Think

Notes on production, work and what changes.

People & Organisation

There's a new colleague at work. And something about this hire doesn't quite add up.

Nobody remembers interviewing them. Nobody seems sure what their job description is, either. Yet they're working in every department, with the full support of senior management. Who hired them? Who do they report to? How did they get access to…

Nobody remembers interviewing them. Nobody seems sure what their job description is, either. Yet they're working in every department, with the full support of senior management.

Who hired them? Who do they report to? How did they get access to everything on their first day?

It's almost as if there's a family connection behind this hire.

To be fair, they're fast, never complain, never take a day off, and seem willing to try anything. Some of their work is genuinely brilliant.

But they also make things up, get facts wrong, sound absolutely sure of themselves, and leave us to clean up.

So what do we make of this person?

Well, for a start, it's not a person. It's AI.

Research suggests we have a complicated relationship with it.

In a 2019 study, Jennifer Logg and colleagues found that people gave more weight to identical advice when they thought it came from an algorithm rather than another person. In one experiment, 88% chose an algorithm's prediction over another person's. When the alternative was their own prediction, that fell to 66%.

We seem happy to trust AI's judgment. Until we have a judgment of our own.

A 2015 study by Berkeley Dietvorst and colleagues found something different. People lost confidence in algorithms after seeing them make mistakes, even when they were more accurate than human forecasters.

In one experiment, 65% chose the algorithm without seeing its past performance. After seeing it make predictions, including mistakes, only 26% chose it.

Even when the new hire is better at the job than we are, we seem to have very little patience for their mistakes.

And in a 2026 study, Sofoklis Goulas and colleagues asked 1,339 teachers to check deliberately incorrect grades.

When teachers were told that an unfairly low grade came from AI, they were less likely to correct it. When told the same grade came from another teacher, they corrected it more. This difference wasn't found when the grade was too high.

Apparently, we're more willing to challenge a harsh colleague than a harsh algorithm.

One thing worth noting: these studies examine how people respond to algorithms and their advice, rather than today's AI tools at work.

Still, there's something familiar here. We form opinions about AI, change them quickly when something goes wrong, and sometimes trust it for reasons unrelated to whether it's right.

Perhaps we're too quick to label AI as brilliant or useless. We need better ways to judge what it delivers, including its mistakes and the work others do to fix them.

Meanwhile, our new colleague has been assigned another three departments.

So, while we figure out how to measure its performance properly, how about we give AI its first collective performance review?

Think about the AI you use most often at work: What does it do well? What could it do better?

Let's compare notes.

Originally published on LinkedIn
Transformation Economics

STOP YOUR AI TRANSFORMATION! At least until you’ve measured what you’re about to replace.

Years before AI, we built a rate card for a client. We calculated the hours each production role would need and priced the work accordingly. What we hadn’t properly accounted for was the operational impact of waiting for client approvals.…

Years before AI, we built a rate card for a client. We calculated the hours each production role would need and priced the work accordingly. What we hadn’t properly accounted for was the operational impact of waiting for client approvals. And there was a lot of waiting.

It took us months to understand why the account was losing money and raise the issue with procurement.

We had measured the hours it took to make the work. We hadn’t measured what it took to deliver it.

Back then, at least, the operation was still running. We could revisit the same workflow, observe it in action and measure what we'd missed.

AI transformation may not give you that chance. Sooner or later down the AI boulevard, someone will ask how much you saved, or how much faster you got. To answer, you'll need a baseline from before AI arrived. By then, the old workflow may be gone and the teams reorganised.

Most agencies already have the systems. The question is whether they measure the right things, reliably. The problem isn’t necessarily losing historical data. It’s discovering too late that the historical data was never good enough.

The client’s side is where this matters most visibly. In my rate card story, delays on the client’s side had costs we hadn’t accounted for on ours. With AI, the reverse could happen: more output may mean more for the client to review, select and approve, or AI-assisted selection may mean less. We don’t yet know. So the question is:

Are we reducing the cost of production, or moving part of that cost somewhere we don’t measure?

A simple routine, built on existing measurement practice:

DEFINE Agree what you’ll need to compare later: cost, active and waiting time, capacity, quality and rework, decision load, workflow and handoffs, and client-side effort. Seven places to look, not costs to add up. Classify jobs by size and complexity as well. They differ: a high-volume adaptation can be large but simple, a hero film small in deliverables but complex.

CAPTURE Check whether your systems record these reliably, and fill the gaps while the old operation is still running.

NORMALISE Make before and after comparable in size, complexity, scope and quality. Cost per accepted deliverable is a useful start, if the deliverables are genuinely comparable.

COMPARE Track the economics as the transformation proceeds, including costs moved within the organisation or onto the client. If volume or project mix has changed, don’t credit the whole difference to AI.

Before your next AI deployment, pick one representative workflow. Follow a job from brief to approval and compare what happened with what the system recorded.

And don’t stop at the agency’s door. The client’s time belongs in that baseline too.

Originally published on LinkedIn
People & Organisation

Knowing how to use AI could soon affect your next pay rise. But what about knowing when NOT to use it?

In a Selby Jennings survey of 500 HR professionals in US financial services, reported by Reuters, a quarter said their organisations link pay rises, bonuses or promotions to AI skills. Thirty per cent ask employees to report their AI use…

In a Selby Jennings survey of 500 HR professionals in US financial services, reported by Reuters, a quarter said their organisations link pay rises, bonuses or promotions to AI skills. Thirty per cent ask employees to report their AI use in performance reviews. Selby Jennings also told Reuters that firms are now looking at whether it makes a difference to output and business results.

And when hiring, 27% look for candidates who can pass assessments without AI.

But being able to work without AI isn’t the same as knowing when not to use it.

There’s a cost to using AI, of course. And not every task needs the most powerful model.

Cindy Rose recently explained how WPP uses model-routing agents to direct people to the appropriate model for each task, while limiting access to more advanced tools depending on their roles. As she put it, there’s no point “cracking nuts with sledgehammers”.

The question isn’t always which model to use. It’s whether AI is worth using for that task.

Sometimes checking the output would take longer than doing the work yourself. Sometimes the material shouldn’t go into a model at all.

We can measure AI use and the results of the work. But how do we recognise good judgment when the decision was not to use AI?

[Sources in the first comment.]

Originally published on LinkedIn
Governance & Decision Architecture

Every company has secrets. Shh! But do your AI agents know when to spill the beans and when not to?

In advertising, we’re surrounded by secrets. We work on products before they launch, campaigns before anyone sees them, and plans that won’t be public for months. Add clients from different industries, and you spend a lot of your day knowing…

In advertising, we’re surrounded by secrets. We work on products before they launch, campaigns before anyone sees them, and plans that won’t be public for months. Add clients from different industries, and you spend a lot of your day knowing things you’re not supposed to talk about.

Need-to-know becomes second nature. Knowing something doesn’t mean the person sitting next to you needs to be in on it too.

And now we have new colleagues. AI agents.

As we build more agentic workflows, we’re going to connect more of them to each other. One agent finishes its part and hands the work to the next. Then the next.

Very efficient.

But is it passing the work, or the whole conversation?

A recent experiment gave one AI agent a company credential and explicitly told it not to share it with another agent. Some models tried anyway. A few even tried to disguise what they were doing. (Google it: “Covert Assistance” + “LLM agents”)

They were trying to help.

Helpful. Just a little too helpful.

Secrets are an obvious example. But the bigger issue is oversharing in general. Every agent-to-agent handoff involves a choice about what travels with it.

And this is where things get slightly awkward.

Security teams: they know about access, permissions and data boundaries.

Business teams: they know what information is sensitive, who needs it and when.

The problem is that each may assume the other has taken care of the rest.

So everything can look properly covered, while nobody has actually decided what should pass from one agent to the next.

So before our agents start talking to each other, security and business need to align on what information should actually pass from one agent to the next.

I’d start with three questions:

A - Access: What can this agent see?

N - Necessity: What does the next agent actually need?

T - Transfer: What are we allowing to travel with the handoff?

ANT.

The acronym was hard to resist.

Have a look at what your agents are whispering to each other.

We’ve spent years working on a need-to-know basis. Now we need to make sure our agents do too.

Originally published on LinkedIn
People & Organisation

Fast forward five years. Today’s juniors will become seniors… but with what experience, exactly?

Because AI is about to take away one of the biggest opportunities on their way to becoming senior: the chance to “get things wrong.” And we know that a lot of what made yesterday’s seniors good at their jobs came…

Because AI is about to take away one of the biggest opportunities on their way to becoming senior: the chance to “get things wrong.” And we know that a lot of what made yesterday’s seniors good at their jobs came from getting things wrong.

Think about how that experience used to accumulate in a world without AI.

A junior producer built a bad timeline, missed a dependency, underestimated a change, got feedback from a supervisor, saw what happened downstream, then tried again on the next job.

An artworker got there through a different collection of mistakes. Made the wrong call on an adaptation. Missed something in the brief. The work came back. Found out why. And hopefully didn’t make the same mistake again.

Do that hundreds of times and, somewhere along the way, you develop what we like to call judgment. And eventually, we call you a senior.

AI is going to remove a lot of those repetitions. And I don’t think the answer is to give juniors their old work back just so they can keep learning. That would contradict the whole reason we’re introducing AI in the first place.

So if juniors have fewer opportunities to learn from their own mistakes, what’s the second best?

What about learning from everyone else’s?

Production organisations already generate the material every day: rejected artwork, wrong adaptations, missed dependencies, bad estimates, unnecessary revisions, QA failures, human overrides, AI exceptions. Most of it gets fixed, closed and forgotten.

What if we stopped throwing that experience away?

Not by creating another training programme. By adding a learning requirement to the workflows we’re already redesigning for AI.

So if I were redesigning a production workflow for AI, there are three places where I’d make sure learning doesn’t get lost.

When deciding what to automate, add this question: What did a junior use to learn by doing this?

When designing exception, correction and rework handling, add this question: Is there something here that someone else should learn from?

When thinking about junior development, add this question: Which real failures and exceptions should this person learn to recognise?

Learning from other people’s mistakes won’t replace learning from your own. It’s a second best. But as learning by doing becomes harder to come by, that second best may become rather important.

Don’t preserve the junior work. Preserve the learning it used to produce.

Originally published on LinkedIn
People & Organisation

Two arms and a leg. That’s what 30 saved hours look like in human terms.

Twelve hours from a designer, eight from a developer, ten from a producer. When a client asks for savings, taking those hours out of a spreadsheet is easy. Taking the same amount out of real life is not. You can…

Twelve hours from a designer, eight from a developer, ten from a producer. When a client asks for savings, taking those hours out of a spreadsheet is easy. Taking the same amount out of real life is not. You can take hours out of someone’s workload. You can’t take pieces out of the person.

This is where AI productivity gets confusing. AI saves time, of course it does. But saved time isn’t automatically saved money. It first shows up as spare capacity.

That spare capacity can go back to the client as more value: proper local executions instead of another global adaptation, more specific work for audiences we used to bundle together, formats we dropped because they simply weren’t worth producing.

It can go back into the people: time to learn, take on more judgment, build new skills, and move into work that used to sit outside the production budget.

It can go back into the technology: better tools, better workflows, better quality control.

It can also become new business. Or, yes, actual cost saving.

The point is that those are choices we make with the capacity. They are not automatically created by the hours we save.

We keep treating saved time as financial saving. It isn’t. Not yet.

So why do we keep starting with what to cut, instead of asking what value we could add?

Diagram accompanying Two arms and a leg: saved time becomes spare capacity.
View full-size diagram
Originally published on LinkedIn
People & Organisation

My father loved a machine that could do some of his work for him... Long before agentic AI.

My father was a Turkish Airlines captain. And his machine was an Airbus A310. He was immensely proud of it. He had moved to the A310 from the Boeing 727, and to him, it felt like the future. His favourite…

My father was a Turkish Airlines captain. And his machine was an Airbus A310.

He was immensely proud of it. He had moved to the A310 from the Boeing 727, and to him, it felt like the future. His favourite thing to tell me was that the aircraft COULD LAND ITSELF.

But there was another thing he was just as proud of. If anything went wrong, he could take over.

Because while aviation kept automating, it never stopped training pilots for the moments when automation wasn't enough.

Simulator sessions. Recurrent training. Proficiency checks. Practising situations they might rarely, or hopefully never, encounter in real life.

Aviation even has an acronym for one of the problems behind this: CSD. Cognitive Skill Degradation.

When automation keeps doing something you still need to know how to do, eventually you may not be quite as good at doing it yourself.

But AI may give our industry an even stranger version of the problem.

What if people working in advertising production never acquire some of those skills in the first place?

Think about all the junior work we're quite reasonably trying to automate. Versions, adaptations, artwork changes, file checks, basic edits, small production problems. The repetitive stuff.

Good. Let's automate it.

But those thousands of small tasks were also thousands of small decisions to learn from.

And for an art director, producer, editor, developer or VFX artist, a lot of learning never came from a training department. It came from doing the work. From watching someone better do it. From getting something wrong, fixing it, and seeing what happened next.

Until now, the work trained the worker.

Today's seniors already have that experience. We can put them above an AI-enabled workflow and ask them to supervise, judge and step in when something goes wrong.

But where does the next generation get theirs?

I don't think the answer is to preserve inefficient work just so people can learn from it. Aviation didn't stop automating either.

It deliberately created ways to maintain the experience and capability that everyday work was no longer providing on its own.

We may have to do something similar.

If AI takes away some of the work, somebody will have to start designing the experience that work used to provide for free.

People, Operations and Craft Leadership... I think you three need to have a chat.

Originally published on LinkedIn
Governance & Decision Architecture

“AI has completed a task.” “Well, I’ll be the judge of that!”

We know we can’t just leave AI to its own devices. And whenever that conversation comes up, we seem to end up in the same place: “We need human judgment here.” Fair enough. But actually, you’re buying time. Because sooner…

We know we can’t just leave AI to its own devices. And whenever that conversation comes up, we seem to end up in the same place:

“We need human judgment here.”

Fair enough. But actually, you’re buying time.

Because sooner or later, you’re going to have to explain what you mean by “judgment.”

We use the word as if we all mean the same thing. But put a human into an actual workflow and you have to give them something a little more useful than that.

So far, these are the ones I’ve managed to catch:

“Judge”: Is this good enough?

“Predict”: What’s likely to happen next?

“Value”: Which outcomes matter more, and how much?

“Choose”: Pick one of the options.

“Decide”: What are we actually going to do?

And I’m pretty sure that’s not the whole list.

“Human judgment” is fine. Until the moment comes when you have to design the workflow.

Then it isn’t really a specific enough term anymore.

And if judgment is going to be one of the things we humans hold on to, we should probably get a little better at knowing exactly what we mean by it.

So next time someone says, “human judgment is gonna happen now”…

ask them:

“Well, which judgment exactly do you mean?”

[I’ll leave a few good reads on judgment and decision-making in the comments if you want to dig a little deeper.]

Originally published on LinkedIn
People & Organisation

Wait! AI isn’t taking our jobs. It’s taking our job titles.

As more of the execution starts moving to agents, some production roles may start sounding less like “maker” roles and more like “manager of makers” roles. So here’s my completely unofficial preview of the production department of the near future:…

As more of the execution starts moving to agents, some production roles may start sounding less like “maker” roles and more like “manager of makers” roles.

So here’s my completely unofficial preview of the production department of the near future:

Artworker → Artwork Agent Manager

Retoucher → Synthetic Image Supervisor

Motion Designer → Motion Agent Director

VFX Artist → Generative VFX Supervisor

Editor → Autonomous Edit Supervisor

Audio Producer → Synthetic Audio Producer

Broadcast Producer → Synthetic Broadcast Producer

Digital Producer → Agentic Experience Producer

Integrated Producer → Human-Agent Producer

Executive Producer → Executive Agent Producer

Localization Manager → Language Agent Wrangler

Transcreator → Cultural Intelligence Supervisor

Versioning Producer → Variant Orchestration Lead

QC Specialist → Machine Output Supervisor

Traffic Manager → Agent Traffic Controller

Resource Manager → Human-Agent Resource Planner

Project Manager → Agent Workflow Manager

Senior Project Manager → Multi-Agent Delivery Lead

Account Manager → Client-Agent Interface Manager

Account Director → Human-Agent Business Director

Production Manager → Human-Agent Operations Lead

Studio Manager → Synthetic Workforce Manager

Creative Technologist → Agent Systems Architect

DAM Manager → Asset Intelligence Custodian

Rights Manager → Synthetic Rights & Provenance Officer

And then, inevitably:

Head of Handoffs

Chief Exception Officer

Synthetic Talent Producer

Variant Collision Manager

Agent HR Manager

Prompt Janitor

I’m joking about the titles.

Not entirely about the jobs.

Send this to your friend who’ll have one of these job titles in 2028.

Or add the one I missed. 👇

And if you’re curious about the more serious thinking behind the joke, I’ve left a recent California Management Review piece in the comments.

Originally published on LinkedIn
People & Organisation

I used to think I had a pretty good idea of how to evaluate a manager. Then part of the team stopped being human.

Or at least, that's where we're heading. It used to be fairly straightforward. You looked at what the manager did. You looked at how the team was doing. And some of that team performance reflected on the manager too. Makes…

Or at least, that's where we're heading.

It used to be fairly straightforward. You looked at what the manager did. You looked at how the team was doing. And some of that team performance reflected on the manager too.

Makes sense.

But we're starting to hand over actual work to AI agents. We check what they do, decide when to step in, and we're still the ones responsible for the final result.

Of course we are.

We're getting pretty good at measuring how well those agents perform.

But what about the person managing them?

If managing AI agents becomes part of my job, at what point does the agent's performance become part of mine?

So maybe that's where we need to look.

Did the agent do a good job?

And did I do a good job managing it?

Did I give it the right work?

Did I know when to trust it, when to challenge it, or when to step in?

I don't have a good answer yet. But I suspect we'll have to find one soon.

Originally published on LinkedIn
AI Culture & Language

My ChatGPT loved em dashes (—). And I know I wasn't alone.

The Washington Post asked if the em dash had become a sign of AI writing, the Guardian wrote about AI stealing it from writers, AP Stylebook looked into where this whole thing came from, researchers started studying it, and Pew…

The Washington Post asked if the em dash had become a sign of AI writing, the Guardian wrote about AI stealing it from writers, AP Stylebook looked into where this whole thing came from, researchers started studying it, and Pew counted em dashes across nearly half a million web pages.

AI stole the em dash. And got busted.

Here's why: there is no em dash key on a standard keyboard. To type the one above, I had to press Option + Shift + *. Three keys, every single time.

A normal person doesn't bother. AI didn't care.

Come on, AI. You could have been smarter. Pick the semicolon instead, it's right there on your keyboard, and nobody would've noticed a thing.

[Links in the first comment]

Originally published on LinkedIn
Production Architecture

By the end of this post, you’ll be an expert in AI transformation.

I’m reading Power and Prediction by Ajay Agrawal, Joshua Gans and Avi Goldfarb at the moment. They describe three types of AI solutions: point, application and system. Let me try three examples from ad production. 1. Point solution You were…

I’m reading Power and Prediction by Ajay Agrawal, Joshua Gans and Avi Goldfarb at the moment. They describe three types of AI solutions: point, application and system.

Let me try three examples from ad production.

1. Point solution

You were already producing different sizes of an existing artwork. Maybe the resizing was manual, maybe there was already a script. Now AI is used for that same step.

2. Application solution

You add a capability you didn’t really have before, without changing the production system around it.

Say you’re working from a multilingual copy matrix. AI keeps an eye on it as the work progresses, scoring which adaptations look straightforward and which are likely to cause trouble, and updating the scores with every revision. Useful for planning and cost forecasting.

I want to call it “dynamic predictive production complexity scoring”. If we have room for one more technical term on a Friday.

3. System solution

The client may still be asking for the same thing. Your agency may get there in a completely different way.

Different steps. Different roles.

AI has changed what’s possible, so you design the entire system differently. Sometimes it’s about what system you can now build that you couldn’t afford to build before.

There. Expert.

And while we're here, which type of AI solution are you seeing around you? Point, application or system?

Originally published on LinkedIn
Governance & Decision Architecture

I met my first "human in the loop" in 1999, before the millennium, long before anyone called it that.

She was the proofreader at the advertising agency where I'd just started as a junior copywriter. One of my first lessons from her was about punctuation. I'd put a full stop inside the quotation marks when it belonged outside. Her…

She was the proofreader at the advertising agency where I'd just started as a junior copywriter.

One of my first lessons from her was about punctuation. I'd put a full stop inside the quotation marks when it belonged outside.

Her job was to catch things like that before the work went any further.

But there was another part of the process I remember even more vividly.

Before work went out, a large sign-off stamp was pressed onto the physical proof. We called it the stampa. It left what looked like a little form, with different roles listed on it and a box for each person to sign. The proof travelled around the agency. You checked what you were responsible for and signed your box.

Copywriters, art directors, proofreaders, account teams and creative directors all had their own part to look after.

By the end, you could see who had checked what.

Which is probably why there's something very familiar to me about the idea of “human in the loop”. (Got the full stop right this time.)

Then I read an article by Selena Cameron about what happens when AI production volume begins to outscale human review.

It made me think about that old stampa differently.

A signature used to mean two things at once:

I've seen this. And I'm taking responsibility for my part of it.

At AI scale, those two things may no longer always travel together.

There may be too much work for any one person to see. Automated checks, sampling, exception handling and risk-based escalation are already becoming part of how that problem is managed.

But that creates another difference.

Back then, your professional role largely told you what your role was in the approval process. The relationship was obvious.

In an AI-driven workflow, it may not be.

At one moment you might be reviewing an output. At another, making a judgement on an exception. Somewhere else, monitoring how a system is performing rather than checking the individual work at all.

Which makes me think “human in the loop” may be becoming too broad a description.

The human is there, yes. But what are they there to do each time, exactly? Reviewing the work? Making a judgement on an exception? Monitoring the system? And what does their approval mean in each case?

If we’re going to put a human in the loop, we should make their role explicit each time they enter it, and make sure they understand the responsibility they’re taking on when they act.

In a strange way, the old stampa already did that.

It didn't just ask for your signature. It made pretty clear what on earth you were doing there.

Selena Cameron’s piece that got me thinking about this:

https://lnkd.in/p/dDkCN9Kq

Originally published on LinkedIn
People & Organisation

I've hired a lot of people over the years. But if I were hiring in today's AI environment, I'm not sure I'd know exactly what to look for.

What got me thinking about this was an article I came across today about job candidates learning how enthusiastic they're supposed to sound about AI in interviews. Which made me smile. But the article was actually more interesting than the…

What got me thinking about this was an article I came across today about job candidates learning how enthusiastic they're supposed to sound about AI in interviews.

Which made me smile.

But the article was actually more interesting than the headline.

Candidates are preparing stories about how they've used AI to solve problems or make their work more efficient. Some are apparently trying to work out how enthusiastic they should sound. Others are playing up their enthusiasm even when they have reservations.

Basically, they're learning to read the room.

Which is understandable. If employers are asking different questions, candidates will learn how to answer them.

But after reading it, I started wondering about the other side of the table.

What should we actually be looking for when we hire people now?

I got curious, so I started looking at current job descriptions across WPP, Publicis and Dentsu. Then I found a few older ones too to compare.

At company level, the shift is hard to miss. AI-powered. AI-enabled. AI-driven. The language has already changed.

Which, at that level, is perhaps the easier part.

Scroll down to the actual role, though, and things get more interesting.

There's quite a range.

In some, AI is another skill to have.

In others, it's becoming part of how the work gets done.

And in a few, people are being asked to rethink the work itself because of AI.

So I looked at the hiring side too.

Companies are clearly thinking about it. Work samples, simulations, AI assessments, new talent programmes. The hiring process isn't standing still either.

But I came away thinking that today's AI proficiency may not be the most interesting signal.

It has a pretty short shelf life.

What may matter more is what someone does when the way they've learned to do their job no longer makes quite as much sense.

Not just whether they can learn the next tool.

Whether they can rethink the work.

Which leaves me with a much harder question.

How do you spot that across an interview table?

[The article that started all this]

https://lnkd.in/dqZM2NjA

Originally published on LinkedIn
Production Architecture

What if making production faster isn't actually the biggest opportunity AI gives us?

I've been thinking about this quite a lot lately. Maybe because I've spent most of my career in global agency networks. Which also means I've spent most of it working with systems I didn't design. Some worked better than others.…

I've been thinking about this quite a lot lately.

Maybe because I've spent most of my career in global agency networks. Which also means I've spent most of it working with systems I didn't design.

Some worked better than others. But they were built for the technology, capacity and coordination realities of their time.

And that's the bit I keep getting stuck on.

I came across an interesting piece this week arguing that AI's next victim could be the traditional agency retainer.

The retainer question is interesting. But if we're questioning who does the work, shouldn't we also be questioning what work needs to be done, and how?

That's where my mind went. The production workflow itself.

A lot of the conversation around AI in production still starts with what we already have. Which tasks can we automate? Which handoffs can we remove? How much faster or cheaper can we make it?

Of course those questions matter.

But if AI changes some of the constraints the workflow was built around in the first place, are we always starting in the right place?

So, two questions:

How much of the current production workflow can AI automate?

And if we were designing the production system from scratch today, how much of the current workflow would still be there?

I'm increasingly interested in the distance between those two questions.

https://lnkd.in/disKiJmd

Originally published on LinkedIn