if I may.

[00:12:38] Okay.

[00:12:39] Regulation will always lag innovation.

[00:12:42] We saw it when Uber and others came in to fore.

[00:12:47] And they took the taxi regulation, or whether it's Airbnb, and they just destroyed governance and regulation.

[00:12:56] So I think the fact that we need to respect that innovation is going to lead regulation is just baked into the situation.

[00:13:07] However, and Esther has done some writing and I've read on some of her stuff.

[00:13:13] You get into a situation where you need to have some guardrails.

[00:13:19] We know this.

[00:13:20] And everyone's afraid we'll, whether it's generative or genic AI, will it turn off grid systems?

[00:13:27] Will it, you know, launch, you know, prevent planes from landing?

[00:13:31] We do need guardrails in this regard.

[00:13:34] So I do think some forward thinking legislators are going to come to fore.

[00:13:40] But it's going to be a lag.

[00:13:41] And in fact, in March, I'm starting with a test your memory.

[00:13:46] In March of 2023, you've talked about a trio of leaders, including Elon Musk, that came together and said we should have a six months moratorium on what's happening.

[00:13:56] Not only did the industry ignore it, those three individuals that were part of that article that she wrote ignored it.

[00:14:04] And so innovators are going to continue to go downhill.

[00:14:08] And what we need is to have the ability to say, OK, is this going to be dislocated to humans?

[00:14:15] Is this going to be dislocated to societal mechanisms, grid systems, energy, what have you?

[00:14:22] And I think we just need to rein that in as the best we can.

[00:14:26] But the net net, I'll go back to it, is innovators are going to run ahead.

[00:14:32] Well, TC, you brought this up.

[00:14:34] So I'm going to ask Esther directly.

[00:14:35] Esther, you did coverage around kind of that call to slow down.

[00:14:40] Clearly, we're not hearing that anymore.

[00:14:42] Give me a little bit of your sense of like which of the leaders shifted their conversation over the past two years and which have been more consistent as you reported on.

[00:14:52] I would say shifted in terms of slowing down.

[00:14:55] No one, no one slowed down.

[00:14:57] Right.

[00:14:58] Absolutely.

[00:14:59] No one.

[00:15:00] So that I mean, I think that you have a lot of the leaders, as you saw last year, went to the White House, had like a lot of conversations with the White House concerning AI.

[00:15:11] And regulation.

[00:15:13] But I don't think I see any benefit of that.

[00:15:16] And as you can, as you know, Elon Musk has definitely calls for more open AI.

[00:15:21] But then he has XAI and his company.

[00:15:25] And there are arguments about whether that is open or not.

[00:15:28] I would say that one thing that we did see last year is this like argument for more openness and more open source.

[00:15:35] And I think that came from all around the industry.

[00:15:37] And there are even questions as to whether MetaLama is open.

[00:15:41] And so I think the conversation is on the current, whether it will lead to regulation.

[00:15:47] I think that we might have to wait a couple of years to see.

[00:15:51] Well, I think that's something we're going to keep an eye on.

[00:15:53] And a reminder for those of you that are watching, if you've got a question, drop it into chat.

[00:15:57] We will definitely address it here.

[00:15:59] Now, for both of you, I want to kind of bring up something else that I've been thinking about.

[00:16:03] And I want to get kind of your perspectives a little bit on it, particularly because I've been thinking about the impact of all of the data center builds that and the energy demands that we're starting to see come out there.

[00:16:15] We're seeing, you know, Microsoft trying to gear up nuclear fusion reactors for a three mile island.

[00:16:22] We're seeing Google and Amazon have conversations about building up there.

[00:16:26] Microsoft just announced $80 million of new investments in data centers.

[00:16:30] It's interesting to me then to note that the model makers themselves aren't making any money.

[00:16:36] The more that they put, you know, Anthropic, OpenAI, not profitable yet.

[00:16:41] And the cloud providers who are making money are hungry for more data centers.

[00:16:45] But the other undercurrent of this is, as reported both by the information and by Business Insider recently, is that this data center hunger is starting to cause real strains on the grid.

[00:16:56] And that may cause costs to get pushed out to end customers.

[00:17:01] And as I've been thinking about this, I see the potential for end customers to get what I'm calling the double whammy of AI costs, where the model makers are going to need to figure out a way to make money.

[00:17:14] And data centers are going to cause strain on energy demands, causing customers to pay more money for energy.

[00:17:22] So in fact, you may be in a situation where AI is causing rising costs on both sides.

[00:17:27] So I'm going to first sort of ask you, tell me a little bit about what the vendors in this space are talking about costs.

[00:17:36] I think that when it comes to costs, it's something that some analysts might even bring up in terms of like, how is this going to make ROI?

[00:17:46] As you mentioned, there isn't like a lot of ROI yet for the creators, like the foundation models creators, that is.

[00:17:53] But we're still seeing a lot of investment.

[00:17:56] I think right now, it's more of a let's push everything out there and see what brings in ROI at the final level.

[00:18:06] Ah, the old throw spaghetti at the wall and see what sticks right.

[00:18:09] Yeah, basically.

[00:18:10] So TC, I'm going to then ask you.

[00:18:12] If I can jump in.

[00:18:15] Please.

[00:18:15] I think what she says is brilliant.

[00:18:18] We're starting to see individual states.

[00:18:22] Data centers are not clean businesses.

[00:18:24] They take up a shit ton of...

[00:18:27] You're allowed on this show.

[00:18:28] Go for it.

[00:18:29] They take up a lot of energy.

[00:18:31] Water, heat, electricity are off the charts usage.

[00:18:37] And you take two simultaneous trends.

[00:18:40] Crypto, AI, generative demands.

[00:18:44] They're off the charts.

[00:18:45] These are not clean businesses.

[00:18:47] They take up a lot.

[00:18:48] And it's great to have them on shore, theoretically.

[00:18:51] But, you know, you check with, you know, communities that are dominated by these businesses.

[00:18:56] It is not a real happy place to have this demand.

[00:19:01] However, to Esther's point, the demand for these are off the charts.

[00:19:06] They're run away.

[00:19:07] It's just going to be endless, endless, endless, endless.

[00:19:09] And communities around the country, there's going to be regulation.

[00:19:12] There's going to have to be state laws.

[00:19:15] You know, who owns the data?

[00:19:17] Who owns the processing power?

[00:19:19] Wow.

[00:19:20] There's just all kinds of implications that have to be discussed and eventually delivered.

[00:19:27] So my question is, yeah, go ahead, Esther.

[00:19:29] And I was going to say, I think that the cost is more than just money, right?

[00:19:32] The cost is also like you mentioned, CC data, right?

[00:19:35] We are having, I have conversations a lot about what is this idea of data ownership, right?

[00:19:41] And, you know, with AI, there is that idea of like, where are you getting your data from?

[00:19:46] Who's sourcing it?

[00:19:47] So I think when it comes to the end customer and when it comes to the consumer, not only are

[00:19:51] you being hit with, you're being hit when it comes to your pocket, but you're also being

[00:19:56] hit when it comes to your data and the data that you own as well.

[00:20:00] So I think she's pivoting into a really interesting branch we could discuss is, let's be honest,

[00:20:08] when you put stuff into chat, the GPT, you upload an, Esther, I'm, I don't know, but I'm guessing

[00:20:13] that on occasion you'll throw an article and say, Hey, could I reword this?

[00:20:18] I do not.

[00:20:20] For the record.

[00:20:21] I do, so I'll take time for that.

[00:20:23] You know, content marketers are, and all of that is getting scooped up.

[00:20:29] And chat GPT and other AI agents, it is the biggest intellectual property ripoff of all

[00:20:37] time.

[00:20:37] Who owns it?

[00:20:38] If you put it in there, if somebody else that you send it to and they upload it now,

[00:20:43] who owns that data?

[00:20:45] There are all kinds of ramifications and Esther, I don't know how much of reporting you've done

[00:20:50] in intellectual property discussions, but do you have any thoughts on that?

[00:20:56] Yeah.

[00:20:56] I mean, that's, that's actually something that I'm pretty interested in, in terms of like ownership,

[00:21:02] right?

[00:21:02] I think that obviously we record, we report and tech target on the business side, but I think

[00:21:08] when it comes to the ownership of like the data of the consumer, I think that definitely

[00:21:13] needs to be a little bit more guardrails and more regular regulation as we were speaking

[00:21:19] about before, because it's everyone, everyone, no one is protected.

[00:21:23] Right?

[00:21:24] And so a lot of times last year you have, you know, whether it's open AI or in Tropic being

[00:21:30] like, well, if you read the fine print, you can opt out of us using our data.

[00:21:35] And it's like, who's reading that?

[00:21:38] Who is actually going to say, I'm reading the fine print that's going to make me opt out.

[00:21:42] And I think there needs to be a lot more, a lot more protection for people regarding that.

[00:21:49] So I'm nerdy enough to have been the person that flipped that switch.

[00:21:53] Cause I did do, did the read that because, but, but actually, but, but what we're hitting

[00:21:57] on exactly what I think the core of this conversation is.

[00:22:00] And I want to sort of reframe it again to get both of your actions, because there's a subtlety

[00:22:04] as certain, the way that you talked about it, and you talked about consumer.

[00:22:07] And I like to think also about customer.

[00:22:09] And what I'm differentiating here is consumer is kind of the mass public that is buying something

[00:22:15] like the user of chat GPD.

[00:22:17] That's an individual user generally using it for personal use.

[00:22:20] There's also what I'm thinking of as customers, which is end companies.

[00:22:25] Typically by the way, small and mid-sized businesses, the ones that I like to think about most, which

[00:22:29] are the ones that are using this more at a company, trying to implement these technologies

[00:22:34] at a company level.

[00:22:35] And what I'm increasingly concerned about is, is kind of this, this now even triple whammy

[00:22:40] of, Hey, how is their data actually being protected?

[00:22:44] Oh, and Hey, how are they going to be impacted by increasing costs of the AI technology itself

[00:22:50] as the companies look to become profitable?

[00:22:53] Oh, and Hey, how are they going to be impacted by energy demands as the grid gets strained?

[00:23:00] And I think we want, if we're putting on our, we are technology advisors, as if you're thinking

[00:23:07] about being an MSP and thinking about a technology solutions provider.

[00:23:11] These are questions that we want to bring up to customers because while I think intuitively,

[00:23:17] I mean, said there are use cases of value from the business perspective.

[00:23:20] I want to make sure we're also measuring the costs completely so that when we measure return

[00:23:26] on investment, it's that.

[00:23:28] And I guess I want to turn around and say like, am I overloading this or am I actually highlighting

[00:23:32] the right levels of risk?

[00:23:35] PC, I'll kick to you.

[00:23:37] Yeah.

[00:23:38] So Esther, Dave and I come from the world of intermediaries, these third party companies

[00:23:44] that are MSPs, VARs, resellers, these organizations that sell technology to end customers.

[00:23:53] And in our years of doing this and familiar, if we're familiar rating ourselves with this audience,

[00:24:00] we know a couple of things.

[00:24:01] One, they don't give a rat's ass about the energy cost.

[00:24:04] That is too downstream.

[00:24:06] Two, we're going to leverage this technology to help generate more leads for their business.

[00:24:14] That's where they're going to, can we use AI to lower our costs through chat bots and what

[00:24:21] have you, and use it to make more leads for our business.

[00:24:26] They're also going to want to sell those services downstream to their customers.

[00:24:30] So they're going to be both users of this technology and also facilitators of this technology.

[00:24:36] So we have that mechanism going on.

[00:24:40] But Dave, I think it's way too highfalutin to think that these guys are going to care about

[00:24:47] energy costs and things like that.

[00:24:50] They don't have the financial models to include that into their businesses.

[00:24:53] That being said, all of the things that I'm going to kick it off to you, Esther, that she said about privacy,

[00:25:00] security, data rights, intellectual property.

[00:25:04] That is something that is going to be that the entire industry, let alone the intermediaries that deliver technology solutions

[00:25:11] and customers are going to have to deal with.

[00:25:14] But Esther.

[00:25:14] Well, I'm not sure if I have much thought, but as you were speaking, I'm wondering like if they don't care about energy, then they probably won't care about, you know, the small, medium sized business.

[00:25:27] Will they?

[00:25:27] They do care about their customers and they do care about their customer needs.

[00:25:31] But I think thinking about downstream energy concerns, kind of like, who am I going to vote for in 2028?

[00:25:39] Yeah, it's kind of fuzzy and hazy, but they don't.

[00:25:43] Well, maybe, I guess what I'm, let me just, let's play this out a little bit, because I like the idea of the debate to see if I'm wrong on this one.

[00:25:51] I'm fully comfortable being wrong.

[00:25:53] But I like what I said, if we get to the point where energy costs broadly jump, say 20 or 30%, which is potentially possible by straining the grid at the level it wants,

[00:26:05] that goes into all supply chain costs, that goes into direct energy consumption.

[00:26:11] And I'm focusing less on the MSP and more on their customers.

[00:26:15] If energy costs broadly jump by a significant amount by strains on the grid, are we looking at a situation where people start caring about energy costs?

[00:26:27] There is no mechanism on earth where a small business, a medium sized corporation or Fortune 500 company can suddenly take a jump in operating costs at 20%.

[00:26:40] Right.

[00:26:41] That will, ears will prick up at that point.

[00:26:45] Yeah.

[00:26:46] But I, and then, so then what is the mechanism to rein that in?

[00:26:50] Do we regulate, you know, crypto farms that are just, you know, soaking up, you know, grid energy off the charts or AI at that point, then yes, 20%.

[00:27:02] Yeah.

[00:27:03] People will respond.

[00:27:04] Okay.

[00:27:04] So cool.

[00:27:05] So, so then let's, let's play a little bit of a game for both of you.

[00:27:08] Like, so we know zero is no reaction.

[00:27:10] We know 20% is, Oh my God, it is.

[00:27:12] Can we, can we give me a, give me a sense of where you might think the tipping point is?

[00:27:16] Does it, is it breaking 10% people care?

[00:27:19] Mr. And I are going to hold up two fingers.

[00:27:21] I'm going to go three, two, one.

[00:27:25] I say six.

[00:27:27] She says five.

[00:27:27] Yeah.

[00:27:28] Okay.

[00:27:29] All right, cool.

[00:27:29] So that's, so that's actually, so, so the usefulness of this conversation is to actually say like, Hey, one of the things we're looking for is potential rises in energy costs to know that that is a portion of this because most of the infrastructure investments required to keep up with energy for the data center expansion will lag significantly.

[00:27:49] The data center expansion itself.

[00:27:52] So it's important actually, if you both land around the same place, we as advisors to small businesses and medium sized organizations, Hey, we need to keep an eye on energy costs rising higher than say five or 6%.

[00:28:04] That's going to be a strain on customer spend.

[00:28:07] That's a really useful outcome of this discussion.

[00:28:12] I think we're absolutely.

[00:28:13] Cool.

[00:28:14] So the question then is, is, is, is okay, then I'm going to sort of wrap this little piece by saying like, okay, I'm thinking about this as a possibility.

[00:28:21] Am I thinking it's coming faster than it probably is?

[00:28:24] Is this, is this even on the radar for 25?

[00:28:29] I wouldn't say it is.

[00:28:31] Okay.

[00:28:32] Awesome.

[00:28:33] TC.

[00:28:33] No.

[00:28:34] Yes.

[00:28:34] I think she's, she's a hundred percent.

[00:28:37] Yeah.

[00:28:37] Awesome.

[00:28:38] Then for listeners, I'll just keep an eye on it and warn you.

[00:28:41] This has been cool.

[00:28:42] So let me turn around then and say like, what is that item for each of you that you're keeping an eye on for a real trend that you're interested to watch over the next, you know, 12 to 18 months.

[00:28:54] How about S or you what you're, you're on the beat watching it.

[00:28:56] What are you watching?

[00:28:57] I think I want to see how, is that what everyone's talking about?

[00:29:00] I want to see how agents is shaping for it.

[00:29:03] And I think this might surprise you, but the way I want to see what agents brings about, right?

[00:29:10] Are we going back to perhaps the meta brace technology?

[00:29:15] You know, I, I want to see if we're going to go into that as well.

[00:29:19] I know it's like, I feel like it's out, but my, my instincts tells me that we might be.

[00:29:24] So.

[00:29:25] TC, what's the one you're watching this year?

[00:29:27] It leverages what she just said.

[00:29:30] It's agentic AI, I think is going to have a breakout year in 2025.

[00:29:35] And people don't understand right now today, less than 1% of enterprise applications are agentic AI centric.

[00:29:44] By 2028, third of them will be.

[00:29:47] So what does that mean?

[00:29:48] Well, generative AI requires input.

[00:29:51] You and I use chat GPT.

[00:29:53] We have to type in our prompts.

[00:29:55] We have to type in our information and then we get stuff back.

[00:29:59] Genegade and autonomously do a lot of what we are trying to do.

[00:30:05] And this is the stuff that gets scary.

[00:30:08] It is the Waymo cars.

[00:30:10] It is the interactive bots that we interact with, but it leverages large language models, LLMs.

[00:30:18] And in a way that is going to be, I think, not only disruptive, but productive.

[00:30:24] So think of it, just go to market strategies.

[00:30:28] Agenic AI can handle email server setup, email automation, warming up.

[00:30:36] It can create automatic campaigns for somebody.

[00:30:39] It just goes on and on about the capabilities.

[00:30:42] That, I think, is what we're going to have to look at in 2025 and beyond.

[00:30:47] And this will then lead to some of the ethical, ecological, environmental, and other conversations that we have.

[00:30:56] So net net, you know, recapping.

[00:30:58] I think, I think, I think agenic AI is going to be the thing that people need to know beyond generative AI, which has blown everybody's mind for the last two years.

[00:31:07] We'll certainly have to keep an eye out there on this because I think it's going to be interesting to see how much of it actually is delivered on.

[00:31:12] Esther, TC, thank you both for joining me.

[00:31:14] Listeners can connect with both of you on LinkedIn.

[00:31:17] Is that the best place to find you?

[00:31:19] Yes.

[00:31:21] Awesome.

[00:31:21] Well, I look forward to having you both back.

[00:31:23] This has been great fun and topical in that I want to preview an interview that's on my YouTube channel already.

[00:31:28] As I talked to Danny Thompson, he's the CEO of Unique.

[00:31:32] We talked about agentic AI.

[00:31:34] He's talking about the transformative potential of AI powered digital humans in enhancing customer interactions and driving brand loyalty.

[00:31:42] We talked about how these emotionally intelligent avatars are going to impact customer service and impact marketing by providing personalized real time assistance that fosters trust and engagement.

[00:31:55] Here's a preview of that interview.

[00:31:56] A unique digital human.

[00:31:59] A unique digital human is an interactive avatar or an interactive character that exists in digital channels, whether it be a kiosk or whether it be a website, mobile app that people can talk to in real time and get assistance and all sorts of use cases.

[00:32:17] I would think that this applies really well in customer service or marketing.

[00:32:21] Well, I think one of the interesting areas this technology is really proving to be successful is just the nature of accessibility and natural interaction.

[00:32:33] With technology like digital humans and AI, now you can be human and the machine knows how to be more human.

[00:32:40] Digital humans are always on.

[00:32:41] And that's a really interesting dynamic with emotional intelligence, not just knowing what you say, but addressing how you feel and responding to make you feel more at ease.

[00:32:53] Well, this video is already out on both my YouTube and podcast feed.

[00:32:59] A lot of my content, it becomes available early for Patreon subscribers who now already have a presentation I put together around MSP success stories for artificial intelligence.

[00:33:11] I really encourage you to dive into both of those.

[00:33:14] And if you want early access to videos, visit patreon.com slash MSP radio.

[00:33:18] Now I want to thank sales builder, our Patreon sponsor whose support makes this show possible.

[00:33:23] Focus on your IT sales workflow with the power of automation and visit them at sales builder.com.

[00:33:29] That's B U I L D R.com.

[00:33:32] And vendors, you too can get your name mentioned on the live show.

[00:33:35] It's a simple monthly subscription.

[00:33:37] Visit patreon.com slash MSP radio for more.

[00:33:40] And listeners, here's how you can support the show.

[00:33:43] Like, share and follow on your favorite platforms.

[00:33:46] If you want to support directly on Patreon, use our give what you want model.

[00:33:51] You set what you think the content is worth and you get access to videos early.

[00:33:56] If you have a question and are listening to the recording, send it in at question at MSP radio.com.

[00:34:01] Thanks for joining me for the business of tech lounge.

[00:34:04] And I will see you next time.

[00:34:18] Bye.