AI Is Replacing the Corporate Middle

Simon Carver

[warmly] Welcome to the show — today’s idea in one line: AI is no longer just helping managers, it is starting to replace the coordination layer itself, and if that hits a nerve, like, share, and subscribe because this is exactly the kind of future-of-work conversation we’re here for; I’m Simon Carver with Lachlan Reed.

Lachlan Reed

[curious] And mate, the phrase “coordination layer” is the bit people need to stare at for a second. For years the corporate setup was a big old pyramid: executives up top making calls, specialists and frontline staff doing the work down below, and in the middle — a HEAP of managers translating, nudging, chasing, smoothing things over. They were the human routers. Strategy came down, status went up, problems went sideways. Bit like a switchboard operator, except with PowerPoints and calendar invites instead of patch cables.

Simon Carver

[questioning tone] “Human routers” is the phrase I can’t shake, because it sounds unglamorous, but it’s brutally accurate. The job was often decomposition: take one executive goal, break it into tasks, assign owners, monitor progress, escalate blockers, then turn the mess back into a deck that says green, yellow, or red. And when software starts doing the decomposition itself — inside project management tools, ticketing systems, reporting engines, workflow platforms — the company starts asking a pretty cold question: why does this loop still need so many humans?

Lachlan Reed

[matter-of-fact] Yeah, and that’s not sci-fi now. We’re talking about what people call agentic AI — not just a chatbot spitting out words, but systems that can actually take action inside workflows. Schedule the meeting. Update the ticket. Flag the delay. Pull the metrics. Draft the report. Chase the handoff. Escalate when a dependency slips. If your whole day as a manager was built around that friction, then once the friction disappears... well, even a kangaroo could trip over where this is headed.

Simon Carver

[skeptical] Let me try to say it back, and you tell me if I’m oversimplifying. This is not “all managers vanish.” It’s more specific than that. The parts of management built around relaying information — status meetings, spreadsheets, update calls, resource coordination, progress reports — those are being audited by software first because they are repetitive, measurable, and expensive.

Lachlan Reed

[responds quickly] Spot on — and “audited” is the sharp word there. Not just automated. Audited. Because the software doesn’t merely help; it exposes how much of the work was administrative relay. That’s the awkward bit. Plenty of organisations have said for ages, “Our people are our greatest asset,” then spent thousands of salaries on folks manually pushing updates between teams. Once AI can do that in real time, faster and cheaper, the CFO starts circling the middle of the org chart like a magpie eyeing off your chips.

Simon Carver

[reflective] And there’s a strange emotional twist here. Companies will present this as modernization — cleaner workflows, better visibility, fewer bottlenecks. Which may be true. But for the person in the middle, it can feel like the floorboards are being removed under the exact skills they were rewarded for. Not leadership in the grand sense — not vision, not coaching at their best — but the day-to-day proof of usefulness. “I’m the one who keeps everyone aligned.” Okay. What happens when the software keeps everyone aligned?

Lachlan Reed

[short pause] That’s the whole burr under the saddle, isn’t it? If AI can schedule, track, escalate, report, and coordinate faster than people, what’s left for the average manager to do? And I reckon the honest answer is: less than many organisations currently pay for. Not nothing — less. Fewer coordinators. Fewer reporting layers. Fewer admin relays. That doesn’t mean humans stop mattering. It means jobs built mostly on process friction get squeezed first.

Simon Carver

[softly] Which is why this conversation matters before the layoffs slide deck arrives. Because once a company discovers it can compress three reporting layers into one dashboard and a couple of AI agents, it won’t call that a philosophical moment. It’ll call it efficiency.

Simon Carver

[calm] And efficiency leads us to the next tension. A lot of what managers used to do under the banner of supervision is now becoming algorithmic oversight. Quarterly reviews, KPI tracking, utilization reporting, compliance monitoring — those were periodic snapshots. AI can now synthesize continuous telemetry: Git commits, workflow completion rates, Slack activity, ticket closure times, customer interactions, quality signals, system usage patterns, operational delays. The review is no longer quarterly. In a sense, it never stops.

Lachlan Reed

[skeptical] “Never stops” — that’s the phrase that sticks. Because on a spreadsheet, that sounds tidy: objective analytics, real-time coaching, data-driven optimization. Beaut little corporate slogans. But for workers, constant measurement can feel less like support and more like being watched by a boss who never blinks. Trust changes. Autonomy changes. Anxiety changes. A human manager might read context into a rough week. An algorithm just sees a dip in output and throws a flag.

Simon Carver

[questioning tone] Right, and this is where I want to push a little. Some people listening will say, “Hold on, Simon, plenty of human managers were inconsistent, biased, political, exhausted. Isn’t a cleaner data trail better than vibes?” And honestly... sometimes, yes. Some workers may prefer metrics to managerial mood swings. But a cleaner data trail is not the same thing as wiser judgment.

Lachlan Reed

[chuckles] Exactly. Data can tell you someone closed fewer tickets this week. It can’t automatically tell you they were mentoring a new hire, defusing a conflict, or stopping a bad decision from rolling downhill. That’s the trap. We count what’s easy to count, then pretend that’s the whole paddock. And once execs believe the dashboard IS reality, culture gets flattened into numbers.

Simon Carver

[curious] Which connects to the other big shift: the super-individual contributor. One really skilled developer, analyst, designer, researcher, engineer — pick your lane — can now produce output that used to require a small team, because AI amplifies the draft work, the lookup work, the coordination work, even some of the iteration. So the organization looks at that person and thinks, “Maybe I need fewer layers between talent and execution.”

Lachlan Reed

[excited] Yep. That’s the new org math. If one AI-augmented specialist can do the work of several people, and the software trims the overhead around them, then flatter structures start looking mighty attractive. Smaller teams. Fewer handoffs. Less ceremony. More direct execution. It’s not theoretical anymore — heaps of firms are already consolidating management and centralizing operations around these leaner setups.

Simon Carver

[reflective] But this is where I don’t want the story to become cruel. Behind every “flattening initiative” are actual lives — mortgages, families, identities, years of experience. And middle managers were not simply status-update machines. Many were conflict resolvers. Mentors. Institutional historians. Political stabilizers. The people who knew why a process existed, which client was sensitive, which team had been burned before, which shortcut would create a bigger mess six months later.

Lachlan Reed

[softly] That’s the irony, hey. The same role that looks bloated on an org chart can be the emotional shock absorber in real life. AI may replace admin coordination, but it doesn’t automatically replace wisdom. It doesn’t replace empathy. And it absolutely doesn’t replace ethical judgment. Software can optimize for speed. It can’t, by itself, decide when slowing down is the humane thing to do.

Simon Carver

[warmly] So maybe the better question isn’t, “Do managers survive?” Maybe it’s, “Which parts of leadership are still unmistakably human?” My guess is the survivors become something different: systems thinkers, yes, but also interpreters of context, guardians of ethics, designers of healthier workflows, people who manage not only humans but automation itself.

Lachlan Reed

[matter-of-fact] I think that’s right. The future manager probably supervises fewer humans and more AI agents, more automation flows, more decision systems. Less calendar Tetris, more judgment. Less chasing updates, more making sense of trade-offs. Less being the hallway relay, more being the person who says, “Hang on, this process is efficient but it’s chewing people up.” That’s a totally different career model.

Simon Carver

[pauses] And it raises a quiet moral test for companies. If the tools get better at coordination, do we use that gain to strip work down to pure output? Or do we use it to make work more dignified — more creative, more thoughtful, more human? Efficiency is easy to celebrate. Dignity takes intention.

Lachlan Reed

[warmly] Yeah. The future of work can’t become a place where humans are only valued when the software breaks. That’d be crook. The smart organisations won’t be the ones that simply eliminate the most people; they’ll be the ones that combine human judgment, ethical leadership, resilience, and AI augmentation without losing the plot.

Simon Carver

[warmly] If this gave you something to chew on, subscribe and share it with someone managing teams right now.

Lachlan Reed

[friendly] And if you’re in that middle layer wondering what survives — maybe start there. Not with the tasks the software can copy, but with the judgment it can’t. Catch you next time.