CC:World Blog

Posted on

This Time, AI Doesn’t Have an Alibi

A poster on a display panel in a landscaped office park shows a hand-painted before-and-after: a tangled blue workflow on the left re-linked into a clean, redesigned one on the right, one step marked in mint.

The work AI is absorbing now is the work we offshored twenty years ago, for the same reason. The advantage won’t go to whoever swaps a human for a machine. It goes to whoever redesigns the work.

AI had an alibi for the vanishing entry-level job until 2025. It won’t have one for what comes next. The work most exposed to it in the short term is the one we spent the 2000s offshoring, from call centers to entry-level coding to the back office, because both waves seek the same thing: how much of a job is written down and how little of it is judgment. That work already left once. The mistake now would be to automate it in place, swapping a machine for a person and changing nothing else. The larger prize, including for the economies that first took this work, belongs to whoever is willing to rethink how it gets done at all.

Orion Playbook banner: prefer to listen? Subscribers get every issue narrated, at orionplaybook.com

A few weeks ago, I argued that AI did not kill the entry-level job, and the timeline backed me up: the decline started years before the models were good enough to matter. This piece is the other half of that story. An alibi that holds through 2025 is not an alibi for 2026 and everything after. The honest question is no longer whether AI will remove work. It is which work, how fast, and what a leader should do about it.

Start with a fact that looks like a coincidence and isn’t. The jobs most exposed to AI today are, to a striking degree, the jobs we shipped overseas in the 2000s. Call centers and first-line support. Claims processing, data entry, bookkeeping, and first-line IT. Plus, a large share of software engineering: QA and entry-level development moved to India, China, and Eastern Europe, while the architects and senior engineers stayed on staff. We have relocated this exact category of work before, and we called it offshoring. It was an arbitrage on geography, the identical work bought in a cheaper labor market. What is happening now is a second arbitrage on the same work, except the destination is not another country. It is a model. AI agents like Claude, ChatGPT, Hermes, or OpenClaw (for the brave ones) now do the tier we once offshored, and then some.

The reason the two waves rhyme is buried in how offshoring worked. To hand a process to a team 10,000 km away, you first had to write it down. Script it, break it into tickets, test plans, and service-level agreements, and strip out the tacit judgment so a stranger could run it to standard. Offshoring was, without anyone intending it, a twenty-year codification process. It turned tacit work into explicit process maps, and those maps are precisely the structure a machine needs to run it. Offshorability, it turns out, was an early read on automatability, because both measure one property: how much of a job is written down, and how little of it is judgment.

That gives a leader something more useful than dread. The test that decided what could be offshored in 2006 is the same test that decides what AI runs well in 2026. A process that is clean enough to send abroad, clearly defined, repeatable, and measurable, is a process clean enough to automate. And that clarity is also what keeps a model reliable: a well-specified process acts as a guardrail, starving the ambiguity a model would otherwise fill with confident nonsense. So one question does three jobs at once. How clearly is this process defined? is the offshorability question, the automatability question, and the hallucination-risk question, one property across three eras. The cleaner and more measurable the work, the closer AI already is to doing it.

This is no longer a forecast. It shows up first where the work was most codified. India’s IT-services firms, the companies that built the offshore delivery model, have cut fresher hiring by roughly 80% from their early-2020s peak as they pivot to AI-first delivery. The tooling is mainstream rather than fringe: by late 2025, around 90% of developers reported using AI coding tools, and over 80% said those tools made them more productive. Boilerplate, test coverage, routine tickets, the scriptable tier is exactly what those tools do best. Support and the back office are on that path too, for the reason above.

The tool, though, is only an amplifier. The 2025 DORA research found that AI magnifies whatever system it lands in: strong engineering practices compound, and weak ones only get faster and less stable. The gains come from the workflow around the model, not the model itself. The bottleneck was never the typing. It was the handoffs, the reviews, the queues, the shape of the work. Offshoring taught this lesson the expensive way: lifting a broken process to a cheaper location did not fix it; it exported the mess. Automating a broken process in place only repeats it faster. A machine bolted onto a human-shaped workflow inherits every bottleneck it has.

A cream editorial card reading "The same clarity that let you offshore the work now lets AI run it," with "same clarity" underlined in mint, above a blue rule and the signature line "Gérard Métrailler - linkedin.com/in/gmetrail".

When the machine is genuinely better, redesign is the reason. My Oura ring developed a failing battery within a year, and the entire warranty replacement was handled by an AI agent, start to finish, in two or three minutes, with no queue and no need to repeat myself to a chain of people. It was faster and cleaner than the human version had ever been. What made it work was not the bot. It was that they automated the friction, the queue, the intake, the status chase, and then trusted the machine with a real decision that cost the company money: issue the replacement. Contrast that with the reflex most firms reach for, dropping a chatbot in front of the old script. In support, AI deflects far more than it resolves; deflection rates run high while verified-resolution rates sit far lower, often in the low 40s. A deflection that never resolves is a customer giving up, the old process with a shinier bottleneck. The principle is small, and it is the whole game: automate the friction, redesign the flow, and point the human at judgment.

That redesign question matters most for the places that took this work first. India’s technology and business-process sector employs more than 5 million people and stands among the largest pillars of its economy; the Philippines’ equivalent employs nearly 1.9 million workers and accounts for more than 8% of GDP. These are national systems built on codified work, now meeting the technology that runs codified work best, and the hiring cuts are the leading edge of it arriving.

The reflex is to read that as a slow-motion disaster. I read it as the opposite. The very fluency that exposes them, running codified processes at enormous scale, is exactly the capability the AI era rewards, provided it climbs a level: from executing the process to designing, supervising, and auditing the AI that executes it. The largest concentration of people who understand how these processes actually work is an asset for building the layer above them. And markets with no legacy to defend tend to skip a generation. Much of sub-Saharan Africa never built out landlines and went straight to mobile; Kenya’s M-Pesa put mobile money in most adults’ hands while much of the West was still standing in line at a branch. An offshore delivery firm has no on-premises workflow to protect and every reason to rebuild delivery AI-first, which is a real path to leapfrogging incumbents who are busy bolting AI onto processes they refuse to change.

Pilots have a phrase for the discipline that all of this asks for: “stay ahead of the plane.” Configure now for the state that is coming, not the one that already arrived, because the pilot who falls behind the plane spends the flight reacting to things that already happened. Staying ahead here means designing for the human-and-machine combination that is arriving, rather than defending the human-shaped process that is leaving. The economics reward it. Making execution cheaper has never shrunk the amount of work worth doing; it enlarges it (economists call it the Jevons paradox), and the roles it opens, redesigning workflows, supervising models, owning the exceptions and trust, are the ones worth moving toward. The winning unit was never the machine alone or the human alone. It is the workflow deliberately built around both.

So the Monday-morning move is a small one. Pick the most codified process you own and resist the urge to pave it exactly as it is. Ask what it would look like if you designed it today from scratch for a person and a machine working together, then build one step toward that, rather than a bot in front of the old script. The work already left once. Whether it leaves you behind this time comes down to whether you are willing to redesign it, or only to re-staff it. Stay ahead of the plane.

A watercolor illustration: a line of old wooden utility poles crosses from the left and stops; from the last pole a single mint arc leaps forward across empty paper, skipping the infrastructure that was never built.
Images source: ChatGPT Images / Claude Opus / Gérard Métrailler

Sources

Originally published at www.orionplaybook.com.

Posted on

The SaaS Bill Just Split Into Two Meters

A watercolor illustration of two rising gauge dials, one small and capped, one large and still climbing, hung as a framed painting on the wall of a modern tech-startup meeting room in downtown San Francisco, with a blue pen and a laptop on the conference table below.

One meter counts how many people log in. The other counts everything the agents they authorized won’t stop doing, and the surviving architecture stacks both.

Your next software renewal will likely carry two prices, not one. One line still counts the people who log in. The new line counts everything the agents they turned loose did while nobody was watching a screen. Salesforce already reports it publicly: Agentforce revenue up 205% year over year, with the seat line holding right beside it. This is a stacking model, not a swap. It changes three things at once: what you sell, what you buy, and how fast a software bill you thought was fixed can run away from you.

Continue reading “The SaaS Bill Just Split Into Two Meters”
Posted on

Every AI Answer Is a Bet Dressed as a Fact

https://www.gmetrail.com/wp-content/uploads/error-bars-deleted-hero-dot-and-spread-notebook.png

Leadership was always about judgment under uncertainty. Now the machine is uncertain, too, with every AI answer arriving as a single number in a confident voice. Re-attaching the odds is the skill LLMs quietly made critical.

Your AI sounds equally sure whether it is right or guessing. That flat confidence is not a quirk; the same training that makes a model agreeable also makes it overconfident. Leadership has always meant deciding before all the facts are in, and the tool used to be the one certain thing on the desk. It just stopped being certain. This is the fourth cognitive lens: reading every answer as one drawn from a distribution, and supplying the calibration that the machine cannot.

Continue reading “Every AI Answer Is a Bet Dressed as a Fact”
Posted on

Your AI Agent Will Never Question the Premise

It was built to reply with the “best” existing answer. The one move it can’t make is the one now worth the most.

Ask an agent to price your product, and it hands you a polished version of what your competitors already do. That’s not a bug; it’s the objective working perfectly. A model is a machine for the most likely next thing, which means it drifts toward the average answer, the one everyone else’s agent is also producing. The expensive human move runs the other way. SpaceX did it with a rocket and found that 98% of the cost was due to inherited habit. Here’s why that move is now the edge, and why a faster agent makes skipping it more dangerous, not less.

Continue reading “Your AI Agent Will Never Question the Premise”
Posted on

When the Parts Got Cheap, the Connections Got Expensive

AI discounted all traditional knowledge work. The value didn’t vanish. It moved to the one place a model still can’t reach, and most leaders aren’t looking there.

AI has made the components of leadership work nearly free: analysis, draft, model, first-pass decision. The value didn’t evaporate. It relocated to the one place a model still can’t reach: the connections between the parts. That shift is why systems thinking, long filed under “nice-to-have,” is now the highest-return skill a leader owns. A nine-second corporate catastrophe this spring shows what it costs to keep watching the parts instead.

This is the second article in a short series on three cognitive lenses for thinking under pressure: tension, connection, and reduction.

Continue reading “When the Parts Got Cheap, the Connections Got Expensive”
Posted on

Your Hardest Problems Aren’t Problems

The ones that keep coming back to the same meeting were never yours to solve. In the age of AI agents, tensions move to the core of the job.

A company handed two-thirds of its customer service to AI, shed the equivalent of 700 support agents, and called it solved. A year later, the CEO said the words every leader dreads: We went too far. It was not an AI mistake. It was a category mistake, the same one that breaks reorgs, rewrites, and roadmaps. Many of your hardest calls were never problems with answers. These were tensions to be managed, and AI agents just turned spotting the difference into a survival skill.

This is the first article in a short series on three cognitive lenses for thinking under pressure: tension, connection, and reduction.

Continue reading “Your Hardest Problems Aren’t Problems”
Posted on

AI Has an Alibi

Everyone blames it for the vanishing entry-level job. The timeline says otherwise.

The entry-level job is disappearing, and the headline has already named the culprit: AI. Here is the problem with the case. The sharp decline set in around late 2022, and AI was nowhere near good enough to replace anyone until late 2025. A cause cannot arrive three years after its effect. So who actually did it? The honest answer is a lineup, and the most interesting suspect never makes the headline: a feedback loop nobody had on a budget line.

Continue reading “AI Has an Alibi”
Posted on

The most durable thing in your software isn’t in the code.

AI can clone a SaaS product in a week and walk an agent straight through a shallow integration. What survives are three things that compound into one another, plus the human judgment beneath them that no competitor can replicate.

A trader at Jefferies, not an engineer, named the “SaaSpocalypse” panic, which tells you what kind of event this is. Software shed close to $2 trillion from its October peak on the theory that anything can now be cloned in a week. The theory is right about features and wrong about moats. The error is picturing a moat as a wall, one thing you build once and stand behind. The durable defense is a loop, and underneath it sits the part no competitor can vibe-code, because it was never in the code: judgment.

Continue reading “The most durable thing in your software isn’t in the code.”
Posted on

The blank canvas and the seasoned eye.

Ten years ago, streaming a movie over Starlink at 30,000 feet would have sounded like fantasy. We are about to be just as wrong about AI, and the people who get the next decade right will not be the ones you expect.

We are reliably bad at imagining ten years out, and we miss in one direction: we underestimate. AI is the next thing we are underestimating. The instinct in most rooms is that the young will lead and everyone else will catch up. That instinct is wrong twice over, and the answer to why is older than any of the tools.

Continue reading “The blank canvas and the seasoned eye.”
Posted on

Your AI budget is already gone.

Editorial documentary photograph of a quiet finance executive's desk; a framed editorial illustration above the desk shows a small stack of token-shaped chips with one corner of the stack burning steadily and a thin trail of smoke rising upward, the rest of the stack intact but visibly being consumed.

Three Uber executives, three different seats, told the same story this spring. The cost category most boards govern quarterly is moving to hourly.

By April, Uber’s CTO had blown the AI budget he set in December. Three weeks later, the CEO said he was metering headcount and leaning further in. Two weeks after that, the COO asked aloud whether any of it was producing value. Three quotes, three seats, one cost category nobody had experience with. Here is why token spend breaks the quarterly cadence finance was built on, and the three questions a board should be asking by the next meeting.

Continue reading “Your AI budget is already gone.”