Latest / Becker’s Healthcare Podcast / Preparing Healthcare Infrastructure for AI and Rising Power Demand

Transcript

Preparing Healthcare Infrastructure for AI and Rising Power Demand · Becker’s Healthcare Podcast

• 0:00
  - Hello, everyone. This is Erica Spicer Mason with
• 0:02
  - Becker's Healthcare.
• 0:04
  - Thank you so much for tuning into the
• 0:05
  - Becker's Healthcare podcast series.
• 0:08
  - Today, we're going to talk about what the
• 0:10
  - rapid growth of AI and digital infrastructure
• 0:13
  - means for health systems,
• 0:14
  - from power demand and grid constraints to long
• 0:17
  - term infrastructure planning.
• 0:19
  - And joining me for this conversation, we have
• 0:21
  - Greg Castle, The US power and grid commercial
• 0:24
  - leader at Schneider Electric,
• 0:26
  - and James VanCannon,
• 0:27
  - health care segment manager
• 0:29
  - for strategic accounts and US Digital Power of
• 0:32
  - Schneider Electric.
• 0:33
  - Greg, James, great to have you both with
• 0:35
  - us today. Thank you for being here. Yeah.
• 0:38
  - Very excited.
• 0:39
  - Absolutely. Glad to be here.
• 0:41
  - Well, glad to have you both on the
• 0:43
  - line. And let's go ahead and get started
• 0:45
  - with some introductions just to give our listeners
• 0:48
  - a little bit of context on the perspective
• 0:50
  - you're bringing to the conversation today.
• 0:52
  - Can you both share just a bit about
• 0:54
  - yourself, your role,
• 0:55
  - and anything else you think is important for
• 0:57
  - listeners to know? Greg, we can have you
• 0:59
  - get us started.
• 1:00
  - Very good. Yeah. Again, I appreciate the opportunity
• 1:03
  - to be here.
• 1:04
  - Greg Castle. So I lead our commercial activities
• 1:07
  - for our power and grid segment in North
• 1:09
  - America for Schneider Electric.
• 1:11
  - That was Schneider Electric for twelve years and
• 1:14
  - in this role for three.
• 1:16
  - Power and grid for Schneider is utilities, renewable
• 1:19
  - developers,
• 1:21
  - independent power producers,
• 1:23
  - and really in the last twelve and eighteen
• 1:25
  - months,
• 1:26
  - really a concerted focus on how we can
• 1:28
  - enable
• 1:29
  - speed to power
• 1:31
  - more for the data center build out that
• 1:33
  - that we're all aware of.
• 1:35
  - So I I guess the perspective I'm I'm
• 1:37
  - bringing is really the the front seat from
• 1:39
  - a utility and power generation aspect.
• 1:42
  - Fantastic.
• 1:43
  - Greg, thank you. And James, how about you?
• 1:46
  - So I'm the, strategic accounts, health care segment
• 1:49
  - manager for Digital Power.
• 1:51
  - I've been with the company for about thirty
• 1:52
  - years in various technical roles as well as
• 1:55
  - in sales roles.
• 1:57
  - My focus has really been in helping a
• 1:59
  - lot of the nation's largest healthcare systems
• 2:01
  - improve their, you know, power reliability,
• 2:04
  - their energy visibility,
• 2:06
  - and their overall operational resilience.
• 2:09
  - And from my perspective,
• 2:10
  - electricity has become,
• 2:12
  - one of those
• 2:13
  - critical things in health care is really like
• 2:15
  - water or oxygen.
• 2:17
  - And really, I see that every clinical system,
• 2:21
  - diagnostic platform, operating room, and now increasingly every
• 2:24
  - AI application
• 2:26
  - really depends on reliable power. So I'm really
• 2:28
  - excited to be able to talk about this
• 2:30
  - incredibly important topic.
• 2:32
  - Yeah. Yeah. Thanks, James.
• 2:34
  - And I'm excited to talk about this topic
• 2:36
  - too because we have many conversations about innovation
• 2:40
  - with AI in the health care space, but
• 2:43
  - not so many conversations about what this looks
• 2:46
  - like from a back end perspective, from a
• 2:48
  - power grid and electricity perspective. And I think
• 2:51
  - it's a really crucial piece, James, as you
• 2:53
  - just noted. So let's go ahead and and
• 2:55
  - start with kind of what the scene is
• 2:57
  - looking like right now.
• 2:59
  - Greg, I know that data center growth has
• 3:01
  - become a real factor in grid planning in
• 3:04
  - the last few years with the growth of
• 3:06
  - AI use.
• 3:07
  - So can you set the stage for us
• 3:09
  - and tell us a little bit about what's
• 3:10
  - driving that demand and and where it's showing
• 3:13
  - up most acutely?
• 3:15
  - Yeah. Absolutely.
• 3:16
  - What a wild ride it's been over the
• 3:18
  - last twelve to eighteen months with the, the
• 3:20
  - introduction and kind of boom of,
• 3:22
  - artificial intelligence,
• 3:24
  - really around inference and generative.
• 3:27
  - I think the most astonishing
• 3:30
  - aspect,
• 3:31
  - for most, especially if you're not as close
• 3:33
  - to the energy side,
• 3:35
  - is the pure scale of of what we're
• 3:37
  - experiencing, specifically in The US.
• 3:41
  - To put it in perspective,
• 3:43
  - you know, just in the last three years,
• 3:45
  - there's,
• 3:46
  - been 487
• 3:48
  - gigawatts
• 3:49
  - worth of announced
• 3:51
  - data center,
• 3:52
  - projects that are planning to be built within
• 3:54
  - the next five years.
• 3:56
  - And even if half of that,
• 3:58
  - actually comes to fruition, it's it's quite astonishing.
• 4:02
  - And if you dig a little bit deeper,
• 4:05
  - about a 150
• 4:07
  - gigawatts
• 4:08
  - worth of that power generation,
• 4:11
  - because of the long five to seven year
• 4:13
  - plus interconnect,
• 4:15
  - challenge that we're all aware of,
• 4:17
  - is planning to go completely off grid.
• 4:19
  - So I landed power generation
• 4:22
  - at the scale of a 150 gigawatts,
• 4:25
  - to be built in the next five years.
• 4:28
  - And I guess just to put that a
• 4:30
  - little more in perspective, that's the equivalent
• 4:34
  - of France and The UK
• 4:35
  - combined from a power generation standpoint,
• 4:38
  - or another perspective would be,
• 4:41
  - 25%
• 4:42
  - of The entire US peak demand today.
• 4:46
  - So the scale of what we're experiencing
• 4:49
  - is really unprecedented and will have
• 4:52
  - ripple effects across many aspects even in ancillary
• 4:55
  - industries from a supply chain standpoint, from a
• 4:58
  - grid reliability
• 4:59
  - standpoint,
• 5:00
  - grid capacity standpoint. And I think
• 5:03
  - the landscape is quickly evolving at a pace
• 5:06
  - we've never seen before.
• 5:08
  - And so it's really critical that,
• 5:11
  - you know, everyone, even if not directly related
• 5:14
  - to data centers, needs to be aware in
• 5:16
  - order to more strategically plan and,
• 5:20
  - be prepared for how these impacts could potentially
• 5:24
  - come down to, you know, health care facilities.
• 5:27
  - Yeah. And Greg, that was exactly where I
• 5:30
  - was hoping to
• 5:31
  - to go a little bit deeper. You know,
• 5:33
  - some of these numbers you're sharing are
• 5:35
  - astounding.
• 5:36
  - And
• 5:37
  - I'm curious,
• 5:38
  - if we boil this all down for a
• 5:39
  - health system, what does
• 5:42
  - this really significant growth, what does it translate
• 5:44
  - to on the ground? Are we talking higher
• 5:46
  - energy costs, risks of outages,
• 5:48
  - or or something else?
• 5:51
  - It's good good question.
• 5:53
  - Hopefully, it won't necessarily impact
• 5:56
  - rate payers and rates in general, but that's
• 5:59
  - largely dependent on, you know, kind of the
• 6:01
  - evolving regulatory and,
• 6:03
  - legislative,
• 6:05
  - landscape.
• 6:06
  - So I think that's definitely an aspect and
• 6:09
  - obviously we've seen that trend over the last
• 6:12
  - two to three years,
• 6:13
  - increase rates.
• 6:16
  - But reliability,
• 6:17
  - you know, I think is also a concern.
• 6:20
  - Schneider Electric's,
• 6:22
  - own research institute has has done some pretty
• 6:24
  - extensive research and can be found on our
• 6:26
  - website.
• 6:27
  - But we're
• 6:28
  - what we're seeing or predicting
• 6:31
  - is really kind of an increase in,
• 6:33
  - reliability
• 6:34
  - challenges as we look at kind of 2029,
• 6:37
  - 2030
• 6:38
  - time frame.
• 6:39
  - And so that certainly has to be a
• 6:41
  - critical factor when evaluating,
• 6:43
  - you know, your current energy spend as well
• 6:45
  - as reliability
• 6:47
  - aspects,
• 6:48
  - you know, in particular for health care, which
• 6:50
  - is obviously paramount and critical.
• 6:53
  - Yeah. Greg, appreciate you putting that all into
• 6:55
  - perspective
• 6:56
  - for our leaders who are listening in.
• 6:58
  - James, I wanna turn it over to you
• 7:00
  - next and talk a little bit about,
• 7:02
  - what this is all looking like at the
• 7:03
  - enterprise level. You know, as organizations are scaling
• 7:07
  - AI from pilots to enterprise wide deployment, we're
• 7:11
  - hearing more about the infrastructure that's required to
• 7:14
  - support those efforts. So
• 7:16
  - from the supply side,
• 7:18
  - James, what are you seeing in terms of
• 7:20
  - power, demand, grid capacity, and facility readiness?
• 7:25
  - So one of the things I I think
• 7:26
  - is the biggest misconceptions, first of all, I
• 7:28
  - just wanna kinda jump back to kinda like
• 7:30
  - what Greg was talking about, was is that
• 7:32
  - AI, preparation is primarily just an IT initiative.
• 7:36
  - I think that that is really
• 7:38
  - kind of a misconception for sure, like I
• 7:40
  - said. But I think of it as one
• 7:42
  - of these,
• 7:43
  - life changing, facility altering changes
• 7:46
  - that we're just beginning to think about.
• 7:49
  - And in reality, the large scale AI becomes
• 7:51
  - a facility infrastructure and operations initiative,
• 7:54
  - of course, very quickly when we think about
• 7:56
  - it.
• 7:57
  - As healthcare
• 7:58
  - organizations move from experimenting with AI to deploying
• 8:02
  - it across the, you know, the clinical side,
• 8:05
  - the the operational, and some of the administrative
• 8:07
  - workflows that have the underlying infrastructure requirements,
• 8:11
  - we start to understand that it becomes really
• 8:13
  - significant.
• 8:14
  - We're seeing increased demand for power capacity, the
• 8:17
  - cooling, the network infrastructure,
• 8:20
  - and of course resiliency planning, which is all
• 8:22
  - about, you know, making sure that we have
• 8:24
  - that patient care taken care of. And then
• 8:27
  - of course, on the supply side utilities across
• 8:29
  - many regions, they're already evaluating, you know, how
• 8:33
  - do we support this rapidly growing load from
• 8:36
  - a hyperscale in the data center space?
• 8:38
  - Internal to Schneider, there's been research and industry
• 8:41
  - reports that show data
• 8:43
  - centers are really,
• 8:46
  - at a point where
• 8:48
  - the consumption
• 8:49
  - is accelerating dramatically like Greg had talked about,
• 8:52
  - and it's and it's kind of driving these
• 8:54
  - AI loads
• 8:55
  - to the point that
• 8:56
  - for a health organizations,
• 8:58
  - they really need to be thinking about power
• 9:01
  - availability and long term capacity planning. And it
• 9:04
  - really just can't be something that's, you know,
• 9:07
  - taken for granted or maybe they'll wait till
• 9:09
  - the next
• 9:10
  - study or something. They really need to have
• 9:12
  - a a,
• 9:13
  - one three five plan now.
• 9:16
  - Yeah. Yeah. I can sense the urgency behind
• 9:19
  - that, James.
• 9:20
  - I'm also curious.
• 9:22
  - You know, there's a lot of discussion in
• 9:24
  - the AI space about cloud capabilities. So
• 9:28
  - for AI workload specifically, how does the on
• 9:31
  - premise versus cloud decision
• 9:33
  - play out differently for a hospital? And what
• 9:35
  - do you think leaders should really be weighing
• 9:37
  - there?
• 9:38
  - Well, that's a that's a very interesting question
• 9:40
  - because
• 9:41
  - for, you know, hospitals and for many people,
• 9:43
  - they think, oh, well, the easy answer is
• 9:45
  - we'll just use the cloud. Right? Therefore, I
• 9:48
  - don't have to worry about anything for my
• 9:50
  - infrastructure
• 9:51
  - at my particular hospital.
• 9:53
  - But it's really matching
• 9:55
  - the workload
• 9:57
  - and the mission
• 9:58
  - to what's there. And so hospitals are are
• 10:01
  - really
• 10:02
  - designed
• 10:03
  - and and really by code required to defend
• 10:05
  - in place.
• 10:06
  - So when everything else goes down, when the
• 10:08
  - infrastructure goes down within the city,
• 10:11
  - they need to keep running. They need to
• 10:12
  - make sure that they have optimal patient care.
• 10:15
  - And so with that,
• 10:17
  - they need to then look at that scalability
• 10:21
  - and what does that mean for the hospital?
• 10:23
  - I can't really all go to the cloud
• 10:26
  - because I can't have all my decisions be
• 10:28
  - there because I may lose communications. So now
• 10:30
  - what do I do? So then that pushes
• 10:32
  - the the,
• 10:33
  - health care
• 10:34
  - facility into a decision of, well, we probably
• 10:37
  - are going to have to look at maybe
• 10:39
  - a hybrid in this particular situation.
• 10:41
  - And so they're gonna be looking at, you
• 10:43
  - know, where basically latency,
• 10:46
  - data governance,
• 10:47
  - cybersecurity requirements, the operational,
• 10:49
  - contingency
• 10:51
  - may drive the organization
• 10:53
  - for localized,
• 10:54
  - computerized,
• 10:56
  - sources being
• 10:57
  - on premise.
• 10:58
  - And what I think healthcare leaders are going
• 11:01
  - to have to think about is really
• 11:03
  - the clinical response and time requirements,
• 11:06
  - the kind of the data privacy and governance,
• 11:08
  - again, like I mentioned,
• 11:10
  - the
• 11:10
  - infrastructure and energy costs that they'll be looking
• 11:13
  - at, and then the overall
• 11:15
  - operational resilience during the outage
• 11:18
  - or disruptions that they may have. Ultimately, I
• 11:21
  - think the right answer
• 11:22
  - in my mind is
• 11:24
  - some combination
• 11:25
  - of the cloud edge and on premises architectures
• 11:29
  - rather than just a single approach.
• 11:32
  - Yeah. That was going to be my next
• 11:33
  - question. Is there one you recommend over the
• 11:35
  - other? So it sounds like a hybrid approach
• 11:38
  - of sorts is what you'd recommend.
• 11:40
  - Great points, James. And something I wanna go
• 11:42
  - back to, you had
• 11:44
  - touched on this idea of
• 11:47
  - this work, this building the infrastructure to run
• 11:49
  - AI at scale. It's not just an IT
• 11:51
  - project. So I'd love to get both of
• 11:54
  - your perspectives on
• 11:56
  - how much coordination really matters between teams like
• 11:59
  - facilities,
• 12:00
  - IT, and clinical leaders.
• 12:02
  - Can you say a little bit more about
• 12:03
  - that coordination and and where you see it
• 12:05
  - getting difficult?
• 12:07
  - Greg, you wanna get us started on that
• 12:08
  - one?
• 12:09
  - Yeah. And I'll let James,
• 12:11
  - pile on too.
• 12:12
  - I think coordination
• 12:14
  - in partnering across many different avenues that we've
• 12:17
  - historically not had to to do is is
• 12:20
  - gonna be critical.
• 12:21
  - And again, just for perspective,
• 12:23
  - you know, to to build on what James
• 12:25
  - said earlier,
• 12:26
  - you know, as we're in the early innings
• 12:28
  - of understanding how to fully leverage
• 12:32
  - the capabilities of AI, I think healthcare,
• 12:35
  - is gonna be
• 12:36
  - an enormous opportunity
• 12:38
  - for many different aspects across
• 12:40
  - healthcare from radiology
• 12:42
  - to digital pathology,
• 12:44
  - patient monitoring, security video, predictive operations.
• 12:49
  - And in order to really
• 12:50
  - take advantage of, you know, the more advanced
• 12:53
  - AI, the inference,
• 12:55
  - AI
• 12:56
  - is likely gonna require, you know, the low
• 12:59
  - latency,
• 13:00
  - on prem,
• 13:02
  - small edge data center.
• 13:04
  - And it's it's, again, worth understanding
• 13:07
  - the energy aspects of that because,
• 13:09
  - you know, we're we're hearing about these massive
• 13:11
  - data centers from a from a training standpoint.
• 13:15
  - But the edge data centers are are fairly
• 13:17
  - significant in size as well. So,
• 13:19
  - you know, if an average 500 bed hospital
• 13:22
  - is around eight to 10 megawatts
• 13:25
  - and you're looking at fully adopting and deploying
• 13:28
  - AI,
• 13:29
  - across your your hospital. You're probably looking at
• 13:32
  - a one to two
• 13:34
  - megawatt on prem data center because of low
• 13:36
  - latency requirements.
• 13:38
  - And so you're looking at 10 to 25%
• 13:41
  - increase in energy just for AI alone, not
• 13:44
  - to mention the other aspects that are increasing
• 13:47
  - the load requirements at a hospital for, you
• 13:50
  - know, EVs and and fleet electrification
• 13:54
  - so that
• 13:55
  - we know
• 13:56
  - the energy increases are coming.
• 13:58
  - And historically, you know, you would you would
• 14:01
  - have, you know, possibly the luxury to go
• 14:03
  - back to your utility and request an extra
• 14:05
  - megawatt, you know, a couple megawatts.
• 14:08
  - You may not have that luxury in the
• 14:10
  - timeframe needed because of the pace
• 14:13
  - at what the industry is moving. So I
• 14:14
  - think really to James's earlier point,
• 14:17
  - looking more strategically ahead at the next five
• 14:20
  - to ten years,
• 14:22
  - to really understand
• 14:24
  - what you can do now to kinda derisk
• 14:26
  - the available energy that that's gonna be needed.
• 14:29
  - Mhmm.
• 14:30
  - Yeah. James, anything you would add?
• 14:32
  - You know, just as as Greg was talking,
• 14:34
  - I was thinking about this very thing. And
• 14:37
  - one of the things is is that the
• 14:38
  - significant load that we're asking to increase at
• 14:41
  - a at a facility
• 14:43
  - and just one simple thing, will the transmission
• 14:46
  - distribution lines be able to handle that?
• 14:49
  - Well, heavens, you know, that's not something they
• 14:50
  - can just pull new wire overnight and and
• 14:53
  - to a facility and make that happen. That
• 14:55
  - has to be planned ahead of time. So
• 14:57
  - if you're wanting it quick, you may be
• 14:59
  - looking at on-site generation
• 15:01
  - or, an awful lot of money to to
• 15:04
  - get power in the place that you need
• 15:05
  - it.
• 15:07
  - Yeah. And what I'm taking away from you
• 15:09
  - both is it's really about getting ahead of
• 15:12
  - identifying
• 15:13
  - and derisking available energy to Greg's point earlier.
• 15:16
  - And, James, I know what you're saying here
• 15:17
  - is along those same lines
• 15:19
  - of preparing,
• 15:20
  - essentially.
• 15:22
  - But I'm curious if you'd add anything, you
• 15:24
  - know, for that leader who wants to get
• 15:26
  - ahead of this challenge,
• 15:28
  - and they wanna derisk
• 15:29
  - available energy or take some of those proactive
• 15:31
  - steps, what's a what's a realistic starting point?
• 15:35
  - Well, when we just step take a step
• 15:37
  - back and look at historically, you know, facilities
• 15:39
  - teams have managed power and infrastructure.
• 15:42
  - And and, of course, the IT folks have
• 15:43
  - managed, you know, basically applications and data.
• 15:47
  - And that world's been pretty siloed
• 15:49
  - up till now. And I think this is
• 15:51
  - going to force a number of alliances
• 15:54
  - that have sort of worked well together but
• 15:56
  - really haven't always kind of had to sit
• 15:58
  - down at the table and really talk things
• 16:00
  - through.
• 16:01
  - And I think, you know, and what I'm
• 16:03
  - referring to is kind of like the new
• 16:04
  - construction and operations side of a hospital.
• 16:07
  - They're going to have to have those more
• 16:09
  - meaningful conversations
• 16:11
  - as the AI boundaries blur.
• 16:13
  - And this is going to then foresee suspicions
• 16:16
  - to be really made by IT leaders of
• 16:19
  - what, you know, those significant implications that they're
• 16:22
  - asking for
• 16:24
  - and what that means to the facility and
• 16:26
  - and and vice versa,
• 16:28
  - and the facility being able to to come
• 16:30
  - up and and and supply that. And then,
• 16:32
  - of course, we have the utility in that
• 16:34
  - mix as well.
• 16:36
  - So all these decisions can directly impact,
• 16:39
  - the technology, the deployment, and the scalability we're
• 16:42
  - looking for.
• 16:43
  - And then I really believe that the organizations
• 16:46
  - that are gonna succeed
• 16:48
  - in in the future, really, in this side
• 16:50
  - is is the ones that bring, here again,
• 16:53
  - the facilities, the IT, the cybersecurity,
• 16:56
  - the clinical leadership,
• 16:57
  - and finance together early. Like Greg's saying, we
• 17:00
  - put together that plan with the CFO
• 17:03
  - to to see what's the investment going to
• 17:05
  - be
• 17:06
  - and what do we need to do based
• 17:08
  - on what we have available to us. To
• 17:10
  - James's point, I think looking creatively at
• 17:14
  - on-site power generation,
• 17:16
  - you know, microgrids
• 17:18
  - historically have been,
• 17:20
  - maybe a luxury or a nice to have
• 17:22
  - and it's quickly, I think, trending to
• 17:24
  - a critical decision factor on how much on-site
• 17:27
  - generation is needed to de risk,
• 17:31
  - the energy needs.
• 17:32
  - But I'd say there's also a lot of
• 17:34
  - creative approaches being developed right now in the
• 17:36
  - market where, you know, you don't necessarily need
• 17:39
  - to spend a lot of CapEx to get
• 17:41
  - it. And there's a lot of independent power
• 17:43
  - producers,
• 17:44
  - energy as a a service providers,
• 17:46
  - even utilities are evolving pretty quick to be
• 17:49
  - able to
• 17:50
  - do energy as a service for large scale
• 17:53
  - on-site power generation
• 17:56
  - that would also protect you from rate increases
• 17:58
  - as well and potentially unlock new revenue
• 18:01
  - streams where the grid gets more and more
• 18:03
  - constrained.
• 18:04
  - You'll be able to
• 18:06
  - to monetize that excess capacity and sell back
• 18:09
  - to the grid
• 18:10
  - and take advantage of, I would say, increasing
• 18:13
  - kinda demand response in virtual power plant,
• 18:16
  - options that, again, I think have have been
• 18:19
  - explored in the past, but I think we're
• 18:22
  - gonna, you know, in the next five years
• 18:23
  - quickly see see that become more and more
• 18:26
  - of the norm versus,
• 18:27
  - the exception.
• 18:29
  - Yeah.
• 18:30
  - Well, it's been such a great conversation with
• 18:32
  - the both of you. A lot
• 18:34
  - of practical advice and considerations for our leaders
• 18:37
  - while also, you know, some fair warnings about
• 18:39
  - what's ahead if that preparation isn't taken seriously.
• 18:43
  - So I wanna thank you again for the
• 18:44
  - time that you spent with Becker's today.
• 18:47
  - Before we close, is there anything that we
• 18:49
  - didn't cover or any final thoughts that you'd
• 18:51
  - like to leave listeners with?
• 18:53
  - I I guess from my perspective, I I
• 18:55
  - know it's it's maybe somewhat of a daunting
• 18:57
  - discussion,
• 18:59
  - which I think is warranted. But, you know,
• 19:01
  - on a positive note, you know, this is
• 19:03
  - driving a lot of collaboration and creativity in
• 19:05
  - the market. So I would encourage everyone to,
• 19:08
  - you know, reach out to your partners and
• 19:10
  - vendors and challenge us in new ways to
• 19:13
  - help help you,
• 19:15
  - solve your problems and kinda co innovate alongside
• 19:17
  - you.
• 19:18
  - James, how about you? Yeah. So, you know,
• 19:20
  - I'd I'd like to add to that too.
• 19:22
  - I I think that if I were to
• 19:23
  - look at a a takeaway from this, it's
• 19:26
  - it's it's really that, you know, AI readiness
• 19:30
  - and energy readiness are becoming really inseparable.
• 19:33
  - You can't think about one without the other.
• 19:35
  - And as Greg said, we're we're just on
• 19:38
  - the cusp of this. We haven't really begun
• 19:40
  - to see
• 19:41
  - the full fruition
• 19:42
  - what is going to be. We're just at
• 19:44
  - that where the s curve starts to move
• 19:46
  - up.
• 19:47
  - And, you know, health care is always focused
• 19:49
  - on resiliency
• 19:51
  - because it's a code compliance and, of course,
• 19:53
  - patient care. That's why we're doing this is
• 19:55
  - to to hopefully save lives and make lives
• 19:57
  - better. But
• 19:59
  - nothing that works in a hospital without electricity.
• 20:02
  - And the adoption
• 20:04
  - of AI and how it accelerates
• 20:07
  - and the electrical grid facing the new pressures
• 20:10
  - that it sees from the new data center
• 20:11
  - growth, the healthcare organizations
• 20:13
  - having the opportunity to strengthen their both their
• 20:16
  - digital infrastructure and their power infrastructure,
• 20:19
  - I think that we need to start planning
• 20:21
  - now
• 20:22
  - to be the in the best position to
• 20:23
  - leverage really more the AI
• 20:26
  - safely,
• 20:27
  - reliably, and at scale
• 20:29
  - while continuing to deliver that exceptional patient care.
• 20:33
  - James, Greg, I wanna thank you both again
• 20:35
  - for your time and your insights today. Really
• 20:37
  - appreciate you doing this. It was a pleasure.
• 20:40
  - Thank you, Erica.
• 20:41
  - Thank you very much. Grateful to be here.
• 20:43
  - Thank you. And we also wanna thank our
• 20:45
  - podcast sponsor, Schneider Electric.
• 20:48
  - Listeners, be sure to tune in to more
• 20:49
  - podcasts from Becker's by visiting our podcast page
• 20:52
  - at beckershospitalreview.com.