HR hires humans. Procurement buys software. AI agents are both — and the collision between those two functions creates something new on the org chart, while exposing a bigger leadership gap upstairs.
Every major enterprise platform this quarter — Salesforce Headless 360, Workday Agent System of Record, Microsoft Copilot Studio, SAP Joule, Oracle agentic, ServiceNow Moveworks, IBM watsonx Orchestrate — is pitching a control plane for your AI agents. But none of them is solving the real problem: who inside your organization actually owns the agent workforce, and who's steering it at the speed agents now act?
In this edition of Lens Four,
🔍 In this episode:
— Why Workday's line — "Organizations wouldn't hire thousands of employees without an HR system to manage them. The same discipline is now required for AI agents" — exposes the HR-procurement collision everyone is about to run into
— Gartner's forecast: by the end of 2026, 40% of enterprise applications will be integrated with task-specific AI agents, up from less than 5% in 2025
— Why Jensen Huang's CES 2025 line — "IT is the HR department of agentic AI in the future" — is half-right, half-wrong, and why Josh Bersin's reframe (HR teams will be the managers and caretakers of AI agents) gets closer
— Bain and IDC agreeing that per-seat pricing is ending: by 2028, 70% of software vendors will refactor pricing around consumption, outcomes, or organizational capability — and what that means for the CEO's agenda
— The contingent workforce market is real money ($171.5B in 2021, projected to $465.2B by 2031 per Allied Market Research) — and why the contingent-labor playbook is the closest analogy for agents
— Aaron Levie's "tokenmaxxing" as the strategic-prioritization problem nobody is ready for
— Why the three vendor vocabularies (employee, contractor, software) are all task vocabularies — and why the agent era needs a judgment vocabulary instead
— The Fourth Lens: the collision between HR and procurement can go two ways (meteor or dressing), but the real steering question lives upstairs with the CEO, COO, and line-of-business leaders
Fourth Lens: The forced consolidation coming over the next twelve to eighteen months solves the plumbing. It doesn't solve the operating model. The organizations that win the next decade of enterprise work will build both the function downstairs that runs the agent roster and the leadership cadence upstairs that sets direction at machine speed.
🔗 Full article and references: seanmartin.com/lens-four/whos-managing-your-agent-workforce
📧 Subscribe to Lens Four: seanmartin.com/lens-four
🎙 Redefining CyberSecurity Podcast: redefiningcybersecuritypodcast.com
🎧 Music Evolves Podcast: musicevolvespodcast.com
🌐 ITSPmagazine: itspmagazine.com
🎬 Studio C60: studioc60.com
Sean Martin is a cybersecurity market analyst, content strategist, and go-to-market advisor with more than 30 years of experience across engineering, product development, marketing, and media. He is co-founder of ITSPmagazine (itspmagazine.com) and Studio C60 (studioc60.com), host of the Redefining CyberSecurity Podcast (redefiningcybersecuritypodcast.com) and Music Evolves Podcast (musicevolvespodcast.com), and co-host of On Location (itspmagazine.com/on-location) and Random and Unscripted (randomandunscripted.com). Learn more at seanmartin.com.
🔎 Keywords: AI agents, agentic AI, digital workforce, Salesforce Headless 360, Agentforce, AgentExchange, Workday Agent System of Record, ASOR, Salesforce TDX 2026, Aaron Levie, Marc Benioff, Joe Inzerillo, Jensen Huang, Josh Bersin, Jorge Amar, Kate Leggett, Gartner AI agents forecast, IDC FutureScape 2026, Forrester agentic AI, Bain SaaS pricing, Deloitte workforce planning, KPMG total workforce planning, McKinsey hybrid workforce, Futurum sameness, Model Context Protocol, MCP, contingent workforce, ManpowerGroup TAPFIN, Allied Market Research, outcome-based pricing, consumption-based pricing, per-seat obsolescence, tokenmaxxing, CapEx vs OpEx AI, systemic HR, superagents, digital employees, HR-procurement collision, total talent management, workforce orchestration, CEO strategic intent, line-of-business leadership, employee vs contractor classification, Sean Martin, Lens Four
[00:00:30] Speaker 2: Who's Managing Your Agent Workforce, and Whose Budget Are They On?
A Lens Four analysis by Sean Martin. Read by TAPE9.
I look at the intersection of business, technology, and messaging regularly through three lenses, how organizations run their programs, how innovation is reshaping the market, and how language is shaping the decisions people make about both. The fourth lens, the connective one, is where the real story usually lives.
This week, the fourth lens is sitting at the intersection of two departments that have spent decades not talking to each other. HR hires humans. Procurement buys software. Between them, they have historically covered most of what shows up on an enterprise invoice, the people and the tools.
And then, over the last eighteen months, a new category of worker quietly showed up on the invoice. It doesn't look quite like either. It's hired, onboarded, given access to systems, assigned tasks, [00:01:30] and reviewed on outputs, all of which sounds like HR. It's also provisioned, licensed, metered, and renewed, all of which sounds like procurement. It's called an agent. And depending on which vendor sold it to you, it's been variously described as a digital employee, a teammate, a co-worker, a digital contractor, a license, a token bundle, or a SKU.
Sometimes all in the same sentence. Usually on the same slide.
If Aaron Levie is right that enterprises are heading for a hundred times to a thousand times more agents than people using the software, and if the broader agentic enterprise story the other major platform vendors are selling is even half-right, then organizations are about to find out something uncomfortable. They don't have a function staffed to manage this workforce. They have two functions half-staffed to manage half of it each. And the invoices are already landing.
That's the visible problem. The one that shows up in budget meetings and procurement reviews. It's [00:02:30] real, and it's urgent, and the rest of this piece is going to walk through it in detail. But it's also not the hardest thing in the room. The hardest thing in the room is whether the CEO, the COO, and the people actually running the business have figured out how to give strategic direction to a workforce that acts in milliseconds, because a function-level collision between HR and procurement is downstream of a much bigger question about who's actually steering the business now.
Lens One. Programs and Business. The Talent Question. Who manages the talent when half the talent doesn't sleep?
HR and procurement don't usually sit in the same meeting. When they do, it's often because someone is arguing over whether a long-term contractor should be re-classified as an employee, a fight that has historically been about tax liability, benefits eligibility, and who writes the paycheck.
That argument is about to get strange.
Workday quietly made the implication explicit in February when it [00:03:30] shipped the Agent System of Record. Workday's framing is direct, organizations wouldn't hire thousands of employees without an HR system to manage them, the same discipline is now required for AI agents. The questions the Agent System of Record is designed to answer are the ones every enterprise is already quietly fumbling. Which agents are running. Who owns them. What is their role. Are they secure and compliant. What do they cost. And are they delivering value.
That's not a software product-management vocabulary. That's an HR vocabulary. Role. Ownership. Performance. Value delivered. Workday has been using it for twenty years, about humans.
Gartner's own forecast captures the speed of the shift. By the end of two thousand twenty-six, forty percent of enterprise applications will be integrated with task-specific AI agents, up from less than five percent in two thousand twenty-five. The leap from five percent to forty percent in a [00:04:30] single year is the real story. An enterprise that hired a few agents as a pilot in Q two of last year is now being asked to operate them at organizational scale, with no new function, no new budget code, and no new playbook.
Josh Bersin has been direct about what that demands from HR. In two thousand twenty-six we will see the most dramatic transformation of HR in my career, with hundreds of traditional processes automated, Bersin said in January, predicting that superagents would eliminate up to thirty percent of traditional HR roles while enabling HR professionals to spend time on hiring, coaching, and managing the AI infrastructure. In a follow-up interview, he framed it plainly. HR teams will need to be the managers and caretakers of AI agents. Translation. HR has a new workforce to onboard, develop, supervise, evaluate, and occasionally let go, and it isn't made of people.
McKinsey has been [00:05:30] running with the same idea from the other side of the org chart. Senior partner Jorge Amar put it plainly. Some pioneering companies in this space are expressing their org charts not only in number of full-time employees, but also in number of agents being deployed in every part of the organization. Amar argued that we are going into a world where you'll have to think about your workforce as both agentic and human, and IT will not be able to do this alone. HR will play a key role, first, to really push the business on what can be done from a hybrid workforce perspective. The shortest version came from Nvidia's Jensen Huang, first at CES two thousand twenty-five and repeated in follow-up interviews. I tell my CIO, our company's IT department, they're going to be the HR department of agentic AI in the future, Huang said. They're going to be the HR department of digital employees of the future. And those digital employees are going to work with our biological ones, [00:06:30] and that's going to be the shape of our company in the future.
Put all of that together and you have a staffing problem that nobody has quite named yet. The CEO and CIO own it, because everything downstream of the decision, capacity, cost, org design, outcomes, rolls up to them. But the operating work sits in a weird middle. HR knows how to hire and manage people, but these aren't people. IT knows how to provision and run software, but these aren't quite software. Procurement knows how to buy services and licenses, but agents behave more like contractors than like subscriptions. And finance knows how to code cost lines, but isn't sure yet whether to book an agent under SaaS, under labor, under services, or under some new category that's still waiting for an acronym.
The honest organizational answer, today, is, nobody owns the roster, because nobody agrees on what the roster is.
And here's the part that isn't in any of the analyst [00:07:30] reports yet. Hiring a contractor used to mean bringing someone in to perform a specific scope of work. The contractor made decisions inside the scope, the company's leadership made decisions about the scope. That division of labor worked because contractors were slow. They billed in hours. They asked questions. They waited for direction. An agent isn't slow. An agent makes decisions inside and about the scope, at machine speed, every time it's invoked. The HR playbook didn't account for that. Neither did the procurement one.
Between the lenses. Which means the next logical question is about the invoice. You can't manage a workforce you can't count, and you can't budget for a workforce you can't classify. The innovation story this quarter is less about what agents can do and more about who is trying to own the ledger that tracks them.
Lens Two. Innovation and Market. The Cost Question. What does it cost to run a workforce [00:08:30] you've never managed before, and on whose line does it sit?
The major platforms this quarter have positioned themselves as the place your agents live, work, and get paid. Salesforce, at TrailblazerDX two thousand twenty-six this week, unveiled Headless three sixty, the next generation of the platform, as Salesforce's Joe Inzerillo described it, built around the fact that coding agents have just gotten really good. The platform now exposes more than sixty new Model Context Protocol tools and thirty or more preconfigured coding skills, with Salesforce framing Slackbot as the front door to the agentic enterprise that knows your users, knows your agents, and orchestrates them to get work done. Workday announced the Agent System of Record. Microsoft has Copilot Studio. SAP has Joule. Oracle has its agentic platform. ServiceNow bought Moveworks. IBM has [00:09:30] Watson X Orchestrate. Each of them is pitching a control plane for the agent roster.
The pitch is converging on the same idea, a system of record for your digital workforce, governed the way you governed the human one.
Which is where the employee-versus-contractor analogy stops being a metaphor and starts being an accounting problem.
Here's what nobody is quite saying out loud. An agent has, functionally, the profile of a contractor. It arrives for a specific scope of work. It's hired through a vendor rather than added to headcount. It's paid by the hour or the task, or in the agent's case, by the token or the outcome. It doesn't get benefits. You don't have to fire it, you let the contract lapse. ManpowerGroup's TAPFIN makes the comparison explicit. Traditional contingent workforce strategies focused on sourcing and managing human labor, contractors, freelancers, and gig workers. But as AI-powered agents take on roles that [00:10:30] mirror human tasks, organizations must redefine what contingent means. Their recommended response is almost verbatim the contingent workforce playbook that's been running for fifteen years, build cross-functional governance that includes procurement, IT, HR, and legal, ensuring digital labor is held to high, human-level standards.
The contingent workforce market is real money. Allied Market Research values the global contingent workforce management market at one hundred seventy-one and a half billion dollars in two thousand twenty-one, projected to reach four hundred sixty-five and two-tenths billion dollars by two thousand thirty-one, growing at a compound annual growth rate of ten and a half percent from two thousand twenty-two to two thousand thirty-one. That money buys vendor management systems, managed service providers, rate cards, compliance frameworks, and a whole industry of specialists whose job is to keep contractor and employee on opposite sides of a dotted line that the IRS actually cares [00:11:30] about.
Now overlay the agent wave on top of that infrastructure. Bain and Company is blunt. If an agent replaces a human task, customers will expect to pay based on outcomes, not log-ons. The fundamental shift is to stop charging for access and start charging for work done. IDC's FutureScape Worldwide Agentic AI two thousand twenty-six Predictions goes further. By two thousand twenty-eight, pure seat-based pricing will be obsolete, with seventy percent of software vendors refactoring their pricing strategies around new value metrics such as consumption, outcomes, or organizational capability. Forrester's Kate Leggett, quoted in the same piece, describes the transition as a gradual rebalancing. A CIO with a subscription for one hundred seats could swap ten or twenty of those seats for consumption-based or outcome-based pricing. Vendors will likely offer different licensing tiers or flex options. [00:12:30] Consultants could come in, manage the agentic AI implementations, and charge a percentage of the outcomes.
Read those forecasts through the HR lens and the picture gets uncomfortable. You are no longer buying software. You are hiring a staffing agency and paying per job done. The vendor you used to call your CRM provider has, functionally, become your contingent-labor provider, except the contingent labor doesn't go home at five and doesn't get a ten ninety-nine.
And read them through the CEO's lens, and they get even more uncomfortable. Per-seat pricing was a business model that matched a world in which software was a tool humans used to make decisions. Outcome-based pricing is a business model that matches a world in which software is making the decisions. You're not paying for access anymore, you're paying for execution against your intent. Which means the CEO has a question to answer that nobody put on last year's agenda. [00:13:30] How clearly, and how quickly, can this organization communicate its intent to the things that are about to be acting on it?
Levie has a name for what happens when the invoice shows up. He calls it tokenmaxxing, internal debates over how much AI compute each workflow is allowed to consume, with enterprises carefully rationing AI usage for high-priority tasks. That's not a procurement debate. That's a capacity-planning debate, and a strategic-prioritization debate dressed up in finance vocabulary. What's a high-priority task is a question the CFO can't answer alone. Neither can HR. Neither can IT. That one lands on the CEO, the COO, and the person running the P and L it's meant to move.
The CFO is still caught in the middle on the accounting side. AI Ireland's Mark Kelly framed it plainly this cycle. AI used for day-to-day operational efficiency, such as customer service [00:14:30] automation or internal process improvement, is typically operating expenses. AI infrastructure that creates durable assets, such as proprietary models or data platforms that generate new revenue, may warrant capital expense treatment. Getting this classification wrong distorts financial reporting, tax position, and ability to measure true return on investment. In plain English, the same agent, doing the same work, can end up on two very different lines of the income statement depending on who signs the contract, and which function got to the deal first.
That's a forced-consolidation setup if there ever was one. The agent-identity startups, the Model Context Protocol governance vendors, the orchestration platforms, the vendor management tools, the HR-tech companies building system of record for agents, over the next twelve to eighteen months, most of them will either be inside one of the five or six enterprise platforms or positioning themselves as the neutral control plane for [00:15:30] buyers who don't trust any single platform to grade its own homework.
Each platform wants to be the one system where that trade-off gets managed. Few want to be the second one you buy.
Between the lenses. Which brings us to the sentence nobody can get right. If the agent is a contractor, say that. If it's an employee, say that. If it's a SKU, say that. But the industry hasn't picked, because each version implies a different buyer, a different budget, and a different set of rules. The messaging isn't confused by accident. It's confused on purpose, because the purpose hasn't been decided yet.
Lens Three. Messaging and Language. The Classification Question. Are we hiring them, contracting them, or buying them? The answer shapes who signs the purchase order.
Here's a fun exercise. Take the vendor materials from any major agent [00:16:30] platform launched in the last six months and circle every noun used to describe what the agent is. You'll find them sorted into three vocabularies that would never normally appear in the same paragraph.
The first vocabulary is the employee vocabulary. Agents are teammates and digital coworkers. They have roles and skills. They go through onboarding and require training. They are managed and supervised. They have performance. Workday says agents are measured like investments, governed like employees, and improved by training and learning. Bersin describes the HR professional's new job as being the managers and caretakers of AI agents.
The second vocabulary is the contractor vocabulary. Agents are hired for specific tasks. They work on scopes. You pay them per outcome or per job. They're provisioned from a marketplace. You onboard them for a project and offboard them when the work's done. TAPFIN's framing [00:17:30] drops the analogy in cleanly. AI-powered agents take on roles that mirror human tasks, a chatbot handling thousands of customer inquiries, an AI analyst synthesizing data for finance teams. These are more than just IT tools, they are digital workers performing business-critical functions. That has a broad range of implications for HR, procurement, compliance, and enterprise strategy.
The third vocabulary is the software vocabulary. Agents are deployed into environments. They consume tokens. They sit in marketplaces with SKUs and tiers. They have renewal dates and usage limits. Salesforce's AgentExchange consolidates more than ten thousand Salesforce apps, more than two thousand six hundred Slack apps, and agents from technology companies like Notion, Linear, and Cursor, backed by a fifty million dollar AgentExchange Builders Initiative. That's a SaaS [00:18:30] marketplace vocabulary, right down to the dollar figure.
All three sets of words describe the same things. The vendors are using all three simultaneously because the market hasn't decided which classification wins, and because each classification pulls the purchase decision to a different part of the org chart.
If agents are employees, HR signs, HR owns governance, HR takes the heat when something goes sideways, and the cost sits on a labor line in the P and L. If agents are contractors, procurement signs, the managed service provider and vendor management system folks own governance, compliance takes the heat on classification, and the cost sits on a contingent-labor line. If agents are software, IT signs, software asset management owns governance, the vendor takes the heat on uptime, and the cost sits on operating expenses.
Futurum's Keith Kirkpatrick flagged the vendor-side symptom at the end of last [00:19:30] year. In a vendor-level analysis of fourteen SaaS providers, he found most vendors strongly or dominantly highlight AI and generative AI, unified platforms and data, workflow productivity, and domain expertise, and concluded that repeated use of the same language blunts its impact, and the lack of clear differentiation makes it harder for customers to understand relative benefits and evaluate which vendor is best suited to address their specific challenges. Kirkpatrick's word for it was sameness. The deeper diagnosis is that sameness sells more when the buyer hasn't decided which buyer they are.
Here's the subtler problem underneath all three vocabularies. All of them, employee, contractor, software, are task vocabularies. They describe something being purchased or hired to complete a discrete unit of work. That was the right framing when humans did the deciding and apps did the helping. It's not the right framing anymore. The agent [00:20:30] isn't performing the task. It's making the decision about the task and performing the task, in the same motion. The vendor vocabulary hasn't caught up to that, because if it did, the pitch would have to stop talking about tools and start talking about judgment. And judgment in an organization doesn't sit with HR, procurement, or IT. It sits at the top.
The language isn't bad writing. It's a hedge. The vendor doesn't know yet whether the next purchase orders for agents are going to come from Chief Human Resources Officer, Chief Information Officer, Chief Financial Officer, or Chief Procurement Officer, so the pitch is engineered to land on any of the four desks. The buyer, for their part, doesn't know who in their own org is supposed to own this, so the pitch sounds like it's addressed to the right department no matter which department reads it.
It's not a conspiracy. It's market ambiguity, dressed in brand voice. And it's going to keep working until the CEO notices that nothing in that vocabulary describes what they actually bought.[00:21:30]
The Fourth Lens. The Collision. When HR and procurement collide, it goes one of two ways, but the real question is who's steering.
Here is what the three lenses add up to.
HR didn't build itself to manage bots. Procurement didn't build itself to manage talent. Neither function is about to disappear, and neither one is fully equipped on its own for what just arrived. What happens next is not inevitable. It is a choice, and most enterprises are making it right now, whether leadership realizes it or not.
One version is a meteor. The two functions keep operating in parallel, each half-managing a workforce that doesn't fit either of their playbooks. Agents get bought without being hired, hired without being tracked, tracked without being reviewed, reviewed by no one with clear authority to act on the review. The roster grows. The invoice grows. The accountability doesn't. [00:22:30] Eventually something goes wrong that traces back to a decision nobody in the org can clearly explain, and the collision is exactly what it sounds like, destructive, expensive, and late to the meeting.
The other version is generative. HR and procurement collide like two colors of paint, or like oil and vinegar shaken into a dressing, neither ingredient disappears, but what comes out is something that didn't exist before. A function that treats the digital workforce as a first-class peer to the human one, with its own hiring logic, its own budget codes, its own performance metrics, and its own lifecycle. Call it whatever you want. Total Talent. Workforce Orchestration. Digital Talent Management. The label is less interesting than what it has to actually do.
The analysts are already circling it from three directions. Bersin has been calling his version systemic HR, all the HR processes can finally interoperate [00:23:30] together. KPMG is calling it total workforce planning, a model in which an organization that is already understaffed will opt to invest in new hires and supplement with contract and or digital labor. Deloitte is arguably closest to the operating reality. Workforce planning should no longer be considered the domain of finance or HR alone, leading organizations are extending workforce planning horizontally to include collaboration across multiple functional areas, including strategy and business operations, IT, and procurement.
All three analyses are polite about the same unsettled fact. The Chief Human Resources Officer, the Chief Information Officer, the Chief Procurement Officer, and the Chief Financial Officer each think they own part of this. None of them currently own all of it.
And none of them, frankly, are supposed to. Here's the part the keynotes don't say.
For three decades, enterprise software has been about tools. Give a human a [00:24:30] better interface, a better spreadsheet, a better ticketing system, a better CRM, and they'll make a better decision on a ten-minute or ten-hour cadence. HR managed the humans. Procurement managed the tools. Finance coded the cost. Most functions downstream of strategy operated against a direction the CEO had set, quarterly, maybe annually, occasionally in a pinch. The cadence of strategy and the cadence of execution were mismatched, but both ran slowly enough that a competent org could keep them roughly aligned.
Agents break that. An agent doesn't wait for a quarterly reset. It doesn't ask what the business is trying to do, it acts on whatever it understood the last time somebody told it. Multiply that by ten thousand agents making decisions every second, and the mismatch between the speed of strategy and the speed of execution stops being a governance inconvenience. It becomes the operating reality of the company. If the intent is fuzzy, [00:25:30] the agent workforce amplifies the fuzz. If the direction is clear, the agent workforce executes against it faster than any competitor still running humans through apps. The difference between those two outcomes isn't a tool. It's a leadership problem.
Which is why the real answer to who manages this sits above HR and procurement. It sits with the CEO. The COO. The line-of-business leaders who actually own where the company is trying to go and how it's going to get there. Those are the people who have always owned strategic direction. The change isn't that they suddenly own something new, it's that their direction is now being acted on thousands of times a second by a workforce that will not stop to ask clarifying questions. They're being asked to communicate intent at a cadence and a resolution that the old quarterly-plan, all-hands-meeting, annual-review rhythm was never built for.
Most of them aren't ready. That's not an insult. It's a structural mismatch. The tools for encoding strategy into [00:26:30] something an agent can act on in real time don't really exist yet. The muscle for reviewing what a distributed agent workforce actually did, at the end of a week, against what you wanted it to do, doesn't exist on most exec teams. The leadership job hasn't changed in principle, it has long been about pointing the organization at a direction and being accountable for the result, but it has changed in velocity, granularity, and exposure. The CEO used to set strategy and trust a management chain to translate it into work. Now the chain is getting replaced, one link at a time, by a roster of agents that translate strategy into work at the speed of inference.
HR and procurement are still going to have to collide generatively, because they're the functions that have to staff and fund the new reality. But the reality they're staffing and funding belongs to the people running the business, not to either of them. HR and procurement build the function that makes the direction operational. The [00:27:30] CEO, COO, and line-of-business leaders are the ones who have to set the direction clearly enough and fast enough that the function has something to be operational against. One without the other doesn't work. A great new workforce-orchestration function with no clear strategic intent behind it is a bureaucracy. A clear strategic intent with no function to operationalize it is a PowerPoint.
And here's the line I keep coming back to. HR and procurement are going to collide into something new to run the agent workforce. But the people steering that workforce aren't in HR or procurement, they're in the corner office, and they're about to find out that setting direction quarterly doesn't work when your workforce acts in milliseconds.
The forced consolidation coming over the next twelve to eighteen months will solve the plumbing. Most of the independent agent-identity, Model Context Protocol governance, and vendor management system for agents startups will get pulled into the platforms or niched into the cracks between them. That's a [00:28:30] good thing, fewer tools to wire together, fewer vendors to audit. But the plumbing is the easy part. The hard part, the part nobody can outsource, is two simultaneous muscles the organization has to build, a workforce function that HR and procurement collide to create, and a leadership cadence that the CEO, COO, and line-of-business leaders have to grow into.
HR and procurement are the mechanics. The CEO, the COO, and the line-of-business leaders are the drivers. Neither can get you where you're going alone.
The Hire Nobody Made.
Two functions. One kind of talent. Three sets of language. Zero agreement on whose P and L it sits on. And one strategic question that used to wait for quarterly reviews and now has to be answered in the moment, thousands of times a second, by an agent workforce acting on whatever it understood the last time somebody told it.
The workforce is being built. The [00:29:30] mechanics of who hires it and who pays for it are being worked out inside HR, procurement, IT, and finance, which is where they belong. But the question of what the workforce is actually for, what it's pointed at, what it should do when the situation changes, when it should stop, that one lives upstairs. It always did. What's new is that the answer matters in milliseconds, and the people upstairs have to learn to give it at that speed.
The organizations that figure out both halves of that equation, the function downstairs that runs the roster, and the leadership upstairs that runs the direction, will own the next decade of enterprise work. The ones that solve for one without the other will find out which one was the hard part in a way nobody enjoys.
This episode of Lens Four was written by Sean Martin and read by TAPE9, an AI audio persona created with ElevenLabs. Human-in-the-loop collaboration, written and edited by Sean Martin, voiced by [00:30:30] TAPE9.
Explore the full article and references at seanmartin.com. Subscribe to Lens Four for a weekly analysis of where business, innovation, and messaging come into focus.
Thanks for listening. Explore more at seanmartin.com.