Rainbow Apparel Co

In April I attended Iterate.ai On, an invite-only AI symposium for enterprise leaders. I walked in thinking I understood the AI landscape reasonably well. Two years of running production deployments at scale. Real experiments. Real stakes. Real outcomes. I walked out knowing something I didn't know when I walked in. Most organizations have no idea what kind of AI they are actually running. Private AI means the organization controls three things: the compute environment, the model itself, and the data that flows through it. That control exists on a spectrum. True private AI runs on dedicated infrastructure with full ownership of models and data. A properly architected private cloud deployment can provide meaningful protection if configuration is verified and contractual terms are airtight. What most organizations actually have is somewhere in the middle. Partial isolation. Protections assumed but not verified. Public AI sits at the other end. Consumer tools like ChatGPT, Claude.ai , Google Gemini, and standard versions of Microsoft Copilot. You control none of those three things. Your input goes to the AI company's servers, governed by their terms of service, which typically permit data collection, storage, and disclosure to governmental authorities. That $20 a month subscription to Claude or ChatGPT your employees are using right now is public AI. Not private. About as confidential as a conversation in a coffee shop. That distinction moved from interesting to urgent in February 2026. A federal judge in the Southern District of New York ruled in United States v. Heppner that documents created using a public AI platform were not protected by attorney-client privilege. A financial services executive typed his attorney's advice into a consumer AI tool. The court ruled the moment he did that, he handed it to a third party. The privilege was gone. He may have waived protection over the underlying attorney-client communications as well. Not just the AI outputs. The original legal advice itself. Enterprise doesn't automatically mean private. Most organizations that believe they have protected AI deployments have not verified what that actually requires. A signed data protection agreement. A confirmed zero training configuration. Infrastructure that keeps your data separated from other tenants. Most Copilot deployments don't meet all three. Most IT teams have assumed rather than verified. A signed contract tells you what recourse you have after something goes wrong. It does not guarantee that nothing will. The distinction between public and private AI needs to be policy, not assumption. #AI #EnterpriseAI #AIGovernance #Leadership

Most organizations ask the wrong first question about AI. Which model should we use? Which vendor should we evaluate? Which pilot should we run? Those are builder questions. They matter. But they are not the first question. The first question is architectural. Before a single line of code gets written, someone has to draw the plan. Where in the organization do decisions actually get made? Where does information slow down before it reaches the people who need it? Where would faster, better information change the outcome? Those three questions, answered honestly, tell you more about where AI can create durable advantage than any vendor demo or model benchmark. In my experience, that diagnostic surfaces two or three places in every organization where AI can fundamentally change how work gets done. Not incrementally. Fundamentally. It also surfaces the places where AI deployment will fail regardless of the technology. Usually because the decision structure is broken underneath. You cannot automate a bad process. You get bad outcomes faster. The second question is organizational. Who owns the outcome? How do you measure it? What changes when the AI is working and what changes when it isn't? Most AI initiatives skip this entirely. The technology gets deployed. The incentives stay the same. The operating model stays the same. And six months later the organization wonders why adoption is low. It is almost never a technology failure. It is almost always a design failure. The organizations that get this right do three things in sequence. Map the decision architecture first. Deploy the technology second. Redesign the incentives and operating model third, in parallel with deployment, not after. That sequence is not intuitive. Most organizations want to start with the technology because the technology is visible and exciting and easy to demonstrate to a board. The architecture work is harder to show in a slide. But it is where the advantage actually gets built. AI is a powerful builder. It still needs a plan. #AI #EnterpriseAI #DecisionSystems #Leadership

Great new workplace for AMMA Verzekeringen - AMMA Assurances !

In 1997 I co-founded a company called PriceSCAN. The premise was simple. Who would buy anything without being able to do instant price comparison? But in 1997 that capability didn't exist. You couldn't comparison shop across retailers in real time. You had to visit stores, make phone calls, clip newspaper ads. We built the infrastructure to change that. Structured product data, real-time price feeds, a consumer decision engine before anyone called it that. We planted a flag before the category had a name. What that flag bought wasn't just a business. It was a seat inside the birth and growth of ecommerce. A front-row view of how a technology wave actually forms. How the infrastructure gets built. How the early architecture compounds. How the window opens and closes. I've watched this sequence repeat across every technology wave since. A capability emerges that changes how people make decisions. Early movers build around it before the consensus forms. The window closes. The early architecture compounds into durable advantage. The organizations that move before the wave peaks don't just get there first. The rules are different early. Access that exists now won't exist later. Some advantages can't be acquired at any price once the window closes. AI is the same wave. Different capability. Same pattern. We are early. Not at the beginning, but early enough that the architecture decisions being made right now will determine who compounds the advantage and who pays to acquire it later. The window is open. It won't stay that way. #AI #EnterpriseAI #DecisionSystems #Leadership

Most UAV failures don’t happen in the air. They start on the ground. After working with VTOL, multirotor, and heavy UAV platforms (up to 90 kg), one pattern is clear: The biggest risks are not “enemy action” or “bad weather”. They are: — poor system integration after assembly — incorrect PID tuning under real payload — unverified communication stability (modems / GPS) — rushed transition from hover to fixed-wing — operators trained on “perfect scenarios only” I’ve seen platforms that looked stable — until real conditions exposed everything. We’ve had forced landings. We’ve had systems behaving unpredictably. But with proper testing and decision-making, outcomes were controlled, and equipment losses remained minimal. What actually makes the difference: ✔ Full validation cycle (not just “test flight”) ✔ Testing under real mission conditions, not lab scenarios ✔ Continuous tuning based on real data (not assumptions) ✔ Training operators for failure, not for ideal flights The gap between “it flies” and “it’s operational” is where most projects fail. Curious how others approach this: What is the #1 failure point you’ve seen in UAV operations?

Noel Garcia is #hiring. Know anyone who might be interested?