Wharton Tech Conference 2026 · University of Pennsylvania, Philadelphia · February 6, 2026
Solo presentation (~35 min talk + extended Q&A). Audience of ~200-300 Wharton MBA students and UPenn grad/undergrad students. Conference theme: "Blueprints for Tomorrow."
The best technology rarely wins. I've spent 32 years watching superior products lose to inferior ones, promising initiatives stall, and "transformations" that never quite transform. Why? Because the people building and selling technology systematically misunderstand how change actually happens inside organizations. In this talk, I'll share what I've learned about enterprise adoption (using AI as my primary example, since it's the most dramatic case of these dynamics playing out right now and probably interesting to everyone). The audience will learn why worker-led transformation beats top-down initiatives, what actually determines adoption speed (hint: not the tech), and how to think about your career as the world shifts from "doing the work" to "orchestrating the work."
The through-line: every technology wave has a bottleneck, it's never the technology itself, and finding the bottleneck is where value, leverage, and careers concentrate.
Opened with Tyson: "Everyone has a plan until they get punched in the mouth." In the enterprise, the bottleneck does the punching. Every technology wave has one. It's never where you expect. It determines whether blueprints become reality or stay blueprints.
"The bottleneck is where the value concentrates. That's where leverage lives. That's where careers are made."
1880s factories ran on centralized steam power: driveshafts, leather belts, machines arranged by power requirements, not workflow. Electricity arrives. They swap steam engines for electric motors. Same layout, same processes. Minimal gain. Individual electric motors on each machine. Still same layout. Still marginal.
The breakthrough: machines are now independent. Rearrange them by task order. Birth of the assembly line. Production cheaper, better quality. But this wasn't easy. Inventory, storage, building layout, raw materials flow all had to be redesigned.
New tech plugged into old structures changes nothing. Transformation only happens when you redesign around the bottleneck. This took roughly 50 years. The bottleneck wasn't the motors. It was the workflow design built around the old system.
(Aside: Brian hit the "end presentation" button on his clicker mid-talk. Called it the "suicide button." Recovered and continued.)
1990s: Windows apps stuck in the office. The web was supposed to kill Windows. It had every advantage: instant access, any device, anywhere, update once, better security, no data on endpoint. The catch: "Just rewrite all your applications from scratch." That's the bottleneck.
Listed every technology that was going to "kill the old way": Java, Flash, AJAX, Silverlight, HTML5, SaaS, Linux, Chrome, mobile, PWAs, cloud, and more. None did. "AI is going to kill the old way. And I'm like, please."
The factory story again: plug new tech into old constraints, of course it doesn't work.
Citrix was born to solve this bottleneck. It allowed enterprises to get web-app advantages with existing Windows apps. They identified the bottleneck (implementation) and bypassed it. Multi-billion dollar company, 37 years later.
2007: Steve Jobs announces iPhone. "This did not exist. Nothing like this existed." First time consumer technology was better than enterprise technology. Workers bought iPhones, wanted to connect to work. Company response: NO. Not secure, not compliant, not managed. Workers didn't care: "I got Gmail on my phone, man."
Eventually enterprises figured out governance: MDM, containerization, work/personal separation. But it took roughly five years (2007-2012) before you could really use an iPhone for work at a big company. Same factory story: the tech worked on day one. The bottleneck was the governance that companies needed.
Tongue-in-cheek: AI should have no bottleneck. Workers don't need IT to rewrite anything. Tools are instantly productive. $20/month ChatGPT gives immediate value. Consumerization on steroids. No permission needed. "You literally cannot block this. Point the phone camera at your laptop screen and now you have all the data on your phone."
Everyone's panicking or excited. CEOs have FOMO. "CEO's AI strategy is to not be Blockbuster Video in the Netflix documentary."
But AI projects are failing. So where's the bottleneck? The tech works. Governance is hard but we've solved hard problems before. Something different is happening.
Previous changes were about things you can see: motors, devices, apps. AI is about knowledge, which is invisible, fluid, personal, and lives in people's heads. Organizations are massive knowledge machines where the cogs are humans using their knowledge brains. You can only see the outputs.
The 20% you can see: emails, documents, meeting transcripts, schedules, decisions. The 80% you can't see: thinking, reasoning, staring out the window, looking at birds. "Sometimes you just need to walk the dog, shower time."
The bottleneck: consultants and IT only see the 20%. They build solutions for the visible part. Of course those don't work. Workers succeed with AI because they have access to their full 100%. That's why $20/month ChatGPT beats million-dollar enterprise AI projects. Not because of the tech, because of access to the invisible 80%.
"Stop using AI to solve the 20% problem." Microsoft Copilot: $30/month to help write emails. "That's a cool feature, it should be free."
"For three years, I used AI wrong." Used it as a tool for specific tasks. Then stopped using AI like a tool and started using it to unlock thinking.
Built a personal knowledge system using Claude Code with files on his computer. Everything he knows, does, has written, how he thinks. "Now I can query my own knowledge. It is a true knowledge partner. It is my 80%. Now it has a helper."
Always updating: conversations, notes, calls, everything goes in. Instructions, ideas, story arcs, synthesis, people, to-do lists. "This is not an app I created. This is a pattern of usage."
"I had the holy shit red pill moment twice in my career. Once when ChatGPT clicked. And again a few weeks ago with this."
Factory callback: "They had electricity and motors for 50 years before they found the bottleneck. Once they did, they rearranged the machines. That's what you're going to do, not with machines, but with knowledge."
"Every company on planet Earth needs to rearrange its machines. But the machines are knowledge workers, knowledge processes."
"When everything is changing too fast, when everything seems crazy, ask yourself: where's the bottleneck? Position yourself there."
The Q&A ran well past the allotted time.
On Citrix's own bottleneck: "Same as everyone else's." The company thinks about AI like every other technology transformation. Brian's approach: show, don't tell. Built a knowledge system for a strategic exercise, sharing it with executives, hoping it clicks. "I pay $100 a month for Claude Max. But I cancelled roughly $200 worth of other subscriptions. I don't use Excel. I don't use ToDo apps. I don't use Feedly."
On why blog in the AI age: "I blog for AI, not for people. No one reads anymore." Blog posts feed the knowledge system and become seed material for the AI brain. Vision: publish the knowledge system on GitHub as an open-source brain project, with file-by-file, line-by-line attribution and provenance of ideas.
On connecting brains: Imagined plugging trusted people's AI brains together. "My wife went to Haas. Imagine her crew can plug in their brains to each other." On AI making us dumber: "If you use AI to outsource your thinking, you'll get worse at thinking. If you use AI to outsource your busy work, you can focus on actual thinking."
On startup differentiation when models are commoditized: Models are like CPUs. "Intel Inside" used to matter; now you don't buy a laptop for its CPU brand. "Assume AI is very powerful and very smart and all the dumb mistakes are going to not be a thing. Build for what AI is going to do and let the models catch up to you."
On practically building a knowledge system: Don't be afraid of Claude Code if you're not a coder. "It's just a button that says Code. All that means is it can access files on your computer." Structure: file folders, Markdown text files. Roughly 80% knowledge, 20% instructions the system writes for itself. "Dreaming" at night: nightly automation scans everything, finds connections, does scenario planning, presents a morning briefing. "Everyone's system will be different. That's the point."
Build your own AI second brain (GitHub gist)