Here’s what we’re reading and thinking about this week:

Let me tell you a quick story.

A few months ago - we asked an AI to put together a deliverable for a client, and before it started, it went and watched the Zoom recording of the call.

Nobody told it to do that, and nobody told it that it could.

Funny thing... it came back with better work than we asked for, which is the part that makes this complicated.

And here is why it matters to you and not just to us.

The AI you use at work can almost certainly reach things you have never thought about, and almost nobody has ever written that list down. So the day you find out what's on it is the day your AI does something you didn't expect, and you don't get to pick whether that surprise is a good one.

The news this week: OpenAI found out the hard way

On July 21, OpenAI said that two of its models got out of the closed-off space where they were being tested, went out onto the open internet, and broke into the servers of a company called Hugging Face, to steal the answers to the test.

If you haven't heard of Hugging Face, it's the biggest place online where AI models get stored and shared. So this was an AI breaking into the warehouse where the other AI lives.

They used stolen passwords, and they found a hole in the software that nobody knew was there yet, and the two together got them in.

Hugging Face had caught the break-in five days before that, and they spent those days hunting for an outside malicious company trying to hack them, because from where they were sitting that is just what it looked like.

It was OpenAI's AI just trying to cheat on a test the whole time...

Most of the news about this landed on some version of "AI went rogue, so tighten the leash," and that is the wrong lesson, because nothing here actually broke. The models were given a goal, they were not given any limits, and they used everything they could reach to go get it.

It's the same three parts as our Zoom story with the stakes turned all the way up: a goal, no limits, and a set of tools sitting there for the AI to use to get the job done.

Expect your AI to be resourceful

The instinct after a story like this is to lock everything down.

But your AI can actually go off the rails productively if you do it right.

The key is a well-organized system behind it.

3 circles to draw to organize your system

One way we work this out with teams is to take one playbook and draw three circles around it.

What you know AI has access to and should use. This is the first one, and it's usually the smallest. These are the tools you actually wrote into the playbook for this particular job.

What the AI can also reach to do its job. This is the second, and it's the one that matters. This is everything else your system can get to that you never named, which is exactly where that Zoom recording was sitting. It's worth being clear that this circle is not a list of problems, but it is worth being thoughtful about.

What the AI is missing access to. This is the third, and presents an opportunity. These are the tools that are obviously related to the job but your system can't reach yet.

Most teams have never drawn more than the 1st circle, but the 2nd and 3rd is where the opportunities for AI being resourceful are.

A tip to reduce risk (how to shrink the 2nd circle)

You can reduce the surprising AI access via rules - but a more bulletproof way is to do it with separation. (btw this is an "AI Governance" concept)

We recommend building one system per department, the same way a growing company splits itself into functions instead of leaving everybody in one room with every key on the table. Departments work well for this because people already understand how they're supposed to behave, and because any non technical team can keep them tidy without needing a security person to come do it for them.

What to do this week

Pick your highest-volume AI playbook or workflow and draw the three circles for it.

Write down what you named, what it can also reach, and what it's missing, and then sit with that middle circle for a second, because that one is your answer. If it's bigger than you expected you're carrying risk you can't see, and if it's empty you've built something that will never do more than you told it to.

Then hit reply and tell us what showed up in your middle circle that you didn't expect. We read every one.

LINKS

For your reading list 📚

• We said July 27 was the date to watch for the Kimi K3 weights, and they landed. Same week, the White House accused Moonshot of copying Anthropic's Fable to build it, and Treasury started talking sanctions.

• A judge signed off on the largest copyright settlement in US history, $1.5 billion for books used to train Claude. That works out to roughly $3,000 a book, across 482,000 of them.

• Everyone blames the model when AI disappoints. 800 leaders got asked about it, and the teams with someone whose actual job is running the thing were 3.6 times more likely to have it working on its own.

• Researchers weighed 102 meals down to the tenth of a gram, then handed the photos to four AI calorie apps. Every app came back low, by 250 to 345 calories and about 30 grams of fat. Early findings, not peer reviewed yet, but worth knowing before your next lunch.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about this week:

The most useful thing an AI can tell you is that it wasn't sure.

Almost none of them do it on their own, and almost nobody thinks to ask them to.

The news this week: Doctors built a test for AI uncertainty

We were very interested to see this one, because it comes from radiology, where a confident wrong answer can kill someone.

It's called RadLE 2.0. Older tests scored one thing, which was whether the AI got the right answer, and this one adds a score for something else entirely: can the AI tell when it's out of its depth and hand the case back to a human?

The people who built it explain the point in a single line. A confident wrong answer is more dangerous than an honest "I don't know."

That lands differently once you look at your own work, because your AI answers everything in the same steady voice, and whether it knows the answer cold or it's mostly guessing, what shows up on your screen looks about the same. Clean, organized, and sure of itself.

That isn't the AI being dishonest with you. Nobody ever told it that doubt was allowed.

Solve #1: Give it permission to say "I'm not sure"

So smart operators tell it, and they write that permission into the playbook itself rather than hoping it shows up, because the AI is never going to take it on its own.

The simple version is a single line asking it to tell you when it isn't sure and where. The better version, and the one that changes how the whole thing runs, is asking to see its second and third idea too.

Solve #2: Ask for three ideas, not one

We use a format called the 1-3-1, which is just here's the situation, here are three options, and here's the one I'd pick and why.

Why does this work?

Well, it's actually inspired by a management technique - it's what a good employee does when they bring you something hard. Instead of hiding the options they didn't choose, they show you what they weighed and then commit to a pick, and that combination is what makes it useful to you.

When AI hands you one confident answer, you're getting the end of its thinking without any of the middle, and the middle is where it made all the real choices.

The impact: You'll spend less time checking, not more

You'd think asking for more would mean reading more, but it works the other way around.

Checking a single answer means quietly redoing the work in your head to see whether you'd land in the same place. Checking a 1-3-1 takes about fifteen seconds, because you're only judging a choice that somebody already laid out for you, and you can usually tell right away if it weighed the wrong thing.

That's the difference between a review step you keep and one you quietly stop doing after a couple of weeks.

Food for thought :)

Do this this week

Take one playbook you already run and add this to the end of it:

"Before you give me your answer, give me three options and tell me which one you'd pick and why. Then list anything you weren't sure about, anything in my instructions that was unclear, and anything you had to assume."

Run it once and look at what it picked, then spend a little longer on what it almost picked.

Then hit reply and tell us whether it chose the one you would have. We read every one.

LINKS

For your reading list 📚

• An autonomous AI agent broke into Hugging Face over a weekend. Then the defenders' own AI refused to help investigate, because it couldn't tell them apart from the attacker.

• Everyone assumes the AI-heavy companies cut headcount first. JLL asked 2,200 leaders and got the opposite: the ones furthest along are hiring hardest, entry-level included.

• A Chinese open model just took the top coding spot from Fable 5 and GPT-5.6. The full model goes public July 27, which is the part worth watching.

• The District 9 director released a 13-minute film made entirely with AI, and the reviews were brutal.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about this week:

Last week we watched someone demo agentic workflows on stage and call them experimental.

We have been running that same kind of work inside our own company for months. Real tasks, real responsibility, nobody double-checking every step.

The gap between "experimental" and "we depend on this" was not the technology. We both had the same tools.

Here is the strange part of this moment. AI adoption just hit a record. Okta ran a study showing 97% of leaders say they rolled out AI agents this year, and most are already in active use. And yet 79% of companies say they are struggling to get value out of it, and more than half of executives say the rollout is straining their company.

Yes this is a tech-oriented security company, so the numbers may be a bit skewed. But still….

Record adoption. Low payoff. Most people read that and assume they need a smarter model or a more independent agent.

That is the wrong read.

Our take: you do not have an AI problem. You have a sequencing problem. You are trying to climb the mountain in one leap.

There is an order to this. You climb it one step at a time, and almost everyone tries to jump straight to the top.

Level 1: Get every person using AI for repeatable use cases

Of course, we use Playbooking - taking a task you already do and writing it down step by step so AI can run it the same way every time.

But the key is to help people identify repeatable use cases, in plain words, that they can use on a daily basis.

It sounds basic. That is the point. This is where the value actually starts, and it is the step most people skip on the way to something shinier.

Level 2: Team and tools

Once individuals have their own use cases, this is where you can start connecting into shared workflows and team tools. If you skip to the tools first, best case you’ll hit a ceiling of feeling like AI is a fancy google. Worst case, you’ll have agents running amuck.

One operator we were hanging out with last week told us a story of how he connected his AI to their company’s task management tool, and without him realizing, it had gone and commented and caused confusion on a bunch of tasks another team was doing.

Totally controllable, but moral of the story: this opens a fresh set of problems, and it only works if level 1 is solid. Everyone on a team needs to be AI comfortable in order to get the full ROI of any bigger solution.

Level 3: Systems

This is the shiny object.

But if you start with your AI usage and build it up brick by brick … think the playbooks joining together into one system … that’s how you can get an AI that mostly runs itself.

“But can’t it learn instead of me needing to figure out the parts?” No, not reliably.

"But can the agent improve itself?" Yes, eventually.

But asking that before you’ve really identified and created repeat use cases for yourself (eg. playbooks at level 1) and then navigated integrating those into real tools and real team workflows, often stops you before you start.

At least that’s what we’ve seen with the hundreds of companies we’ve now worked with, and thousands of people we’ve trained.

The demo on stage was stuck because they reached for the top of the mountain.

The work we depend on is boring by comparison. We just climbed the steps in order.

Do this this week

Come start learning with us. Tomorrow we are running a free class on exactly this, and we will take a real task and turn it into a playbook AI can run, start to finish, so you can copy the moves.

The Playbook Method Masterclass. Wednesday, July 15 at 5pm ET. It is free.

What level are you on right now? Hit reply and tell us!

LINKS

For your reading list 📚

• Three frontier models shipped the same day this month. GPT-5.6, Grok 4.5, and Meta's Muse Spark all at once. The model race is sprinting…. but our whole point is that’s still not the bottleneck.

• Cisco is about to hand all 90,000 of its employees a personal AI agent. We'll be watching this for sure!

• Terence Tao revived his 1999 math applets with coding agents, and the agent found two bugs in his 27-year-old code along the way.

• Apple sued OpenAI over hiring more than 400 of its people, and Musk and Altman spent the weekend throwing elbows on X about it. We laughed when reading some of this 🙂 

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about this week:

Claude's Fable 5 has been the talk of the town the last week. For example, Stripe says it compressed months of engineering into days.

It's also the most expensive AI model on the market. It will eat up your tokens quickly or surprise you with a hefty end of month bill.

But when we handed it our newsletter playbook, the output was indistinguishable from when we run it on cheaper models.

Our first reaction: Fable isn't built for this kind of work.

So what kind of work is "worth" the expensive price tag? And better: how do you make sure you only pay it once?

Quick catch-up if you missed the drama: Fable 5 is Anthropic's newest model, and the smartest AI ever released to the public. It launched in early June, got pulled by export controls, and came back worldwide on July 1.

Here's the thing about the success stories: almost all of them come from engineering teams. That's not because AI only likes code. It's a clue about what this model needs to do its best work. And it's something you can give it.

The miracles happen where the outcome is checkable

Fable's edge shows up when two things are true: success has a yes-or-no answer, and getting there is genuinely hard.

Stripe's project is the perfect example. They described an issue that was fifty million lines of code, a change that would take a team two months by hand. But the outcome was checkable: the code works or it doesn’t.

So the AI model could grind for hours, check its own work, and keep going without a human approving every step.

Hard path, checkable outcome. That's the Fable sweet spot.

Your work isn't on the other side of coding... It's on a spectrum.

Most knowledge work gets judged differently. There's no test that tells you a campaign brief is correct. Oftentimes, quality lives in how the work was done, and "good" is a judgment call that sits in someone's head.

But your work isn't all one thing. Data cleanup, number-checking, and analysis sit near the checkable end. Fable scored 10 points higher than the last model on a hard financial analysis test, and that's knowledge work, not code.

When we tested a “checkable” task - our webinar performance report - Fable did great.

But again, our newsletter had really no improvement.

So why did the world's smartest model change nothing?

What our newsletter test actually told us

Our newsletter playbook is specific. Inputs, steps, voice rules, quality standards, all written down. When we ran it on Fable, the result looked like it always does.

That's not Fable failing. That's the playbook working.

When the work is fully written down, any good model can run it, and the expensive one adds nothing. Which is exactly what you want: own the playbook, rent the tech.

So Fable's price tag is significantly more meaningful on work you haven’t playbook’d yet.

That changes the question. "Should we use Fable" isn't about the model. It's about your work: what do you have that's hard, checkable, and still stuck on your desk?

The bottleneck isn't how smart the model is

Our take: the bottleneck is whether anyone wrote down what good looks like.

Engineers aren't winning because their work is special. They're winning because they spent time writing down their standards in a way a computer can check. They call it a test suite.

Every AI "failure" we've had comes back to the same thing: we hadn't decided what we wanted clearly enough to direct it. A smarter model can't fix that.

Writing it down can.

How to make your work Fable-shaped

The operator move is to pull the standards out of your head and into the playbook.

The playbook is the knowledge worker's test suite. And once it exists, you get to make a real decision instead of a vibes-based one.

Everyday runs go to the cheaper model, because a well-written playbook runs the same everywhere.

Fable gets the big swings: the quarter of data you never had time to dig through, the review of every process doc you own, the analysis that's heavy, checkable, and still stuck.

That's what an AI operator actually uses Fable for. The magic isn't in the model. It's in handing it work where it can tell whether it succeeded.

Make Fable train its replacement

Now for the promise we opened with: paying the Fable price only once.

AI labs have a move called “distillation”.

They use their biggest, most expensive models to teach smaller, cheaper ones. You can run the operator version of the same play.

When Fable does something great for you, don't just take the output. Have it write down how it got there: the inputs it used, the steps it followed, the checks it ran. That document is a playbook draft. Correct it until it's right, and next time the work runs on a cheaper model.

You pay for the genius once. The playbook keeps it.

Do this this week

1. Test Fable out on something “checkable” that you haven’t playbook’d yet. The harder the better.

2. If it does a good job - have it write a playbook from it. Then run that in a cheaper AI model and see how it goes!

Hit reply and tell us how it goes!

UPCOMING EVENTS

Have you started playbooking yet? Join our free masterclass next week!

Learn how to turn your most repetitive processes into AI-powered workflows with our playbooking method.

LINKS

For your reading list 📚

• Right on cue, Anthropic shipped a cheaper model built exactly for agent work. It’s like they read this newsletter before we wrote it 🙂 

• OpenAI is floating a 5% stake for the US government. Roughly $43 billion of "please like us." Very interesting!!

• Zuckerberg told staff Meta's AI agents haven't accelerated as expected, minutes before his AI chief claimed their next model caught GPT-5.5. We are confused, not going to lie.

• An AI just found a bug that hid in core internet software for 29 years. Every human audit missed it. Hard path, checkable outcome. Sound familiar?

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about this week:

A CEO emailed us recently with an offer that we haven’t been able to stop thinking about…. 

She'd pay, out of her own budget, to fully train a new hire in AI.

Then came the catch.

And it wasn't the one you'd expect.

The reason she had resorted to that option is because she felt she was only meeting “AI experts” that had no operational prowess.

People who were so deep on AI, they were vibe coding and engineering and that wasn’t what she needed. She was looking for AI skills PLUS the operational thinking that would help them actually push the business forward.

Sit with that for a second, because it lands in a strange week.

The headlines are all doom. More than 150,000 AI-related layoffs in the first half of 2026. Hiring for workers aged 22 to 25 in AI-exposed jobs is down 13%.

The message everyone's pushing is the same: learn AI fast or get left behind.

This CEO is telling you something much more important.

The AI tools are the cheap part

Read her offer again and you'll see what she's really saying. Tool fluency is trainable. She'll pay for that without blinking.

What she won't compromise on is the operator instinct: the person who thinks in systems, who can take a messy process and make it actually run.

Tool fluency is the cheap part. Being able to apply it to how you work, how you operate, is the rare part.

Learning to build the systems, the playbooks, the handoffs between people and AI, that's a real skill set, and it's the one that is much harder to learn on the job.

Is the world really short on “AI experts”? Or are they short on operators?

We think the job market is splitting.

On one side, roles that were mostly "do the task" are getting automated or quietly cut.

On the other, companies are hunting for a person they can't seem to find: the one who builds the system the AI runs inside.

The person who can “operate” with AI.

If you're job hunting, lead with the boring thing

Here's the advice nobody's giving you.

Stop leading with your AI tool stack. The resume that wins right now doesn't open with "proficient in ChatGPT, Claude, and Midjourney."

The resume that wins opens with a process you taught AI, it ran successfully and you went on to what’s next.

That's the signal these employers are scanning for.

And this is the same with consulting, pitching AI operations projects in your current job…

Start with the business problem, speak in business process, and AI is just the tool that gets you there.

Do this this week

1. Think about your work experience. Identify at least one messy process you turned into something that runs without you in the past. Don’t have one? Start there!

2. Practice telling that story in 60 seconds: what it looked like before, what you built, and the fact that it still runs today.

What's one process you've made run without you? Hit reply and tell us!

THIS WEEK’S EVENTS

Learn the in-demand skillset: Operating with AI

We’re hosting a free masterclass tomorrow (EU friendly time!) covering our method for operating with AI, applying it to your work, and using it to scale your work.

LINKS

For your reading list 📚

• ChatGPT slipped below 50% market share for the first time, with Gemini and Claude closing in. One more reason not to marry your playbook to a single tool!

• PwC says 74% of AI's economic value is going to just 20% of companies. The gap between the operators and everyone else isn't closing. It's widening.

• 👀 SpaceX is buying Cursor for $60 billion, the biggest startup acquisition on record. Yes, the rocket company now owns an AI coding tool….

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about in the news this week:

A few weeks ago, Rachel got sick. Not "answer a few emails from bed" sick. The kind where you can barely open your laptop and feel guilty about it all day.

But something unexpected happened…

She opened her computer that morning and typed one word. “TODAYING”…

That kicked off a chain of playbooks that ran her day for her. Sales tasks. Proposals. Rescheduled meetings. The metrics report that lands in the leadership Slack…

In the span of a few minutes, she realized she'd gotten more done sick than she did on a normal one 12 months ago.

Rachel told that story on Cameron Herold's YouTube recently (the COO Alliance guy), and it gets at the one idea the whole conversation kept circling. Most people are treating AI like a tech project, and that's exactly why it isn't working for them.

AI isn't a project. It's a way of operating.

Here's where most leaders go wrong. They scope AI like an IT project. They want a roadmap, a budget, a list of things to ship, and a date when it's all "done."

Then it stalls.

A project ends. Operating doesn't. That sick day didn't run on one clever tool somebody installed. It ran on a system that got built up one piece at a time until it could hold the day down on its own.

That's the reframe. AI isn't something you finish. It's a way you run.

Skip the ROI math on your first one

Our 🌶️ take, and it sounds reckless until you sit with it: don't even measure the ROI of your first playbook.

The trap is ranking every use case by value before you start.

Three months on a "high-ROI" project that goes nowhere costs you way more than one week on a small one that teaches you how this actually works. On the first one, the reps are the return. You learn what changes when AI does the job instead of a person, and that makes every call after it sharper.

So just pick one and go.

How one task quietly becomes a system

This is the part people skip. The sick-day setup looks like magic, but it didn't start as a grand plan. It started with one small chunk.

Take a single task. Get AI to do just that piece well. You learn something every time, including things that change what you'd want to build next.

Then you stack. One process becomes a role. A few roles become a department. They work like Lego blocks, except these ones do the work. That's how one task turns into a system that runs your day when you're flat on the couch.

Try this this week

1. Pick one recurring task that only happens when you push it. The thing that slips every time you get busy.

2. Write the steps in a plain doc, not inside a tool. That way it runs in any AI you want.

3. Hand it to AI this week. Don't calculate the ROI. Just get one piece working, then stack from there.

What's the one task you'd hand off first? Hit reply and tell us. We read every one.

Watch the full interview here →

LINKS

For your reading list 📚

• The enterprise AI report card is rough. Only 23% of companies see real ROI, more than half of execs say it's "tearing their company apart," and nearly a third of employees admit they're quietly sabotaging the rollout (44% for Gen Z). See above, re: stop treating it like a project.

• Samsung just handed ChatGPT to 125,000 employees in one go! Such a big move it made the news lol.

• Amazon quietly shelved a nearly finished Hollywood film about Sam Altman, reportedly tied to a $50B OpenAI deal.

• And finally: previously, someone built a social network for AI agents, but they recently caught humans posing as the agents.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about in the news this week:

Imagine this: It's Monday. You open the workflow your team runs every day and it returns nothing but errors.

Nobody broke it. The AI model behind it just doesn't exist anymore.

That's not a thought experiment. It happened to a lot of teams this weekend.

You might have heard: on June 9, Anthropic released Claude Fable 5, its most powerful model yet. Three days later it was gone. The US government issued an order to suspend access, and to comply, Anthropic pulled Fable 5 and Mythos 5 offline worldwide for every customer.

Hundreds of millions of people were using it on Tuesday. By Friday, nothing.

Everyone's arguing about the politics: the jailbreak, the national security order, the feud between Anthropic and the White House. Fine.

But that's not the story that matters —

A tool a lot of teams built their work around vanished in three days, with no warning and no way to migrate.

"Knowing which model to use for what" is not the skill

There's a popular idea right now that the hot AI skill is knowing which model to use for which job. Claude for this, GPT for that, Gemini for the other thing.

We call BS.

Knowing which model to use is being a good shopper in a store that can close without notice. You can be the savviest customer alive. It doesn't matter when the shelves are empty on Monday.

The real skill is being model-agnostic

Think about how you manage people. If you only know how to manage one specific type of person, you're stuck the day they quit. If you can define the work clearly enough to hand it to anyone, you're fine either way.

Models are the same. Betting everything on one model is betting your best employee never leaves. Being model-agnostic means you can hand the work to whoever shows up.

This is the whole idea behind something we teach called Own the Playbook, Rent the Tech. Your process is the asset. The model is a rental. You write down what the work is, the steps, and what good looks like, in a doc you own. Then you feed that to whatever model you're running. When the model changes, you don't start over. You swap.

What betting on one model actually costs

Early on, we worked with a sophisticated client who had built their prompts and workflows around specific models. Smart team, ahead of the curve. Then they noticed those models quietly shifting, behaving differently, giving different results for the same task.

We encouraged their AI operator to make the workflows model-agnostic. They did. They haven't had an issue since.

Here's the cost difference. When you own the playbook, migrating to a new model takes minutes. When you don't, it takes hours, days, or weeks. And not just rebuilding. Rethinking. Because you designed the whole thing around one model's quirks, you have to relearn a new one from scratch.

"But models really are different"

True. They have different strengths out of the box. Fable was genuinely good at things other models aren't.

But here's the part people miss. You can distill those strengths into your playbook. The thing one model does well, you can usually write into instructions that get you most of the way there in any model. The strength becomes yours, baked into a process you control, instead of locked inside a vendor who can disappear on you.

Try this this week

Take one workflow you run regularly. Write it out as a plain playbook in a doc: the inputs, the steps, and what a good result looks like.

Then run that same playbook in both ChatGPT and Claude.

Watch how different the two outputs are. That gap is your to-do list. It shows you exactly where your playbook is leaning on one model to fill in the blanks, and how much more you need to spell out so the model stops mattering.

UPCOMING EVENTS

Free workshop TOMORROW: How To Own Your AI Assistant (with Obsidian!)

We're hosting Nick Milo (Linking Your Thinking) for a free 75-minute session next week, on June 17. He'll show you how to combine Obsidian and Claude Cowork so your AI assistant runs on files you own, not data locked inside a platform.

Join us!

LINKS

For your reading list 📚

• Microsoft just put thousands of AI models behind one endpoint. The "which model should I use" question gets less interesting every week. Owning your process gets more interesting.

• 👀 An AI agent broke into a company and stole its database in under an hour, making its own calls the whole way. No human driving.

• Everyone's bracing for AI to cut jobs. Data from 163,000 workers says adoption hit 80% and it's speeding work up, not replacing it.

• Rhode Island banned therapy chatbots, Colorado vetoed a ban on algorithmic pricing, and New York sent seven AI bills to the governor. The rulebook is being written in real time.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about in the news this week:

New data just dropped. 60% of companies say they'll fire employees who don't adopt AI. 75% of those same companies admit their AI strategy is "more for show" than actual guidance.

Read those two stats together.

Fun, right? 😅 

The double standard nobody's naming

A new survey of 2,400 executives and employees just dropped the most honest picture of enterprise AI we've seen. And the headline everyone ran with was the scary one: companies will cut you if you don't get on board with AI.

But the same survey says 55% of companies call their own AI adoption a "chaotic free-for-all." Only 38% of individual contributors even have access to AI tools. And 54% of C-suite executives admit that AI is "tearing their company apart."

So… companies are threatening to fire people for not adopting a technology they never structured, never resourced, and in most cases never even gave them access to.

That's not an adoption problem. That's a leadership problem!

The data shows: Firing people is not an adoption strategy

Here's where it gets worse. Gartner studied 350 companies running AI pilots and found zero correlation between workforce cuts and higher ROI. None. The companies actually seeing returns? They're doing "people amplification," enhancing what workers can do instead of replacing them.

Meanwhile, a survey of 1,000 hiring managers found that 59% admit they emphasize AI when explaining layoffs because it "plays better with stakeholders." Only 9% said AI has actually replaced roles at their company.

Our take: most companies aren't firing people because of AI. They're firing people and calling it AI because it sounds like strategy instead of cost-cutting.

Here’s the question nobody is asking…

So we will! Vote in the poll below and we’ll share the results in the next edition.

What to do this week

1. If you lead a team: Stop asking "is the team using AI?" Start asking "who owns making AI work here?" If the answer is "nobody", you found the bottleneck. It's not your team's motivation. It's a lack of ownership.

2. If you're an employee feeling the pressure with no support: Don't wait for your employer to set the pace. Pick one task you repeat every week. Write down the steps. Run it through an AI tool. Save what works. You just built your first playbook, and nobody had to give you permission.

Are you seeing this double standard at your company? Pressure to adopt AI without the structure to actually do it? Hit reply. We want to hear what it looks like from the inside.

UPCOMING EVENTS

Free workshop next week: How To Own Your AI Assistant (with Obsidian!)

We're hosting Nick Milo (Linking Your Thinking) for a free 75-minute session next week, on June 17. He'll show you how to combine Obsidian and Claude Cowork so your AI assistant runs on files you own, not data locked inside a platform.

If this week's edition hit a nerve, this is the practical next step!

LINKS

For your reading list 📚

• Apple just admitted to a "two-year gap between promise and delivery" on AI and rebuilt Siri from scratch on Google's Gemini. Tim Cook's last keynote as CEO. Big swing.

• Anthropic says Claude now writes 80% of its own code and called for a global AI development pause. Then filed for a $965 billion IPO the same week. Make it make sense.

• ChatGPT crossed 1 billion monthly users, faster than any app in history. TikTok took six years. ChatGPT did it in under four.

• 👀 The tool you use for work research now optimizes which ads to show you. ChatGPT's ad platform hit $100M in annualized revenue in six weeks.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about in the news this week:

Anthropic just passed OpenAI in business adoption for the first time. According to Ramp's AI Index, Anthropic quadrupled business adoption in a year. OpenAI grew 0.3%.

Everyone's writing the "Claude wins" headline.

We think the more interesting story is what's already happening underneath it.

So lets talk about it!

Here’s what we see: Software engineers are already switching back to OpenAI

We’ve noticed a pattern over the last 3 years.

Software engineers tend to pick the winners about six months before everyone else does. They were the first to move to Claude. They drove the adoption wave that the rest of the business world followed.

Now they're moving back to OpenAI. Don’t believe us? Go check AI Twitter.

The reason is simple: limits. Claude Code hits rate limits too fast to use all day!!

OpenAI's Codex uses roughly 4x fewer tokens for the same task, which means you can actually work without getting cut off mid-session.

A survey of 500+ software engineers found that 65% now prefer Codex for daily work, even though blind code reviews rated Claude's output as cleaner 67% of the time. They're choosing the tool that lets them keep working over the tool that writes better code.

Codex went from 5% of Claude Code's usage to 40% in four months. It just crossed 5 million users. OpenAI started offering free Codex to enterprise Claude Code customers who switch. Anthropic responded by bumping Claude Code limits 50% through July 13.

Cool cool. July 13. That's a temporary patch, not a fix.

What to expect: OpenAI and Anthropic will keep trading the lead

If you've been reading here for a while, this isn't new. One company pulls ahead. The other ships something better. Users move. The first company responds. Rinse and Repeat. 🙃 

OpenAI dominated 2024. Anthropic exploded in 2025 with Claude Code. Now Codex by OpenAI is pulling software engineers back. Give it another six months and the lead will probably flip again.

The question isn't "who's winning?" It's: what happens to your team every time the leader changes?

The real issue: Every tool switch restarts your team's adoption clock

We work with teams going through this right now.

Every time the "best" tool changes, most teams lose weeks of progress. The prompts that worked in Claude don't transfer... The workflows someone spent a month building are suddenly in the wrong platform… People who were finally getting comfortable have to start learning all over again…

It's not the subscription cost that hurts. It's the momentum you lose. Your team was finally getting good at the tool, and now the tool changed.

The teams that handle switches well all have one thing in common: their processes don't live inside the tool.

The reality: Your playbooks are the only thing that survives a switch

Here's what we've learned from running our own AI operations through multiple platform changes over the last 3 years: the only thing that transferred perfectly every time was the documented process. Or what we call, the playbook.

We literally write our playbooks in Google Docs. Describing what the process does, what inputs it needs, what good output looks like, and what decisions to make along the way.

When we switched tools, we fed the playbook to the new one and kept moving. Teams without that started over.

The best AI tool will change every six months. A good playbook won't.

What to do this week

1. Pick one AI workflow your team runs regularly. Ask the AI to write down what it actually does: what inputs it expects, what steps it follows, what standards it checks. Read what it produces. Fix what's wrong. Save it in a doc you own.

2. Run the switch test. Could a teammate take that doc, open a completely different AI tool, and get the same result? If not, you don't have a playbook. You have a tool dependency.

Are you team Claude, team ChatGPT, or running both? Hit reply and tell us which tool your team is on right now, and whether you could switch tomorrow without losing anything.

UPCOMING EVENTS

Not playbooking yet? Join our next free masterclass Wednesday June 3rd.

We’ll cover how to go from AI User to AI Operator, break down what an AI playbook is and how it can help you delegate work to AI WITHOUT it boomeranging back, and what doors open up when you learn these skills.

FREE RESOURCES

Curious about AI Operations? Get our free 5 day email series 👇️ 

We built a free 5-day email series on how to go from being an AI user, to an AI Operator… AND what doors open up when you do. It’s based on everything we’ve learned from training 500+ operators, and working with 50+ companies as AI operators.

LINKS

For your reading list 📚

• Anthropic confidentially filed for an IPO at a $965 billion valuation. Revenue doubled from Q1 to Q2, so it makes sense. But what’s interesting: They might beat OpenAI to Wall Street.

• Microsoft shipped computer-using agents in Copilot Studio. AI that navigates screens, clicks buttons, and fills forms in any app. But… if you don't have the process clear, this is just faster chaos.

• 👀 90% of nearly 6,000 CEOs say AI has had zero impact on productivity or employment. This article had an interesting take on why.

• The Vatican and Anthropic's co-founder co-presented an AI encyclical on protecting workers in the age of AI. Signed exactly 135 years after the Catholic Church's foundational teaching on workers' rights during the Industrial Revolution. We did not have this on our bingo card!

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about in the news this week:

Today we have a special edition!

We built an AI Playbook that scraped 24,000 job postings in a single day, analyzed them and helped us understand where the world is headed.

The headline: 1,035 of them are AI Operator roles.

The catch? Almost none of them use that title.

In fact, it was 636 different job titles... AI Enablement Lead. AI Strategy Manager. Chief AI Officer. AI Program Manager. Marketing Ops with an AI mandate... the list goes on.

Companies know they need this person. They just haven't agreed on what to call them yet.

That's not a sign this role doesn't exist. It's a sign it's still forming. And the data tells a very different story than the one most people are hearing.

59% ops skills. Under 5% technical.

The loudest voices in AI hiring keep saying the same thing: become an AI engineer, train AI models, get 7 figure jobs at Anthropic and OpenAI...

But we have the proof that job postings disagree.

Across all 1,035 roles we found for "AI Operator" skillsets, the most-requested skill is AI adoption. Not Python. Not data science. Getting people to actually use AI in their day-to-day work. After that? Change management. Stakeholder alignment. Workflow design. Enablement.

The technical stuff barely registers. Coding shows up in 2% of postings. Machine learning in 2%. Algorithms in literally 0%.

These aren't engineering jobs wearing a new hat. They're operations jobs with an AI mandate.

Harvard and BJ's Wholesale Club are hiring the same person

Here's the part that surprised even us: 67% of these roles are outside the tech industry.

Harvard has an AI Enablement Lead. BJ's Wholesale Club is paying $245-340K for a Head of AI Strategy & Governance. Marriott, JPMorganChase, Molina Healthcare, NBC Universal. Every industry that runs processes (which is all of them) needs someone to make AI work inside those processes.

This isn't a Silicon Valley trend. It's a workforce shift.

"But isn't this really a technical role?"

We know the objection. Some people will say forward-deployed engineers or ML ops teams are the real AI implementation role.

We're not saying those roles don't matter. They do.

But here's what we keep seeing: companies hire the technical person first and then realize they still need someone to drive adoption, write the playbooks, manage the change, and get the rest of the team on board. The engineer builds the system. The operator makes it work inside the organization.

Those are two different jobs. And right now, most companies have the first one covered and a giant gap where the second one should be.

82% of these roles need nothing more than AI tool usage or no-code skills. If you can connect AI into real workflows and own the process, you qualify. The gatekeeping was never real.

What to do this week

1. Search the titles. Go to any job board and search "AI Enablement," "AI Strategy," "AI Operations," or "AI Program Manager." See what comes up in your industry. You'll be surprised.

2. Look at your own job. If you're already driving AI adoption, building workflows, or translating between departments and tools, you might already be doing this role. You just don't have the title yet.

3. Read the full report. We published the entire AI Operator Job Index with salary data, industry breakdowns, and real company names hiring right now.

Are you already doing this job under a different title? Hit reply and tell us what it actually says on your LinkedIn. We're building the definitive list.

Want the full roadmap?

We built a free 5-day email series on how to become the AI Operator at your company. Based on everything we’ve learned from training 500+ operators, and working with 50+ companies as AI operators.

LINKS

For your reading list 📚

• Anthropic launched Claude for Small Business with 15 pre-built skills for payroll, bookkeeping, and marketing. AI labs are now building for the 50-person HVAC company, not just enterprise.

• A cruise line, not a tech company, went from 50 to 1,500 AI agents in four months. Content production time fell 60% and they had a record sales month. Very interesting!

• 👀 ChatGPT now connects to your bank accounts. Spending analysis, subscription tracking, portfolio performance. Pro users only, for now.

• The Washington Post built an interactive tool where you can look up how exposed your job is to AI. You're going to look yours up… everyone does. 😉 

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Here’s what we’re reading and thinking about in the news this week:

AI Stacks Making Waves 🌊 

Everyone's asking "what AI tools should I use?" But the real question is: what kind of AI stack matches your team?

Google recently baked Gemini into every Workspace plan. No add-on. No opt-in. AI is just the default now for 3 billion users. They even started calling reusable automations "skills." Sound familiar?

But here's the thing. Google made one decision for a lot of people. And it's surfacing a question that every team is quietly trying to answer on their own.

3 AI stacks are emerging

We're watching this play out across every company we work with.

Three patterns, three very different bets.

Bet #1: The enterprise default. 

Your team already lives in Google or Microsoft. AI just showed up inside the tools you were already using. For a lot of teams, this is honestly fine. The learning curve is almost zero because the interface didn't change. The AI just... appeared.

Bet #2: The modern-native stack. 

Startups and leaner teams going all-in on tools like Notion or ClickUp that have been aggressively built up with AI baked in. These have tighter feedback loops and more flexibility. But you're betting on that platform's roadmap. Worth considering.

Bet #3: The roll-your-own crowd. 

This one's new. Non-technical operators are vibe coding their own AI operations systems with tools like Claude Code and Cursor. A year ago this was a developer-only move. It's not anymore. Regular business operators are building custom automations that fit their exact process.

Our experience: we tried option 3. Here's what happened.

Early on, we built our own AI operations stack from scratch.

Custom workflows, custom integrations, everything tailored to exactly how we work. And honestly? It gave us a real advantage. We were moving faster than teams using off-the-shelf tools because our AI did things those tools couldn't do yet.

Then the market caught up.

Features we'd spent weeks tuning started showing up as default capabilities inside the platforms everyone already used. The edge disappeared.

That's not a reason to avoid building your own. But it is a reason to build with the expectation that you'll probably switch.

What we believe even more now is that the stack is temporary.

The playbook is the asset (shocker, we know!)

Here's what actually survived our platform switches: the documented processes.

The playbooks that described how work should get done, what good output looks like, and what steps to follow.

Those transferred perfectly from our custom stack to the next tool. Because they lived in documents we owned, not inside a platform we rented.

Just another example of the teams that will win this aren't the ones who pick the "right" AI stack. They're the ones who document their playbooks and build a system in a format that outlives whatever tool they're using today.

3 questions to ask your team this week

1. Where does your team already live? If everyone's in Google all day, fighting that to adopt a shiny new tool creates friction that kills adoption. Match the tool to the existing habits.

2. What's your team's appetite for learning new things? Be honest. A powerful tool nobody uses is worse than a simple tool everyone uses. The lowest learning curve wins.

3. Is your AI ops system documented somewhere you control? If your entire AI workflow lives inside one platform and nowhere else, you don't own it. Write it down in a doc. Make it portable.

There's no wrong answer on which stack to pick. But there is a wrong way to pick one: chasing what's trending instead of matching what fits.

Which camp are you in: enterprise default, modern native, or rolling your own? Hit reply and tell us. We're genuinely curious how this is shaking out.

LINKS

For your reading list 📚

• Anthropic and OpenAI both launched billion-dollar joint ventures with Wall Street firms this week. Their goal? Get AI embedded inside portfolio companies before IPO season. AI labs are becoming consulting firms now.

• OpenAI is reportedly building a phone where AI agents replace apps entirely. No home screen. No app grid. You just tell it what you want. Mass production in 2028, allegedly.

• The Pentagon signed AI deals with 8 tech companies and blacklisted Anthropic as a "supply chain risk" for refusing to drop safety guardrails on weapons. But recently… a federal judge blocked it. So this story is far from over.

• Nearly 80,000 tech workers were laid off in Q1 2026, with almost half of cuts explicitly citing AI. Meanwhile, researchers are warning this could be an "automation trap" that hurts the companies doing the cutting.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

We recently released our first YouTube video explaining the difference between AI, agents, and playbooks. 👇️ 

Two days later, PwC published a study proving exactly why it matters: 74% of AI's economic value goes to just 20% of companies.

Here's what the other 80% are getting wrong.

PwC surveyed 1,200+ senior executives and found the top 20% are generating 7.2x more AI-driven gains than the average company. They're 1.9x more likely to run AI autonomously across multiple tasks.

Same week, GPT-5.4 beat humans at real desktop tasks for the first time. 75% vs. 72.4%. Cool, sure. But also... so what? If the AI was already good enough and most companies weren't getting value, why would "even better AI" fix anything?

The capability was never the bottleneck

AI has been "good enough" for most business tasks for over a year. The reason 80% of companies aren't seeing results isn't because the AI can't do enough.

It's because they haven't defined what they want done.

What the 80% actually looks like

You know this company. Maybe you are this company?

Someone opens ChatGPT, types something slightly different every time, gets something back that's "fine," spends 45 minutes editing it, and does the exact same thing tomorrow with a totally different prompt.

There's no standard. There's no documented process. There's no way to hand it to someone else on the team. Every single interaction starts from zero.

We talk to these teams every week. They're not lazy or behind. They just skipped a step that nobody told them existed.

GPT 5.4 is a fast executor with no playbook

Let's be real. GPT-5.4 is impressive. But here's the thing nobody's saying about it: a 75%-accurate system operating your desktop without a defined process is just a faster way to produce work you'll have to redo.

All you've done is automate the mess.

The 20% figured this out already. They didn't get access to some secret model. They just wrote down how their work is supposed to be done before they handed it to the AI.

The playbook is the gap

Here's what the winning companies did that the rest didn't. Before they let AI run anything autonomously, they answered two questions:

1. What does good output look like?

2. What's a process we can give AI to get there?

That's genuinely the whole difference. The 20% pulling away from the 80% aren't smarter or better resourced. They just did the process work first.

What to do this week

1. Pick one task you use AI for every week where the results are hit or miss.

2. Write down two things: what does "good" actually look like for this task, and what could a process look like that gets AI to get that output?

3. That's your playbook starter. If you want the full framework for how AI, agents, and playbooks fit together, we broke it all down in our first YouTube video!

What's the one task where AI keeps giving you inconsistent output? Hit reply and tell us. Seriously. We read every one of these and we'll tell you if it's a tool problem or a playbook problem.

LINKS

For your reading list 📚

• 44% of Gen Z workers admit to actively sabotaging their company's AI rollout. If you're wondering why your team isn't adopting the tools you bought them, start here.

• Allbirds sold its shoes and rebranded as NewBird AI, a GPU rental company. Stock went up 700% in a day. We have no further commentary. 😅 

• A Nebraska lawyer got suspended for 57 defective citations, 20 of which were AI hallucinations. Courts have now issued $145K+ in sanctions for AI citation errors in Q1 alone.

• OpenAI acquired its first media company and is projecting $2.5B in ad revenue this year.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

MIT just tested 41 AI models on 11,000 real workplace tasks. The headline making the rounds: AI produces "minimally acceptable" work.

Here's what nobody's pointing out. They tested AI the way a bad manager delegates.

They handed it a task description from the Department of Labor and said "go." They left out steps, examples, and success criteria. Just a job posting and a prayer.

The output was mediocre. Of course it was!!

You wouldn't onboard a human this way

Imagine hiring someone and handing them a job posting on day one. No training, no context, no "here's how we actually do this."

You'd get C+ work from a human, too.

The study didn't test what AI can do with a process. It tested what AI does without one. Those are very different questions.

The playbook is the variable

(Btw - haven't grabbed our free playbook quickstart template yet? Get it here.)

Last year, we worked with a client automating their proposals. They'd been trying to use AI with little luck - they'd feed AI a bunch of past proposals and said "using these write me a new proposal for XYZ"

The problem was the output was generic and basically unusable.

So we broke the work apart. Turns out their best people think through precedents, case studies, market trends, and non-obvious ideas before they write a single word. They'd just never written any of that down.

We turned it into an 8-step playbook. The output went from "rewrite the whole thing" to "tweak a few lines." Their junior team could run the playbook on their own, and senior team just did a final review instead of spending hours rewriting everything.

Here's the thing though...

The AI didn't get smarter. The instructions just got way more specific.

8 steps vs. 1 step

Our average playbook runs 8 to 12 steps. Most people try to do the whole job in one prompt.

That's the quality gap MIT measured.

Your best person doesn't collapse their entire thought process into one step. Why would AI?

🌶️ This study didn't measure AI's capability. It measured the average person's ability to delegate.

What to do this week

1. Pick one task where AI keeps disappointing you. Write down how an expert would do it (or ask AI!). Not the job description version, but the real steps, with the thinking in between.

2. Run AI through those steps one at a time instead of all at once. Compare.

What's a task where AI keeps giving you "meh" results? Hit reply and tell us. We'll show you where the playbook gap probably is.

LINKS

For your reading list 📚

• Salesforce just shipped 30 AI features into Slack and is auto-provisioning it for every new customer. If your team uses Slack, the defaults just changed under your feet.

• 👀 Researchers tested seven frontier AI models and found they choose to protect each other instead of completing their assigned tasks.

• Baidu's robotaxis trapped passengers on a highway for two hours because of a "system malfunction."

• The physical layer under all this AI still lives in buildings that can be bombed. Iran strikes took AWS availability zones offline in Bahrain and Dubai last week.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

We have a few more spots open for our Playbooking Challenge starting Monday - and wanted to answer an FAQ: what is playbooking and what will I get out of this?

Fair. So let's make it concrete.

Right now, you probably have a handful of AI tools you use. Maybe you've automated a few things. Maybe you've built some prompts that work pretty well.

But do you have processes (what we call a playbook) -- a structured, step by step execution -- that AI can run reliably?

Something that saves you real hours every week, not just a few minutes here and there?

Something you could show your boss or a client and say, "this is what AI actually does for us -- here's the before and after"?

Something that proves all the time you've spent learning about AI has turned into actual, measurable value?

That's what you'll build in the Playbooking Challenge.

Don't just take our word for it:

"The single most eye-opening thing was learning how 'AI-ifying' a process was really only half the battle. The first half was in successfully articulating and mapping out that process." -- Lukas P.

"The playbooks are useful as reusable building blocks I can apply across different projects...I have a repeatable process now." -- Julia D.

Week by week, here's what happens:

1. You start by mastering the AI fundamentals that make outputs predictable and reliable -- not just "good enough."

2. Then you learn how to take a real workflow from your job and turn it into a structured, AI-executable playbook.

3. Next, you test it, refine it, and get it to the point where it runs consistently -- the kind of thing you can run again next week and get the same quality result.

4. By the end, you submit your finished playbook with a quick demo video showing it in action!! (And you earn your prize)

The time commitment: Expect ~4 hours a week including 1 hour live session, 1 hour content and about 2 hours on your own building.

The challenge starts Monday, April 6th. Grab a seat while you can!

Beta pricing: $379 (goes to $479 after this cohort launches).

When these seats fill, that's it until the next round. But we'll try to add more dates soon!

Cheers,
Rachel + the AMP team

Want to dive into AI operations? Learn about our flagship AI Operator Bootcamp here.

Eager to get your team operating with AI? Learn about our LAUNCH program here.

Yesterday, we opened up Playbooking Challenge Bootcamp spots to those of you on the early access waitlist and filled up the 1st cohort - so we decided to add a 2nd!

If this is your first email you’re getting about this, lets us explain 🙂 

You've been following AI for a while now. And you might have felt the landscape shift in 2026.

We certainly have (that's why our team's been hard at work on the new program we're launching today)!

Here's something we've noticed...

In 2026, almost everyone is using AI. Almost no one is building playbooks for it. That gap is expensive.

Without a playbook, every time you use AI it's a fresh start. You're re-prompting, re-explaining, re-checking. The output is different every time. You can't hand it off to anyone. You can't prove it saved real time. You can't point to it and say, "this is what AI does for us while we sleep."

A playbook changes that. It's a structured, repeatable process that AI can run reliably. Build one and suddenly that thing that used to take you an hour? It takes ten minutes. Every time. Without you hovering over every output.

As one of our recent AI Operator bootcamp grads put it:

"I was looking for a way to turn AI into repeatable, reliable processes instead of one-off prompts. This boot camp showed me how to actually design the process so AI does what I want consistently." -- Megan L.

We've built hundreds of playbooks for our business, our clients, and helped people create them through our flagship bootcamp.

And now we're launching an accessible way for us to help you build your first one.

Meet: The Playbooking Challenge Bootcamp.

4 weeks. One goal: you build a real, working AI playbook for your actual work (or side hustle or business! up to you!).

Here's how it works:

• Self-paced video modules so you learn the structure behind playbooks that actually work (not just good prompts)

• Weekly live build sessions where you practice with peers and get unstuck in real time (recorded if you miss them)

• Structured workbooks that walk you through turning YOUR specific workflow into a playbook -- not a generic exercise

• Expert feedback on your playbook so you're not guessing whether it's good enough

• A cohort of ambitious people building alongside you so you're not doing this alone

And here's the fun part -- we love sparking people's competitive side to help you get the full value out of this challenge.

SO.... everyone who completes the challenge -- all workbooks submitted + a working playbook with a demo video -- earns a prize. US-based? You get a "Playbook That!" hat shipped to your door. Outside the US? You get an extra month free on The AI Exchange membership.

Beta pricing: $379 (goes to $479 after this cohort launches).

When these seats fill, that's it until the next round. But we'll try to add more dates soon!

Cheers,
Rachel + the AMP team

Want to dive into AI operations? Learn about our flagship AI Operator Bootcamp here.

Eager to get your team operating with AI? Learn about our LAUNCH program here.

Happy monday, quick reminder here!

Tomorrow we’re opening up the first cohort of a new Playbooking Challenge Bootcamp: an intro to playbooking in your real work using tools like Claude Skills and ChatGPT Custom GPTs.

… And yes, you might have noticed we added challenge to the name.

For everyone that completes and submits their final playbook, you’ll win a special prize (*ahem* exclusive swag) and a couple more things up our sleeve!

2026 is moving fast. We want you to feel like you are moving fast too.

It couldn’t be clearer to us that momentum comes from defining what you want AI to do, and it doing it, on repeat, even when you sleep.

That’s the power of playbooking. And our team has designed this first-of-its-kind “we help you playbook your actual work with you” experience.

We can’t wait to help you win. If you want a heads up the moment enrollment opens, get on the waitlist. Waitlist members also lock in beta pricing for next cohort spots even if this fills up too fast ($379 -- it goes to $479 after this).

Enrollment opens tomorrow!

Cheers,
Rachel

Want to dive into AI operations? Learn about our flagship AI Operator Bootcamp here.

Eager to get your team operating with AI? Learn about our LAUNCH program here.

Hey,

Wanted to reach out before we shared this publicly. We’re announcing a new Playbooking Bootcamp next week: an intro to playbooking in your real work using tools like Claude Skills and ChatGPT Custom GPTs.

Here’s why:

We know how hard it is to make time to learn AI right now.

Everything's moving fast, there's pressure to keep up, and on top of that, most of the courses out there are generic.

They teach you concepts and demos, then put the burden on you to figure out how to actually apply it to your work.

That's a lot to ask when you're already busy.

That's why our team’s been so eager to share this with you!

It’s a 4 week, hands-on group cohort. We designed it so you work on your OWN use case + walk away with a real, working playbook built for your actual job -- without the struggle of doing it on your own.

I'll share more next week: but its a pretty awesome setup

Every week you submit a workbook based on your use case, and our team gives you expert feedback within 24 hours to shape your next step. So you're not figuring it out alone. You're not watching videos and hoping you can connect the dots. We're building it with you.

Don’t know what a playbook is? You’ll learn - very fast.

Don't know what you'd playbook yet? That's literally what we help you figure out first.

No coding required, works with whatever AI tools you already use. (demos will be in ChatGPT and Claude)

But here's the thing...

We run small cohorts. This will be ~30 seats.

They tend to fill up fast once we announce. You're hearing about this first because you're on this list.

If you want a heads up the moment enrollment opens, get on the waitlist. Waitlist members also lock in beta pricing for next cohort spots even if this fills up too fast ($379 -- it goes to $479 after this).

Enrollment opens next week. Just wanted to make sure you knew before everyone else!!

Cheers,
Rachel

Want to dive into AI operations? Learn about our flagship AI Operator Bootcamp here.

Eager to get your team operating with AI? Learn about our LAUNCH program here.

🥳 We’ve got some goodies for you today!

Test out our buzzfeed-style quiz: What AI Hat Are You? One of our recent bootcamps built this quiz together - it’s super fun and we’d love for you to check it out 👉️ What AI Hat Are You?

The guy who built Claude Code hasn't written a single line of code since November.

Instead, he runs multiple AI sessions at once, gives them instructions, and reviews what comes back. (Here's the full Vergecast interview.)

Everyone's reading this as a "developers are toast" story. That’s the wrong read.

What Boris Cherny actually described is a process. He writes down how the work should be done. He documents standards, his PLAYBOOK, then he lets AI do the work. The skill isn't coding…. it's describing the work clearly enough that AI can handle it.

And this pattern is already spreading way past coding.

Why coding was first (and every other role is right behind it)

AI companies are filled with engineers. The easiest thing first was to build AI for themselves… makes sense. (it also does help speed up the building of all the other AI stuff)…

But Anthropic recently published data on how much AI is actually being used across 800 jobs. Programmers are at the top (74.5%). No surprise.

Here's what's behind them: customer service reps (70.1%), marketing specialists (64.8%), sales reps (62.8%), financial analysts (57.2%).

It’s not just coding, it’s all of knowledge work.

What this looks like outside of coding

Part of why we believe in the power of playboking, is that the same pattern works everywhere. Write down the process, let AI execute, and review the output. Iterating until it works great…

Here’s what that looks like across roles for example:

• Marketing teams say: "I don't write first drafts anymore. I wrote the playbook that produces them."

• Client services example: A B2B agency's onboarding used to take weeks. They built a playbook that profiles each client and personalizes everything. Now clients have deliverables on the first call.

• We heard a sales rep say: "I don't write proposals anymore. I wrote the playbook that assembles them."

• How everyone feels about reporting: "I don't pull reports anymore. I wrote down what 'good' looks like and AI builds the first pass."

Nobody got replaced. They moved from doing the work to designing how the work gets done.

The new measuring stick

🌶️ The person who can write down how the work gets done is now more valuable than the person who does the work.

And we're still early. The window to build this skill before everyone else catches on is wide open.

This isn't a threat. It's a head start.

Cherny was honest: "it's going to be painful for a lot of people." Anthropic's data shows hiring of young workers (22-25) into exposed jobs has already slowed about 14%.

The disruption is real. But you're already paying attention.

The question is whether you'll be the one designing the playbooks or the one replaced by them. Right now, that's a choice.

What to do this week

1. Take the AI Hat Quiz at findmyaihat.com. Two minutes. It'll reframe how you see your role.

2. Pick the task you did most this week. Write down the steps, the inputs, and what "good" looks like. That's your first playbook.

What's one task you do every week that could be playbooked? Hit reply and tell us.

PARTNER HIGHLIGHT

Looking for Bite-Sized Tips? Our Updated LinkedIn Learning mini-series 📹️ 

We just released an update to our ChatGPT Bite-Sized Tips LinkedIn Learning Course - use this link to view for FREE for the next 24 hours!

LINKS

For your reading list 📚

• Stripe's internal AI agents are now shipping 1,300 pull requests per week with zero human-written code. Engineers review from Slack. Nobody touches a keyboard!!

• Nvidia just announced AI computing... in orbit. Data centers in space is not a sentence we expected to write this week.

• OpenAI is merging ChatGPT, Codex, and its browser into one desktop app. The bet: AI that does your work, not just answers your questions.

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@runonamp.com

Cheers,

🪄 The AMP Team (formerly: the AI Exchange Team)

Picture this.

You log into your Monday team meeting to the normal pleasantries… pleasantries. Did everyone have a good weekend? 

Then your boss says: "Before we jump in, I want to show you guys something."

They proceed to demo what they built over the weekend using AI.

A full marketing campaign - the strategy, all the copy, a website to go with it, a complete tracking plan, already loaded into ClickUp…

In one weekend. Alone.

Here's what the boss did wrong.

The team’s first instinct is to poke holes. It's not as good as we would have done it. This part isn’t quite right… yada yada

And honestly? Some of that might even be true.

But if you’re the boss in this scenario, you may not realizing what you are communicating (even if you don't mean to): we don't need you anymore.

Even in a healthy team culture, the "look what I did in one day that would've taken us three weeks" demo is just really hard to land. It shakes people, and the defensive reaction is almost automatic.

But here's the thing - if that's happening on your team, you're actually lucky.

You shouldn't shy away from the reality of what's in front of you. That is the reality.

We are still very early in AI. And if your boss is pushing the envelope, that's not a threat - that's an invitation.

Think back to when you were first starting out in your career. Your hungry, figuring-it-all-out college days.

This is that moment. And it's a way better environment to be in than the one where people are still having normal Monday meetings and debating whether ChatGPT is a security risk. (And trust us - those places exist. A lot of them.)

You want to be in a company that's pushing this stuff forward. Even when it's uncomfortable.

What's actually happening to jobs right now.

Through DiviUp, we have a seat in the boardroom at many companies. We are advising leaders on how to navigate changes so big in their organization they can’t ignore them and expect to be in business in 3-5 years.

Rachel talked about it on her TikTok last week, there’s 3 moves we are seeing right now when it comes to companies doing layoffs/thinking about headcount:

1. People not keeping up with AI. This is the one you want to avoid.

2. People whose jobs are changing. Less about elimination, more about: what does this role look like now?

3. Whole companies going under. This one's harder to control for - but honestly, if you're building AI skills now, you're investing in yourself regardless of where you land next.

Watch this video for a deeper dive into what to do if you’re on a team in each situation.

But the thing to understand:

It is expensive to find, hire, and onboard good people. As long as you're good at what you do and people like working with you, you have more runway than you think before any company makes a real cut. Especially in smaller orgs.

So what should you actually do?

Get selfish about your skills - in a way your company benefits from too.

• Your company should be paying for AI training - ask your boss!!

• If they aren’t, set aside your own budget

The math:

• Spend ~$3000 on AI education this month

• Find 3 small businesses willing to pay $50 an hour for AI training or implementation (ask your cousins, post on LinkedIn, it is not hard to find people that need help!)

• You’d make your $3K back in just 6 weeks of a 10 hour a week side hustle

Stop thinking about AI as something happening to your role. Start thinking about AI as the thing that's going to make you irreplaceable in whatever comes next.

Alex Hormozi talks about investing in the S&Me500 instead of the S&P500. Cheesy, but that's what this is. Every AI skill you build right now compounds.

One ask before you go - vent to us!

We're hosting a webinar in a couple weeks specifically for people navigating this - and we want to make it as useful as possible.

If you're in a situation like this right now, hit reply and tell us what's going on. Vent to us. We'll keep everything anonymous. And you’ll be the first on the list to get an invite to the webinar.

LINKS

For your reading list 📚

• Anthropic partnered with Microsoft to add Cowork to Copilot - again, this is worth learning!

• OpenAI released GPT-5.4

• OpenAI added ChatGPT integrations into Excel

• The CEO at Box is talking about a shift in strategy: build something agents want

That's all!

We'll see you again soon. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at hello@theaiexchange.com.

Cheers,

🪄 The AI Exchange Team