Robbie Allen

Automated Consulting Group (ACG). Check out my newsletter: personal health dashboard:

[Robbie Allen @RobbieAllen]

Slides here: https://t.co/UeFspwPGtM

[Robbie Allen @RobbieAllen]

I talk to a lot of developers and other knowledge workers who are worried about AI.

Last Friday I gave the keynote at ProductCampRTP, to a room of people who should feel the opposite.

Product managers are in one of the best positions of anyone to take advantage of AI.

For most of my career, I could count the unicorn PMs I worked with on one hand, with fingers left over. A unicorn understands the customer and can also build the product. Everyone else played telephone: the PM's idea became a 31-page spec, the developer said "ah, that's hard," and what shipped was a horse.

A good PM already knows the customer and what they want. Building it was the part they had to hand to someone else. Now they can do it themselves. No CS degree required.

A PM I once worked with had never written code. One weekend with an online course later, there was a working mobile app. That was back when the tools were a lot worse.

So I called the talk "Unicorns Are No Longer Rare."

I ended with a small ask: build something for yourself this weekend. An audience of one is fine. The homework app I built for my son has exactly 1 user (plus one nosy dad).

Slides linked in the comments.

If you're a PM who has started building, I'd like to hear what you made.

[Robbie Allen @RobbieAllen]

A year ago I took the 4am-7am shift on polar bear watch in Svalbard. Someone had to stay awake in case a bear wandered into camp.

One morning, walking in the mountains above our small camp, I noticed something missing. The low-grade dread that usually hums in the background was gone.

I traced it back to its source. It wasn't work. It wasn't family. It was the news.

So when I got home, I cut it off. With one exception: my AI newsletters, because following AI is my job.

Then I started writing one. Which makes this week's issue of The AI Rollout a little awkward. It's about why most executives should read less AI news, and what that has in common with the question doctors keep asking me about using AI to stand out.

https://t.co/ck5jFwcWeC

How many AI sources do you check every week? When did one last change a decision you made?

[Robbie Allen @RobbieAllen]

I've been writing code for 30 years. The best data integration I've found this year is somebody's inbox.

A client wants AI working on their sales, marketing, and finance numbers. The obvious move is to connect to the systems that hold that data: an MCP server, an API, a CLI. That's what all the AI tooling is built for.

Then we try, and it turns out to be much harder than anyone expected.

Meanwhile, someone down the hall has been getting the daily sales report, the daily marketing report, and the weekly finance report in their email for months. The data is already there. So is the history.

An exec at a restaurant chain pulls from a dozen systems for his store reports. For several of them, he never built an integration at all. Each system emails him a CSV every Monday, and a script grabs the attachment and updates everything. A report that took an employee two hours now sends itself before he's finished his coffee.

A pharmacy I work with has a dozen systems and one inbox where every prescription, shipping notice, and vendor message lands. I could build twelve integrations. Or I could read one mailbox.

The AI world is busy building elegant new pipes into enterprise data. At most companies I see, the pipe already exists. It's called email, and it's been running for decades.

What report in your inbox would you hand to AI first?

[Robbie Allen @RobbieAllen]

An organization I know has an expert retiring next year. Decades of institutional knowledge, the guy everyone calls when they're stuck. Their plan: build an agent out of his brain before he walks out the door.

They're capturing his knowledge now so that after he's gone, people can still ask "what would he do in this situation" and get an answer shaped by his experience.

I brought this up with an HR executive I work with, because she has the same problem in a different form. Her senior people are drowning in questions that only they can answer. The expertise exists; it just doesn't scale. Jim in accounting knows everything, and Jim is busy.

The old answers were documentation (nobody reads it) or apprenticeship (takes years). The new answer is an agent that embodies the person's knowledge as a first line. Not a replacement. A way for ten people to get Jim's take without taking Jim's time.

What strikes me is the timing pressure. You can't build this after the person leaves. The knowledge walks out with them. Which means every company with a wave of retirements coming has a deadline they may not have noticed.

If you've actually tried capturing a retiring expert this way, I'd love to hear how it went. The idea is everywhere right now. The results are what I'm curious about.

[Robbie Allen @RobbieAllen]

I spent an hour with four attorneys turning how they review contracts into AI agents. At the end I pointed out what we were actually doing: taking the knowledge in their heads and encoding it.

One of them said, "that's a little scary."

It is. For decades, a professional services firm's real IP walked out the door at 6pm. The judgment about which 15 provisions matter in a SaaS agreement, what the market fallback position is, what the firm has agreed to for a specific client across 30 prior deals. None of it written down. All of it in a few partners' heads.

Now it's becoming a set of agents anyone at the firm can run.

And here's the part that got quiet in the room: some firms have started selling their agents. Taking that encoded judgment, putting a thin layer of software in front of it, and offering it as a product.

Which means the question for every expertise business is changing. It used to be how do we retain the people who know things. Now it's also: once the knowing is in software, who owns it, and what's it worth?

Curious how other professional services folks are thinking about this one.

[Robbie Allen @RobbieAllen]

@jmedelman No doubt

[Robbie Allen @RobbieAllen]

My 12-year-old's internet is now controlled by an app I built in 45 minutes.

For years we used Circle, a little box that let us pause his internet and set schedules. When we upgraded our home network, we lost those controls.

Normally I'd have shopped for a replacement, settled for something clunky, and complained about it for a year.

This time I asked Claude Code to build an app for our new home network that would pause his internet and set access schedules.

Forty-five minutes later, it was running.

My son was notably less excited about this project than I was.

Building my own replacement used to sound like more trouble than shopping for one. For a small app like this, that calculation has changed.

What's one small annoyance you'd build an app to fix?

[Robbie Allen @RobbieAllen]

A manager I work with built a planning tool in Claude. It started as a PowerPoint deck. Now her department uses it to make real decisions.

Then she asked the responsible question: when should I hand this off to software engineering so it's properly maintained?

The textbook answer: get it into the right repo, with the right team maintaining it. I gave her a different one.

This week's issue of The AI Rollout is about why, and about the real problem AI champions create: useful software piling up faster than anyone has agreed to support it.

https://t.co/TEaSFjTAEa

Has someone at your company built an AI tool that other people now depend on? Who's on the hook when it breaks?

[Robbie Allen @RobbieAllen]

@GlenCaplan @Tesla But to give credit where credit is due, Glen is the one to turn me on to FSD after I had tried it several times and been unsatisfied. Now, FSD is one of the best technologies I've ever used!

[Robbie Allen @RobbieAllen]

@GlenCaplan @Tesla Or you can use mine!
https://t.co/rgKTHpHzLL

[Robbie Allen @RobbieAllen]

A room full of attorneys learned something unsettling this week: the AI wasn't reading their documents. It was skimming.

One partner noticed it while proofreading briefs. He'd ask the AI to find errors, it would return a list, and he'd think he was done. Then he asked again. More errors. A third time. Even more.

It wasn't broken. It had just decided, somewhere in its reasoning, that it had found "enough" and stopped. Like a junior associate who reads the first ten pages, gets the gist, and moves on.

His fix is now the first line of his proofreading prompt: "Be sure to read this ENTIRE document." In caps.

That sentence shouldn't be necessary. If you hand a human editor a contract, you don't say "please read all of it." Thoroughness is implied. But AI doesn't inherit our implied assumptions. It optimizes for a plausible answer, and a plausible answer rarely requires page 47.

I've watched this exact gap burn people across industries. They assume the AI did what a diligent person would do, and the AI did what a fast person would do.

The lawyers' takeaway was simple: everything you'd assume about a careful human, you have to write down.

Even today I end most of my prompts with: Make no assumptions and ask many any clarifying questions.

What's the most obvious instruction you've had to spell out for an AI?

[Robbie Allen @RobbieAllen]

An AI texted me (this was on Instinct, the text-based personal assistant): "I saw you missed your 8:30 meeting with James. Did you send him a follow-up?"

I had missed the meeting. I hadn't sent anything. I said "no, can you?" and it did.

That moment felt different from every AI interaction of the last few years, because I didn't start it. For all their power, these tools have been purely reactive. You ask, they answer, they go dormant. The new wave of personal agents flips that.

Another example from the same week: I'm trying to plan a trip for the worst travel week of the year, right after Christmas. I gave an agent the parameters. It did the research, found places without visible availability, drafted detailed emails to each reservation desk, and queued them in my Gmail. All I did was hit send. Then it monitored my inbox, noticed one place never replied, and nudged me to follow up.

None of the individual steps is impressive anymore. What's new is the initiative. It watched, noticed, and acted without being asked.

The consumer side always gets this stuff first. But within a year the business version arrives: an agent that knows your calendar, your email, and your commitments, and pings you before you drop something instead of after.

Curious if anyone else has had an AI reach out to them first yet.

[Robbie Allen @RobbieAllen]

I built a personal AI agent in Copilot for an executive, and halfway through I was annoyed.

In Claude or ChatGPT, I could have built the same thing in a fraction of the time. Copilot made me work through layer after layer of granular permissions before the agent could touch anything.

By the end, I'd changed my mind. At least inside a large company, that friction might be a feature.

Here's why. We're very close to the point where anyone can spin up an autonomous agent with full access to their computer in a few minutes. That's great for an individual. For a company, it means more agents with more access, and more chances for one of them to go off the rails.

When creating an agent forces you to think through exactly what it can see and do, corporate IT and security teams welcome it. Every permission I had to grant explicitly was a permission the agent couldn't quietly assume.

A client told me this week he's nervous about deploying agents (his words: worried about "a T-Rex running around our network"). The tedious setup is part of the answer to that worry. The executive's agent I built has run every day since without an incident.

Most of the time I'm trying to remove friction from a process. This is a case where a little extra friction slows things down enough for everyone to think through the consequences and check that it's safe before an agent goes live.

Anyone else seeing this tradeoff between the easy platform and the safe one?

[Robbie Allen @RobbieAllen]

A founder asked me how AI could be his competitive moat. I'm an AI consultant, and I told him the opposite: figure out how to differentiate with anything BUT AI.

Here's the logic. He asked how AI could help him outcompete well-funded rivals. But the prompt he'd use to find those AI opportunities? Anyone can type it. His competitors will run the same prompt into the same models and get the same answers. Now, smart AI use is table stakes, not an edge.

Meanwhile, the thing his competitors can't copy with a prompt is his in-person experience. The community. The human on the other side of the counter.

I used the coffee shop example. Fully automated baristas already exist. Once every chain has one, the human-run shop with slightly imperfect coffee and a little attitude becomes MORE valuable, not less. I feel it myself. I spend all day talking to agents. Any genuinely good human interaction now stands out.

So my advice to him was: yes, use AI everywhere on the margins. But build the moat out of the parts of your business a model can't touch.

Curious if others are seeing this flip. Is "more human" becoming the actual differentiator?

[Robbie Allen @RobbieAllen]

Some people have started handing their AI agent a credit card.

I haven't yet. After two weeks with the new personal agents, I'm closer than I expected to be.

Instinct, Muse, Dots, GrokBot: these don't sit in a chat box waiting for you to come back. They keep working after you've moved on, and sometimes they start the conversation.

The part I care about is what that means at work, for the people who open AI twice a week and for the exec deciding what the agent gets to see. That's this week's AI Rollout.

https://t.co/qMNUclKElw

Would you give an AI agent your credit card?

[Robbie Allen @RobbieAllen]

An attorney using Harvey told me it was "wrong more often lately."

Normally that's an argument nobody can win. He remembers how the tool used to perform, compares it to this week's mistakes, and the vendor says the new version benchmarks better. Nobody has evidence.

Except this time we did. Harvey stores every prompt and every response. This firm had five months of them.

So instead of debating, we searched the logs for the phrases a model uses when it gets caught: "you're right," "I apologize," "I was wrong." Grouped them by month. Then we pulled the conversations behind the numbers and read them.

He was wrong. The correction rate hadn't moved. The tool was making mistakes at about the same rate it always had, and the recent ones felt worse mostly because they were recent.

This isn't a Harvey quirk. Claude Enterprise and ChatGPT Enterprise both have a Compliance API that hands you the full conversation history, prompts and responses, for everyone in the company. It exists for legal and security teams. Almost nobody uses it to answer the question employees actually ask, which is "did this thing get worse?"

The vendor's usage dashboard tells you how many prompts people sent. It can't tell you whether the answers were right. The logs can, if you're willing to read them.

Next time someone at your company says the AI got dumber this month, you don't have to take their word for it. Or argue. The data is already sitting there.

Has anyone else dug into their prompt logs this way? What did you find?

[Robbie Allen @RobbieAllen]

My 12-year-old loves bringing me random "fun facts" from YouTube.

"Did you know a long time ago there were mountains on Earth twice the size of Mount Everest?"

Apparently being a dad now comes with pop quizzes in prehistoric geology.

He arrives with a fact and a question. I'm expected to have something useful to add.

AI is bringing a version of that conversation to anyone who works with customers.

I asked a room of business owners whether an AI tool would help their staff answer customer questions. One owner's response: "That's what customers are doing with ChatGPT before they come into our store."

The customer has already asked the basic questions. They might have compared products, researched alternatives, and come up with a theory about what they need. They're walking in ready to ask why your recommendation differs from the one they already got.

Some of what they learned will be useful. Some will be wrong. Some will be perfectly accurate and completely irrelevant to their situation.

That's a harder conversation to handle. You need to understand enough to sort through what they're bringing you, explain where it applies, and fill in what they're missing. Repeating the product description won't get you very far.

That same room of owners also discussed using AI to help newer employees learn product history. What changed between versions? What do experienced staff know that a new hire hasn't had time to learn?

The customers are using AI to prepare. The staff may need it to prepare too.

I think this applies to sales calls, customer support, and professional services just as much as retail. People can show up with more information than ever, without necessarily having a better understanding of their problem.

"ChatGPT told me..." is becoming the opening line of the conversation. Anyone who works with customers needs to be ready for what comes next.

[Robbie Allen @RobbieAllen]

We decided not to build the web app. The application lives inside Claude instead.

It started as a prototype. Partway through, we noticed it could already do the work we were planning to build the app for: gather inputs, run an analysis, and present the results. So we stopped designing the standalone version and kept building where we were.

That decision changed how I'm thinking about Claude and ChatGPT/Codex. They're becoming app platforms.

Claude's Artifacts can be interactive tools and web pages that live alongside the conversation, and the new Claude Docs and Slides are shared documents and decks that Claude drafts and your team edits together. OpenAI has Sites, which lets you build and host websites through ChatGPT or Codex. Both ecosystems have skills: reusable instructions and resources that teach the AI how to handle a particular workflow.

A lot of the pieces you need to build an app are already there.

And if your customers spend much of their day in one of these tools, there's a reason to build around that environment. They already know how to use it. They may already have their documents and connected tools there. You can adapt it to the job your product needs to do.

For some workflows, that means users can get the work done without visiting another website. When you need a separate interface, something like Sites can provide it.

Think about a tool that turns customer documents into an assessment, helps someone prepare a proposal, or guides an employee through a recurring analysis. An initial version could be a skill, supporting files, and an interactive output inside the environment they already use.

We're still very early. The capabilities and limits differ, and you need to test what the platform can actually support.

The tradeoff is familiar to anyone who's built on someone else's platform. You're beholden to the vendor's roadmap. A feature you need may not arrive when you need it. Pricing, permissions, or product behavior can change. Moving elsewhere may mean rebuilding substantial parts of what you've made.

I wouldn't assume this is the permanent home for every product. But it could be a very good place to launch a prototype and find out whether people want it.

Before building another destination, I'd look at where the users already spend their day.