AI Isn’t Eliminating Work. It’s Moving the Bottleneck

In this episode of the Macro AI Podcast, Gary Sloper and Scott Bryan examine one of the most important questions facing business leaders today: is AI eliminating work, or is it changing where work gets stuck? 

While much of the public conversation focuses on job replacement, the bigger strategic issue may be that AI is moving the bottleneck. AI can make individual tasks faster — from writing and research to coding, customer support, forecasting, and design — but that does not automatically make the entire enterprise faster. In many cases, AI simply exposes the next constraint: approvals, data quality, governance, implementation capacity, supplier readiness, field labor, compliance, or physical infrastructure. 

Gary and Scott discuss why the labor market is not yet showing a simple AI-driven job-loss story, why entry-level career paths may be one of the first pressure points, and why individual productivity gains do not always translate into enterprise-wide economic gains. They also explore how AI can create new work by making ideas, experiments, and business models cheaper to pursue. 

The episode highlights examples across healthcare, manufacturing, banking, retail, telecom, and software, showing how AI shifts the constraint from knowledge production to workflow absorption. The discussion also explains why physical bottlenecks — including data centers, power, cooling, manufacturing capacity, clinical capacity, logistics, and supplier readiness — will matter more as AI accelerates planning, design, analysis, and demand generation. 

The key takeaway: AI is not just a labor replacement technology. It is a throughput technology. The companies that win will be those that map their workflows, anticipate where bottlenecks will move, redesign early-career training, modernize their workflow layer, and use AI for growth — not just cost cutting. 

About your AI Guides

Gary Sloper

https://www.linkedin.com/in/gsloper/

Scott Bryan

https://www.linkedin.com/in/scottjbryan/

Macro AI Website: 

https://www.macroaipodcast.com/

Macro AI LinkedIn Page:  

https://www.linkedin.com/company/macro-ai-podcast/

Gary's Free AI Readiness Assessment:

https://macronetservices.com/events/the-comprehensive-guide-to-ai-readiness

Scott's Content & Blog

https://www.macronomics.ai/blog

Transcript

00:00
Welcome to the Macro AI Podcast,  where your expert guides Gary Sloper and Scott Bryan navigate the ever-evolving world of artificial intelligence.  Step into the future with us  as we uncover how AI is revolutionizing the global business landscape  from nimble startups to Fortune 500 giants.  Whether you're a seasoned executive,  an ambitious entrepreneur,

00:27
or simply eager to harness AI's potential,  we've got you covered.  Expect actionable insights,  conversations with industry trailblazers  and service providers,  and proven strategies to keep you ahead in a world being shaped rapidly by innovation.  Gary and Scott are here to decode the complexities of AI  and to bring forward ideas that can transform cutting-edge technology  into real-world business success.

00:57
So join us,  let's explore, learn  and lead together.  Welcome back to the Macro AI podcast.  I'm Gary Sloper joined as always by Scott Bryan. Today we're revisiting a topic that we've touched on a few times here on the show. We've discussed it quite a bit outside of the show,  but the facts just keep rolling out in some great studies. And the topic is what is AI really doing to work? Yeah, exactly. I think the

01:25
question that we easily get a few times a month is how many jobs is AI going to replace?  Some version of that question. Yeah.  And I think that's a understandable question, especially with the media and causing a lot of hype here. And  generative AI can write, it can summarize, analyze uh code, create images. Many people use a lot of the public LLMs to create funny images. We've seen that. uh Other organizations have

01:54
have branched into generating reports and assisting with their customer support, maybe drafting legal language and even taking action inside software tools. So this feels different from prior waves of automation because it's not just touching the factory floor  or the call center script like we've seen with other, you know, kind of tech evolutions. It's reaching into the  work of educated professionals and it's already causing a lot of college students to reevaluate their target.

02:24
majors, lot of organized post-secondary schools are starting to change how they teach. And if you look at a lot of the recent data that shows that AI may not be eliminating work as quickly as people expected, which I think is a bit of a surprise. So the more important story is that AI is moving the bottleneck. Yeah, no, exactly. And I think for business leaders, which are, you know, a lot of our listeners are business leaders,

02:53
ah That's a more useful way to think about it, ah that AI is moving a bottleneck. so the question is not, shouldn't really just be, you know, which jobs will AI eliminate? I think maybe the better question is ah when AI makes parts of the business dramatically faster, where does the bottleneck move or, you know, suddenly become visible? And I think that's where the strategy focus really needs to be is  the new bottlenecks that are appearing. And what I mean is,

03:23
AI can make certain tasks faster, but that doesn't automatically mean the entire company gets faster. It may simply expose the next constraint. So whether that's, know, approvals or  data quality or implementation capacity, physical infrastructure, whatever it is,  or even management decision-making at that new bottleneck. So today we're going to dig into that idea.  So framing AI as a throughput technology and not just

03:53
you know, labor replacement. Yeah. Yeah, it's a good point. And I think the  labor market is not telling a simple job loss story. So let's start with this job loss narrative. There are absolutely jobs and functions under pressure. I  don't disagree with that. I think the data is there. We've seen it. We know people impacted. uh And we're seeing AI affect areas such as freelance markets, uh content creation.

04:20
Customer service is a big one, especially from an agentic standpoint that we've talked about  here on the show. Software development, a lot of people have all of sudden overnight become coders  that knew nothing about application development in the past. And also some  entry-level knowledge roles. And I think this one's important uh because  when you look at the overall labor market, the evidence so far does not show the clean immediate collapse that some people predicted. ah

04:49
That doesn't mean AI isn't having an impact. It means the story is a little bit more complicated and we'll kind of dive into that a little bit more here  as we continue in this episode. Yeah, definitely.  And I think that's where executives need to be really careful. A lot of people are confusing a couple different things. So first, know, what AI is technically capable of doing. uh Second, whether their employees are actually using it.  And then third, whether the company has

05:18
redesigned workflows around it. And then lastly, whether those productivity gains actually translate into measurable business outcomes.  And those aren't the same things.  AI might be capable of performing parts of a job, but that doesn't mean a company is ready to integrate it,  completely trust it, restructure their teams,  and actually go in and reduce headcount around it. Yeah, Scott, that's a critical distinction. uh

05:46
that you just made. capability does not equal adoption. And I think that's a big one, right? Just because you can do certain things doesn't mean the whole organization has embraced it or put it into their process.  And that doesn't equal productivity. And productivity does not automatically equal labor reduction.  So a lot of companies are still in this early phase. Employees are experimenting. We've seen in many organizations publicly talk about how much they're burning through tokens, for example. ah

06:16
teams are using AI to draft and summarize research or brainstorm for their company, but converting that into enterprise level transformation is much, harder to do. Yeah, exactly. And I think that's why the labor market  signals are uh mixed. We're not seeing  no impact. We're seeing the early phase of a  much larger, basically total reconfiguration of work. And I think AI is already changing expectations.

06:45
It's, changing what, uh, how people perceive good performance  and how quickly people are expected to produce. But the organizational model hasn't fully caught up to this,  uh, everything that's happening yet. Right. Right. And I mentioned earlier, and I kind of want to touch upon this now, which is, you know, one area where the impact is showing up earlier is in the entry level work. Yep. So there are many of us that have been in tech or other job roles. Uh, think about.

07:15
you know, that first entry level position. A lot of junior roles  historically have existed because organizations need people to do repetitive knowledge work. It could be, you know, drafting the first revision of a document, pulling down research, summarizing the meeting and meeting notes and, you know, next steps, uh preparing a spreadsheet or, you know, pivot tables and other things from an analysis standpoint, reviewing tickets,  creating the basic

07:43
you know, analysis around that, or even just writing, you know, the first few lines of code. So, so a lot of that uh work was not always glamorous,  but it's really how young professionals learned. I learned that way. I learned in an application development environment, doing very small little bits of work because I was entry level. Same thing, you know,  like you just kind of, that's, that's kind of the on the job training that you received at those, at those junior levels.

08:11
Yeah. And a lot of us have, uh, have kids thinking about, uh, college and what they're going to do. Right. And I think, yeah, it's, it's tough.  Um, and I think that AI is, is really good at a lot of those first draft and first pass tasks.  companies might look at junior roles and say, you know, why do we need someone to do that? If AI can do it faster, right?  Um, but the problem is those tasks were also training grounds for all their entry level folks. And, and that's how, that's how these people learned.

08:41
judgment and what good work looks like. ah It's how they learned oh context and risks and trade-offs. And also really importantly, it's where a lot of entry level people get experienced with clients and client expectations. So that's a  key area. So if  AI compresses the  early career ladder too much, then companies really could kind of cut off  their  future talent pipeline.

09:08
Yeah, and that's a really important management issue there, Scott. And I'm glad you kind of touched upon that. Companies still need people with judgment. Still need people that can be creative and have domain expertise. uh Empathy is another one that I can't stress enough when  I talk to folks, especially early in their career.  And that problem framing ability. How do you control your learning curve and build that strategic thinking? So, you know, the...

09:36
In hindsight here, if the traditional apprenticeship path disappears, where do those skills come from? Yeah. Yeah. I think  the companies that are successful here  aren't going to simply replace junior work with AI. I think the real winners are going to redesign those junior roles. And a lot of them are starting to do that based on some of the research we did. So instead of asking entry level employees to spend all day producing drafts, they might spend more time.

10:05
you know, reviewing AI outputs,  validating AI outputs and facts, and actually handling some of the exceptions that come out of those cases.  And then probably a little bit more time working with the customer base and learning how decisions are made. And that's kind of a slightly higher bar than some of the traditional entry level roles. But if companies build the right training model, I think they can actually create stronger

10:33
professionals right out of the gate with a little bit more, know, jumping right into a little bit more responsibility. Yeah, I 100 % agree with you. And especially the point you made around validating facts. I what a great way to enable a young professional to kind of go through whatever the output is to make sure that it's accurate. Right. And we've seen what happens when things are not accurate.  So,  so for business leaders listening right now, this is not just an HR topic. It's a strategic

11:01
capability topic. If you do not redesign the early career path, you may save some money in the short term, ah but you will almost certainly damage your future leadership pipeline, right? They need to be able to build that same pedigree that you  were afforded in a much more manual environment, maybe  five, 10, 15 years ago.  Yeah. So if you think about productivity gains,  I think another major point is that individual productivity  and

11:31
enterprise productivity are really not the same thing. AI can absolutely make individuals seem faster. You know, like a customer service agent can resolve multiple issues faster. know, marketers can create a lot of different variations on a campaign. Developers can make a lot of prototypes, et cetera. And an executive might, you know, be able to synthesize a large

11:58
set of documents and minutes instead of hours. And those are, those are definitely real gains. But if the company doesn't automatically become more profitable,  it just, they just don't automatically become more profitable because someone saved a couple hours. Right. And that's  such an important point. Those two hours can disappear into more meetings,  additional rework, more review cycles,  added low value activity, or simply more output that no one acts on.

12:27
You might generate more reports, but not make a decision faster. It's that old, know, paralysis through analysis. ah If your org generates more ideas, you need the operating capacity to implement them. Otherwise you're just, you're going to be bottlenecked. Yes. Yeah. You'll need the operating capacity to act on those ideas. Totally agree. And I think AI creates leverage and acceleration, but leverage  then has to go somewhere. So if the workflow  is actually broken,

12:56
AI is just simply going to accelerate the broken workflow right up to the point of the next bottleneck.  I think another example would be, approvals are slow, AI is just going to simply accelerate it right up to the queue waiting for the next approval. m So if you've got a bunch of managers in a flatter organization that are already overwhelmed, which is the case in a lot of businesses, AI is just simply going to accelerate that information, and it's going to be more than they can absorb.

13:23
Yeah. And so this is where companies run into what looks like an AI productivity paradox. Employees feel faster. Teams produce more, but the enterprise doesn't  see the full economic benefit because work jams up at the next bottleneck. kind of, you know, those stops. Yep. And I think that's,  that's because the real work isn't just adding AI tools. The real work is, is, is getting in, identifying the workflows and redesigning them.

13:53
Uh, the governance around them,  the data architecture, um, and the, the levels of accountability. And that's where AI starts to become business transformation rather than just individual productivity improvement. So consultants  and managers need to be able to map out where exactly AI can add acceleration and what step the work will then pile up at. Yeah. Good point. So.

14:22
Let's shift to another interesting idea around how artificial intelligence will create work by making ideas cheaper.  lot of the job loss discussion assumes there's a fixed amount of work. And that view, if AI performs a task, a person loses that task and eventually a job disappears. But economies don't really work that way. When the cost of producing something falls,  demand often expands. Yeah.  I think that's a huge point. So AI

14:51
lowers the cost of cognition, of thinking,  of writing, designing, analyzing, coding, all those things that we've been talking about. ah But that means that companies will try more things. There'll be an influx of ideas. So a product team that could test three ideas might now be able to look at 30.  And then in marketing, I was a marketing major, team that could personalize campaigns for just a few segments can now personalize for hundreds.

15:19
And a finance team that could model just oh a handful of scenarios can model thousands. So there's just more,  more ideas, more potentials. And I think a common one that we're hearing, we've talked about this all the time, a software team that can oh go and just prototype one feature can now prototype lots. Exactly. And that explosion of ideas driven by artificial intelligence creates new work. So.

15:48
some of it's technical work, some of it's managerial or human work, and some  sort of it moves more quickly to the point  which  it becomes more physical work. So just keep that in mind with this explosion of ideas driven by AI. Yeah, yeah, so physical work. So when you have a whole bunch of ideas, I think a lot of people miss this,  AI is just maybe just digital, but it creates real world demand.

16:17
So all the things behind it, so the data centers, power, cooling, fiber, chips, ah those details we've covered pretty extensively in other shows. But outside of the technology, AI can increase demand in areas like uh manufacturing, logistics, uh healthcare, and supply chain operations. So I think  the most important new jobs might not even have AI in the title. may be in the industries  and the functions that have to absorb the demand that AI generates.

16:48
And I think this demand is really just starting to kind of just beginning to ripple out into the macro economy. Yeah, that's a good point. And it will be interesting because if those jobs do not have quote unquote AI in the job title, as you mentioned, will that, I wouldn't say turn off potential employees, but they may be looking for that just as that stepping stone in the overall ecosystem, right? So if you're an HR or

17:17
people organization, just something to be mindful of as well that  people that are looking for these roles may be looking for some sort of AI component in their job description. And I guess that's just, that's the human element of this, right?  And this  also brings us to the core idea of  the episode. So before AI, one of the  major constraints in many companies was the speed of human knowledge production. And I think we can all agree on that.

17:46
really centered around how long does it take to write the report or analyze the data? ah What is the time it takes to build the presentation or respond to the client in that ticket and that contact center environment as we talking about earlier? Or  what I think a lot of organizations uh use some of these AI tools around coding, how long does it take to draft the code, that first rev? Can we increase our sprint cycles? And AI reduces many of those constraints.

18:13
ah which is why we wanted to have this episode and kind of talk through all this. Yeah, exactly. Because once those constraints are reduced, the bottleneck simply moves. The new constraint  is  the rest of the company. How can the company absorb, uh validate, decide, and  scale? And that's a very different problem.  Yeah. So let's jump into examples that we came up with before the show, Scott. I know we wanted to...

18:41
kind of think about this before we started recording, but maybe we start, ah you know, the first one, I think you were talking about healthcare. Yeah, I think that's a good way. The examples are good way to kind of clarify it. um So in healthcare and AI, well, AI uh might identify patient risk faster.  And that's a, that's kind of a common healthcare use case, but then the bottleneck might become appointment availability, uh clinician capacity, reimbursements.

19:10
care coordination, you all of the downstreams from, uh, from identifying patient risk faster.  Um, and then just take another one in manufacturing. AI will, will certainly generate better product designs much faster than the bottleneck then shifts to  all the other things, tooling,  supplier lead times,  plant capacity, quality control or regulatory testing, all those downstreams from the ideas. Yeah. Good examples.

19:39
I wrote down banking, so in banking,  think  when you look at that particular vertical, artificial intelligence will accelerate things such as underwriting, which definitely speeds up the loan process, which I think many of us would appreciate. uh But the bottleneck can then shift to other areas.  Think of  compliance review, model risk management. So can you repay this? Data lineage or even  trust of that particular customer.

20:09
So these are the things that  can speed up but also kind of get bogged down in other components. And then if you look at retail, AI may personalize things like demand generation and spark a big rush on products that explodes on social media, maybe because of AI, for example.  But the bottleneck then shifts to inventory, fulfillment, merchandising, uh last mile delivery,  who's delivering the package.

20:38
and just overall customer service. So, you know, that that's where you could see that from a retail perspective. Yeah. Yeah. I think we're already seeing that. Yeah. And then a few more uh in telecom, which Gary, you and I live in. ah It's I think AI can design networks faster, but then the bottleneck may be the order processing permitting if necessary, delivery or equipment availability or even field labor. So all the, all the downstreams of

21:08
of being able to come up with excellent designs to use network as a service. Yes, telecoms are certainly getting closer to  NAS network as a service, there  are certainly other downstream hangups after you click  to order a circuit.  it's on that, that's one thing, but to your point, if it's a new location, you have all those other physical constraints that as of right now, uh there's not the physical AI component in telecom. Yep, exactly.

21:38
Um, yeah, was throw it one more, um, in, software, we've talked about this a few times and the coding, uh, idea is, everywhere in the, in the news. Um, so AI is already generating much more code than the bottleneck moves to where it shifts to architecture, security review, you know, technical data across the organization, product management, and then, you know, even further downstream customer adoption. So.

22:04
you know, every, everything else, once that code has been jammed out by AI becomes a bottleneck. Yeah.  And that is a great way to frame it. Um, AI does not eliminate constraints. It reveals the next constraint after,  after the first one.  Um, yeah. So I think, I think,  you know, that should change how executives think about AI strategy. The question isn't just, you know, where can we automate? next question is if this part of our business becomes

22:34
50 % faster or 90 % faster, what breaks next? And that's the bottleneck question. think that's where a lot of really good AI consultants are starting to spend their time and organizations are bringing in more AI consultants just simply to identify bottlenecks or see things that people inside the business just when they can't see the forest through the trees. Yeah. Yeah, that's a good point. And I liked your earlier topic around telecoms, obviously not just because you and I have spent a lot of time there, but you know,

23:03
If we, if we think about physical bottlenecks and maybe we spend a few minutes there on the physical side of AI, because this is a major executive blind spot. There is a tendency to think of AI as purely digital. So software, models, apps, agents, automation, et cetera. But AI is deeply physical. And we mentioned telecom as one example, but you know, AI definitely has that physical component that you can't lose sight of. Yeah. I mean, AI like we've talked about depends on

23:33
data centers, chips, power, pooling, fiber, water,  construction,  everything related to the data center. And the digital economy increasingly depends on physical infrastructure that  can't scale at software speed. And I think this pattern applies beyond AI infrastructure market and  goes back to those examples that we mentioned. So AI is making planning faster than the execution.

24:00
and it's making demand generation faster than the actual fulfillment. And then, you know, a couple other examples like it's making design faster than the actual manufacturing process. And that healthcare example, it makes diagnosis and administrative work faster than the actual treatments. Right. So in a lot of industries, artificial intelligence may actually make physical constraints more visible. They appear a lot.

24:28
Easier. You can generate the strategy quickly. You can generate the plan quickly. You can generate the design quickly, but can you build it? Can you staff it? Can you supply it? Could you deliver it? Yeah. Yeah. I think that's the key. I think the companies that are going to be successful in the, in the near term and going forward in the, in the world of AI will be the ones that really understand where the bottleneck moves and they'll invest ahead of it until eventually they become more AI native.

24:58
Yeah. Yeah. Good. Good point. All right. So let's,  know, from a business standpoint, let's  let's shift focus and talk about strategy uh because AI can reduce costs and streamline certain workflows. We've talked about that countless times on this show uh or just parts of workflows. uh And that allows the efficiency improvement. So every executive should be looking for those opportunities in their business.  I think it's table stakes. But there is a mistake here. Yeah. Yeah. I think that

25:28
We talked about this. think the mistake is treating AI only as a cost cutting tool. And the  bigger opportunity that  seasoned executives should be thinking about is growth. ah So AI can help companies quickly enter new markets, uh serve a lot of their smaller customers more profitably. think a lot of the telecoms are struggling with that. ah They can actually go and personalize products for you.

25:57
Um, accelerate R and D, reduce cycle times, find hidden demand across your customer base  and even create new, new service models. So the same technology can produce very different outcomes depending on the strategy driven by the business. Yeah. And to your point, I think this is where a lot of organizations get out in front of their skis too quickly. I mean, building that center of excellence, we talked about one of our very first episodes.

26:26
to really understand what your strategy is going to achieve because,  and I think we both learned this when we were taking classes at MIT, Scott,  is the strategy to make you cheaper,  more efficient, reduce headcount, cut costs, ah and just build that competitive landscape. There's a lot of things you need to answer as a company to make sure that it's worth your while.  one company may use AI to reduce support headcount,

26:54
that'll probably save money in the next quarter, but another company may use AI to provide better 24 by seven support. ah Use the technology to identify upsell, cross-sell opportunities, improve retention, serve customers that were previously unprofitable, maybe they only purchased one product.  And generally really focus on that customer experience, that customer success environment for your clients.

27:19
That second company may create a new work and new jobs in the customer success arena. And that could be around analytics, product design and implementation. So same technology, different strategy. And that's where your organization may be unique from your competitors or  folks in the industry. And you really have to kind of suss that out. Yeah. I think you nailed it with, yeah, you could save money in the next quarter or you can figure out how to reinvest right away,  which could create new jobs. uh

27:49
Yeah, just another thought on manufacturing. So, you know, one company could use AI to, you know, reduce engineering time, which would be, know, faster drawings, faster documentation, faster reports, faster production planning. But then another company could use AI to redesign the way products move from concept to production. So actually redesigning the whole process. And I think that's what a lot of people are failing to look at. Um, you know, so you can use it to accelerate design iterations.

28:19
simulate failure points, predict maintenance needs,  and then look at your supplier chain, improve supplier coordination, reduce scrap  and spot quality issues before they become expensive problems. ah But then the bottleneck moves. So maybe the engineering team can design faster, but the tooling team can't build fixtures fast enough, or maybe  demand forecasting improves, but the suppliers can't keep up, or the plant can identify defects.

28:47
but then it needs better sensors, better data, more technicians, uh skilled technicians out on the floor. So the strategic question is not whether AI replaces tasks  because it does. think the strategic question is whether the company uses that freed capacity to simply cut costs  or  to improve throughput across the entire operation. So from design to sourcing to production, quality and delivery. So you really have to look at the full picture to take advantage of it.

29:15
and reallocate jobs and talent. Yeah, those are all good points. So let's  talk for a minute about what executives and AI consultants  really should focus on now.

29:31
Yeah, I think the first step, kind of what I was alluding to in that last piece is, you know, map out end to end workflows, not just the jobs, but the entire workflows. The job titles are just too blunt. ah I think leaders should look at where the work is repetitive, where judgment is required, you know, how their data looks and where data might have weak points, ah where approvals  move through the process slowly.

30:00
And then of course, where the physical constraints actually limit execution down the line. Yeah. And I would say probably a close second is, is measuring the bottleneck migration. before deploying your, your AI environment, ask if this task becomes 50 % faster, for example, what breaks next does a work pile up in legal or finance or operations, wherever that, you know, may be kind of halt.

30:28
And the physical supply chain, like, is that another area that you could see a bottleneck as well? So any new acceleration will make work pile up at the next step and being able to measure that is very critical. Yeah. Yeah. And I'll just revisit the early career pathways again. I think AI is going to change how junior employees learn. So companies are going to need apprenticeship models that teach judgment and not just, you know, production.

30:58
And then they need to learn how to use AI, validate it and understand the business context behind the output. Yeah. I would say the, the, the next or kind of fourth step would be to map out how to modernize the work for a layer from the ground up. So a lot of companies are trying to apply AI on top of various tech debt and disconnected systems,  uh, which all have inconsistent data and formal processes for the business. know who you are. You're not.

31:26
alone, especially if you've had some M  &A, that will limit results. You really need to look at it through a whole new lens with the explosion of AI capabilities and really assess how these flows work. And maybe this is a good time to kind of just almost start fresh and retire old. Right. Yeah. And then they need to connect it with physical capacity. I think for some businesses, the limiting factor isn't the model. It'll be something physical.  And like we talked about, you know, the clinical

31:56
capacity and healthcare manufacturing capacity or supplier readiness. So AI planning has to include all the physical components in the physical world as well. Yeah, yeah, I agree. And I'd say the kind of the last year is we kind of bolt this out as number six, use AI for growth, not only for efficiency. So the biggest value may come from asking things like how can AI help serve more customers, innovate faster, make better decisions,  reduce cycle times and maybe enter markets.

32:26
you know, we as an organization could not profitably serve before  getting back to a lot of that, you know, customer service environment, you know, so, so how do you use  artificial intelligence for growth? Not just for efficiency. Yeah. And not just cutting jobs. Yep.  Yeah. So I think the big takeaway is that AI is not simply a labor replacement technology. It's a throughput technology. And I think some of the  reports and data that are coming out are showing that the companies that are successful are

32:56
not using it simply as labor replacement.  They're using it as a throughput technology. So it increases the throughput of ideas,  which are very valuable, and it gives businesses huge opportunities to move quickly and to generate new revenue.  And  that will replace some tasks, it'll replace some jobs, but it'll also create lots of new roles, lots of new companies. We're seeing a lot of that right now, especially around coding, because that's one of the first areas. uh

33:26
And then of course we're seeing a lot of new infrastructure needs to support it  and uh overall a lot of new competitive dynamics are starting to emerge in the macro economy. Right. And it is organizations learning to operate at a higher rate of change. think we can all agree there. Right. And  AI will replace work. AI will create work. But most importantly AI will reveal where the enterprise is too slow, too fragmented, too manual.

33:53
who physically constrained or too dependent on legacy workflows to absorb the productivity, it now has the potential to generate.  I mentioned, you know, kind of that M &A environment or legacy tech debt. So that is the real bottleneck. And for executives, that is the opportunity. That's it for today's episode of the Macro AI Podcast. Thank you for listening and we'll see you next time. We appreciate you sharing uh our show out to your colleagues and friends. uh

34:22
Please continue to send us  emails  on LinkedIn with any questions.  And until next time, we'll see you soon.