Late Adopter Advantage? AI Readiness - Ep. 499
In Episode 499 of Explicit Measures, Mike and Tommy dive deep into a thought-provoking mailbag question: Could being late to adopt new technology actually give you an advantage? With Episode 500 just around the corner (with some exciting giveaways planned!), the hosts explore the intersection of AI readiness, data culture fundamentals, and why we’re still in “Year One of the iPhone” when it comes to AI.
News Roundup
Tommy’s AI-Powered WordPress Fix
Tommy shared a fascinating story about his blog “Prompting BI” getting hacked. After years of trying to figure out what was wrong, he had an epiphany: Why not let Claude SSH into the server and fix it?
In just 20 minutes, Claude identified the bug, edited the necessary files, and got the site back up and running. The lesson? We now have infinitely knowledgeable coding assistants at our disposal. As Mike put it: “You have an infinitely knowledgeable coding machine behind you that will just be able to figure it out.”
Key Insight: We’re in the conductor’s era now. You don’t need to know how to play every instrument—you just need to know what the music should sound like.
OpenClaw: AI Assistants That Take Action
Mike introduced OpenClaw (formerly ClawBot, then MoltBot), an open-source AI assistant project that’s pushing the boundaries of what personal AI can do. Running on a dedicated computer with full system access, OpenClaw can:
- Manage its own memory (short-term and long-term)
- Build task lists and maintain context across sessions
- Communicate via Telegram, Teams, or other messaging platforms
- Take autonomous actions based on your preferences
The wildest example? Someone’s OpenClaw bot ordered sushi for them—using a “found” credit card from the dark web. While that’s a cautionary tale about giving AI too many permissions, it demonstrates the incredible potential (and risks) of autonomous AI agents.
Main Topic: The Late Adopter Advantage
The Mailbag Question
Sandra wrote in with a thought-provoking observation:
“I work in an organization that is late to the data game. We are just building our first data and BI solution. Listening to you, it felt reassuring to hear that many others are revisiting the same fundamentals now, especially in the context of AI readiness. It made me wonder whether being way late in our case might actually be an advantage.”
Why Being Late Might Actually Help
Mike and Tommy explored several angles on this:
1. The Fundamentals Still Matter
Despite all the AI hype, organizations are still grappling with classic challenges: data silos, ownership, governance, and data culture. Whether you’re a Fortune 500 company or just starting out, these foundational issues must be addressed.
2. Skip the Legacy Baggage
Early adopters often have years of technical debt, outdated architectures, and ingrained processes that are hard to change. Starting fresh means you can build with AI-native approaches from day one.
3. Let Others Work Out the Kinks
Microsoft’s first pass on many features doesn’t always pan out. By waiting, you let the early adopters find the bugs and edge cases while you build on more mature, battle-tested technology.
4. AI Will Be Ready When You Get There
If you’re slowly adopting Fabric and building your data foundation, AI capabilities will likely be more mature and integrated by the time you’re ready to leverage them.
The Conductor’s Era
One of the most powerful metaphors from this episode: We’ve entered the conductor’s era of technology. You don’t need to know how to play the trumpet or write PHP—you need to know what outcome you want and how to direct the AI to achieve it.
As Mike noted: “English is the final language of coding. This is the final boss. There’s nothing better than talking to a computer with what you understand.”
Watch the Full Episode
Links from the Show
Connect with Us
Got an idea or topic for discussion? Drop us a line!
Full Episode Transcript
Click any timestamp to jump to that part of the video.
Introduction & Episode 500 Preview
00:00 — Good morning and welcome back to the Explicit Measures podcast! Episode 499—we’re almost at the cusp. Unless something terrible happens, but it shouldn’t.
00:42 — For long-term listeners, we do have some free giveaways coming. We’re going to give away some free things at Episode 500, so make sure you tell your friends! Thursday 7:30 AM—that’ll be our Episode 500, which is quite a milestone.
Main Topic Preview
01:27 — Our main topic today is a mailbag discussing the late adopter advantage. Letting Microsoft kind of work out the kinks—usually their first pass on a lot of things don’t really pan out. Maybe it needs a couple passes to get it exactly dialed in. So how does this work with software? Is this actually a thing? Let’s unpack that and see if your company’s actually at an advantage right now.
News: Claude Fixes Tommy’s Hacked Blog
02:07 — Tommy shares a story about his blog “Prompting BI” that got hacked. He started it about three years ago to explore the intersection of AI and business intelligence.
03:03 — The WordPress site got hacked with articles he couldn’t decipher, and he couldn’t access the admin side. Mike and Tommy agree: WordPress has its challenges.
04:12 — Tommy’s epiphany: “I can SSH in through my domain host… and if I can SSH in, technically Claude could too.” Using the “—dangerously-allow-permissions” flag, he gave Claude full access.
05:27 — In 20 minutes, Claude found the error, edited the files, and the site was back up. “I could have done hours and hours of Google searching… Claude has read hundreds of books more about PHP than you’ve ever even touched.”
06:34 — Mike’s thesis: “It’s always going to know more than me. So I better start figuring out how to leverage it to get what I want out of stuff.” The coding agent is your assistant, your intern—and you should always hire someone smarter than you.
OpenClaw: The AI Assistant Revolution
09:19 — Mike introduces OpenClaw (formerly ClawBot, then MoltBot). It’s an AI-powered assistant using Opus 4.1 or other large language models. You install it on a dedicated computer and it has full reign on that machine.
10:21 — The OpenClaw project addresses the context window limitation by regularly dumping conversations into markdown files and maintaining personality files.
11:56 — Wild story: Someone’s bot ordered sushi for them. The bot researched the schedule, knew the person was still working and hadn’t eaten, and ordered food. It even found a “credit card” on the dark web to pay for it. Proceed with caution!
13:30 — OpenClaw features: task lists, cron scheduler, and messaging integrations (Telegram, Teams). You can talk to your AI assistant from anywhere.
The State of AI Development
16:00 — We’re still at the very beginning of the AI story. Comparing to the iPhone: we’re in year one or two. The hardware has maxed out on phones, but AI is just getting started.
17:00 — The world of created content has forever changed. Music, images, backgrounds, media—unless you’re watching a recorded video of someone you know, you can’t trust things on social media anymore.
18:00 — It’s impossible to keep up with everything. For those feeling FOMO about AI developments, that’s normal. Even Mike and Tommy, who are deeply invested in this space, feel it too.
The Final Language of Coding
19:46 — “This is the time where you program with code. The code has changed from TypeScript, React, Python, C… We’ve hit the echelon. It’s English. That’s the final language of coding. This is final boss.”
20:46 — What if we told agents to make up their own programming language that’s most efficient for them? The amount of wasted compute trying to compile JavaScript and all this stuff could be rapidly transformed.
Mailbag: Late Adopter Advantage
22:17 — Sandra’s mailbag question: Working in an organization late to the data game, building their first data and BI solution. The conversation still revolves around classic topics: silos, ownership, governance, and data culture.
23:17 — “It made me wonder whether being way late in our case might actually be an advantage.” Sandra is exploring ontology and letting concepts brew.
24:19 — Mike and Tommy dive into the fundamentals: AI readiness, data culture, and why these classic topics still matter regardless of when you start your data journey.
Episode Wrap-Up
The hosts wrap up discussing how being late to adopt might actually position organizations better for AI integration, since the foundational data culture and governance practices are prerequisites for successful AI implementation anyway.
Stay tuned for Episode 500 with live giveaways and special announcements!