Hey, it’s Aaron. It’s a beautiful fall morning in Idaho, and as I write this, I’m sitting in my study at home with a great cup of coffee.

Team Contio is getting ever closer to the MeetingOS launch. It’s super fun to use the product in beta. We’re adding a ton more beta users over the coming weeks and looking forward to expanding our feedback loop. We can’t wait to launch it publicly for the world to see!

I’m so excited for this product — the AI operating system for meetings is going to be a BRILLIANT meeting strategist, a SUPERHUMAN memory, and a HYPER-INTELLIGENT Chief of Staff who has you prepped, organized, and following up better than the President.

All while snapping right into your existing calendar, working with every video call platform, and working for your in-person meetings too.

One last chance for all of you who subscribe to this monthly email. If you want into the early beta group, just hit reply and tell me — I’ll pull some strings, and get you to the top of the list right away.

Welcome again to all the new folks! I love hearing from you, so send back questions, thoughts, ideas, or just a hello any time by hitting reply.

Remember, great leadership is contagious. Thanks for being here.

AI and the End of Secrets

I remember the first time that I used ChatGPT Deep Research. I wrote a pretty decent prompt, put it in deep research mode, and the AI went off for 45 minutes to think and dig around on the Internet.

When it came back with a 35-page report, I knew the world had changed.

Now — let’s be clear. ChatGPT Deep Research is a bit like a junior research analyst right out of school. I would take everything that analyst said with a grain of salt. I’d fact check anything that didn’t seem right. I’d try to guard against confirmation bias and dig deeper on anything that felt too good to be true.

But there was no mistaking the reality: $200/month for ChatGPT Pro was buying me the equivalent work of a junior research analyst, who was remarkably gifted at connecting the dots and finding meaning from little data points.

Immediately the thought hit me — is AI leading us to the end of secrets?

Secrets by Obscurity

For decades, the business world has depended on obscurity to keep secrets. If you weren’t one of the Magnificent Seven companies with reporters crawling all over you, it was possible to build new products or develop new initiatives, and make extensive progress before announcing it to your customers and to the world.

When I was CEO at Riskalyze, I’d do somewhere between 30-40 interviews a year with reporters or podcasters. Of course, I’d never leak our unannounced strategy to build that new product or enter that new market. But undoubtedly, a smart reporter or podcaster would ask a question, and wanting to be consistent with my projected self in the future, I’d make a few remarks that expressed openness to expanding in that direction.

There was no easy way to piece together the little nuggets I’d dropped on podcast X, with the data points I’d mentioned in my interview with journalist Y, with the LinkedIn groups one of our product managers was joining, and the job openings we’d posted on our careers page.

Those little micro-data points used to be very difficult to assemble into strategic insights.

Not any more.

Secrets Aren’t Dead, But Their Lifespan is Shorter

AI hasn’t made it impossible for a company to keep secrets. But it has dramatically shortened how long most secrets stay secret.

Strategic roadmaps? Easy to spot via hiring patterns, public messaging, interview comments, and partner activity.

Algorithmic breakthroughs? Models can now reverse-engineer behavior, scan research exhaust, look at newly-released API docs, and piece together working theories that are “close enough.”

Your latest pricing test? Already scraped, modeled, and mimicked by someone who’s never even logged into your admin panel.

AI doesn’t need a leak. It just needs breadcrumbs—and businesses leave those behind constantly. Stray comments, a shift in social media post wording, code commits, API behavior, conference chatter, shipping data, and patent filings…it’s all discoverable, siftable, and analyzable now.

The Formula

So, if you’re working on a secret project, how do we think about its discoverability in this new era?

I think the formula is something like this: Discoverability = Exposure × Observability × Inference ÷ Countermeasures

• Exposure: How many surfaces are emitting signals? Think repos, contracts, public demos, hiring trails.

• Observability: How machine-readable and real-time are those signals?

• Inference Power: This is where AI shows up—connecting dots that humans would miss.

• Countermeasures: The technical, legal, and operational steps you take to reduce discoverability.

AI dramatically increases inference and observability. That means our job as leaders is to reduce exposure and strengthen countermeasures—especially when secrets need to hold long enough to deliver value.

What Secrets Hold Up?

Not all secrets are doomed, and here are some categories of secrets where tomorrow’s edge still lives.

• Air-gapped work: If the bytes never leave, the model never sees them. Simple.

• Private, non-replicable data: Exclusive data sets, sensor networks, or customer flywheels no one else can copy.

• Privileged compute or fabrication: Custom chips, secured firmware, or closed-loop manufacturing that stays dark to the outside world.

• Compartmentalization: If no single system or person knows the full picture, even a leak gives partial signals.

• Execution-embedded secrets: The supply chain trick, or the weird process nuance that doesn’t live in code.

If you want durable advantage, don’t rely on keeping an idea quiet. Rely on making it hard to replicate, even when people know it exists.

The New Playbook: Secrets in the Age of AI

I suspect we’re in for a huge amount of change in how companies begin to protect their competitive secrets in the AI era. Corporate environments are going to start looking a lot more like military or intelligence ones.

We’ll see a lot more “default private” architecture — no public logs, limited CI/CD mirroring, zero egress by default.

We’ll see compartmentalized teams and credentials — Steve Jobs was famous for setting this up at Apple. Nobody has more than they need, and the blast radius from leaks is limited.

We’ll see more companies red-teaming their footprint regularly — using AI agents to scan what an adversary could infer from their public data, and building internal OPSEC bots that automate detection of leak vectors in meetings, Slack threads, and third-party integrations.

We’ll see organizations engaged in creating noise and misdirection — posting generalized or even misleading job descriptions, shipping decoy branches, or obfuscating API behaviors.

And most importantly, organizations will begin designing execution for speed — working to ensure that even if the idea leaks, they’re already 50% to 75% of the way ahead of any competitors seeking to fast follow.

The Bottom Line

AI collapses the cost of connecting dots. Inference is practically free. Scraping is trivial. Pattern detection is instant. That means your secrets — especially the idea-level ones — have a half-life compared to the past.

But that doesn’t mean you’re out of weapons. It means your strategy shifts.

From keeping ideas hidden, to making them hard to copy. From strategy secrecy, to speed of execution. From hoping for silence, to assuming fast leaks and winning anyway.

If someone connects the dots, make sure they’re still months behind.

That’s the new game. Play to win.

Talent on the Prowl

I know of an incredible person who is newly available — her background is in HR, people operations, and program management — she’s someone who is incredibly great at supporting teams, building process, and helping people feel seen and supported. She’s worked in both in-person and remote roles. If you’re ready for the magic touch of this individual in your organization, shoot me a note back and I’ll make an intro!

Best Reads + Listens of the Month

• My Podcast with Morgan Housel. I know I’m a bit biased on this one, but the brilliant Morgan Housel agreed to do our 15-minute “Worst Meeting Ever” podcast, and as usual, his insights were out of this world. Give it a quick listen on YouTube, Spotify, or Apple Podcasts!

• Resiliency and Scale. You all know by now that I’m a huge Ben Thompson fan. His analysis at the intersection of technology, business, and geopolitics is incisive, smart, often contrarian, and usually ends up right on the money. His piece tying together the recent AWS outage and Chinese export controls on rare earths was an absolute gem and will change how you think.

If I can help, let me know.

That’s it for this month! As a reminder, I am coaching five CEOs or leadership teams. That’s all I have time for, but I’m having fun doing it, and it’s allowing me not to pay myself a lot from the new company while we are in the build phase. 😉 

I don’t think I could have been nearly as effective for my team without the help of a CEO coach. There’s just something about having someone who has walked that journey before as a “safe” sounding board who you can speak freely to.

At the moment, my coaching slots are full, but I am building a wait list as I have two slots that have been taken temporarily. If you want to discuss, click here to learn more about my approach, and hit reply if you’d like to talk about it.