Many people think of AI as the train. In reality, AI is the railway. The real shift is the infrastructure being built today: compute, energy, talent, and digital stack. Countries that invest in this infrastructure will shape the future of AI. My conversation with @BCG at MWC.

AI-powered wildlife alerts on rural roads are a useful application of connected technology, especially with measurable speed reductions during testing. @TMobileBusiness Partner

Two patients have the same diagnosis. Would the same treatment be right for both? Clinical decision support helps clinicians examine the differences. Patient information supplies the context. Medical knowledge helps interpret it before a decision is made. Microblog @antgrasso Consider a clinician preparing to prescribe a medication. The patient record contains kidney function results that may affect the appropriate dose. A clinical decision support system can combine that information with medical references and flag a possible dosing concern for review. Now the clinician has something to examine. Are the patient data current? Does the recommendation make sense in this particular care situation? The software can bring relevant evidence to the attention of the clinician while the decision is being considered. The clinician interprets that evidence and decides how to proceed. The purpose is to make medical knowledge available in the context of the patient, without transferring clinical judgment to the software. #ClinicalDecisionSupport #DigitalHealth

Data privacy depends partly on where AI runs. Gradium's Phonon 1.0 generates speech on the user's device, keeping input text local and avoiding per-request cloud fees. You can hear the result in this French-speaking game demo.

Cybersecurity is a business decision. Connect AI governance, identity protection, and cryptographic readiness to clear ownership and measurable risk reduction. Make resilience a leadership priority. Source @Gartner_inc link gtnr.it/4jjawSk via @antgrasso

The next tech advantage won’t come from AI alone. It will come from the ability to turn it into the physical world—through energy, chips, infrastructure, talent, and operating models built to make innovation work at scale. @McKinsey Link buff.ly/WBTnQUb via @antgrasso

AI adoption does not automatically build trust. With 30% of Americans worried about its pace of development, leaders need to make AI's value tangible: show measurable results, explain its limits, and make human accountability clear. Source @StatistaCharts via @antgrasso

A KPMG survey finds that 89% report measurable value from AI, but only 12% consistently assess that value against its cost. The figures raise a question: what can organizations conclude about AI economics if costs and benefits are not compared consistently? Microblog @antgrasso In its Global AI Pulse Q3 2026, KPMG reports that 61% review AI costs during approval and 59% monitor them once systems are running. Consider an AI application that processes supplier invoices. The finance team may know how much it costs each month. But how much time does it save once the manual corrections are taken into account? Without that comparison, deciding whether to expand the application is harder. Among organizations reporting established ROI, KPMG finds that 48% consistently compare AI value with cost. Financial control tells an organization what it spends. Connecting cost to results helps it decide whether further investment makes sense. #AIEconomics #AIManagement

If you use AI to avoid thinking, you are going in the wrong direction. Use it to become more capable without giving up your judgment. AI can expand what you can examine and imagine. Let that additional capacity make you more curious, not more passive. Microblog by @antgrasso Saving time on routine tasks is useful, but accepting an AI output as a decision already made is another matter. Consider a team redesigning a customer service process. AI can examine patterns across many requests and suggest alternatives the team might not have considered. Good. Now you have more to evaluate. One proposal may reduce waiting times while making it harder for some customers to speak with a person. Would you accept that compromise? Ask AI to help examine the consequences, too. Then decide what better service means and which alternative to pursue. The responsibility for the outcome remains with people. #AIAugmentation #HumanCenteredAI

If enterprise architecture makes you think of buildings, think again because it maps how business and technology fit together. EA shows the connections across the organization, so decisions are made with the whole system in mind, not one piece at a time. Microblog @antgrasso Digital transformation can improve how an organization works, but it also makes business and technology more dependent on each other. Enterprise architecture makes those dependencies visible before changes are made. Suppose a company wants to introduce same-day delivery. The business process may need to change. The order application may need new functions. Inventory data has to be available at the right moment, and the infrastructure must support the workload. That one business decision now touches process, application, data and infrastructure. Change one part and another may be affected. Replacing an application or moving a workload to the cloud can have the same effect. The technology decision sits inside a business system. Enterprise architecture gives you the map of that system, so you can see what depends on what before you change it. #EnterpriseArchitecture #DigitalTransformation

People often say a qubit is 0 and 1 at the same time. That is simple, but incomplete: think of it as many possible shades between black and white before measurement gives one result. The difference lies in what can happen before you read the final result. Microblog @antgrasso A classical bit is easier to picture. It is like a bucket that can contain white paint or black paint: 0 or 1. With a qubit, the state before measurement can take many possible forms. Physicists call this superposition. The shades are only a metaphor. A quantum computer does not read an intermediate gray value. The state is described by numbers called amplitudes, associated with the possible outcomes 0 and 1. While the computer is working, quantum operations change those amplitudes. Think of waves. Some reinforce each other while others cancel. Amplitudes can combine in a similar way. This is called interference. Then comes measurement. You do not read a shade. You get a classical result: 0 or 1. So the interesting part is not that a qubit somehow gives you two answers at once. It is what the quantum computer can do with those possibilities before measurement gives you one answer. #QuantumComputing #Qubits

Digital leadership is no longer about access alone. The advantage comes from turning connectivity into business capability—using digital infrastructure to make faster decisions, improve customer value, and scale with less friction. Source @VisualCap via @antgrasso

Humanoid robots may soon be everywhere around us, so rebuilding their AI from scratch for every new task would be a poor use of resources. The more tasks we expect them to handle, the less sense it makes to rebuild the intelligence behind capabilities. Microblog by @antgrasso A robot foundation model changes the starting point. Instead of developing a separate model for every capability, developers begin with a pretrained model that has already learned patterns that can transfer across robotic tasks. What can be reused depends on the model. It may already contain representations that help the robot understand a scene or decide how an action should proceed. Take a warehouse humanoid. The base model may already help it recognize boxes and understand an instruction such as moving an item from one location to another. Adapting the robot to another warehouse may then require less task-specific training than building that capability from zero. Now move the same idea to a factory. The robot may face different parts or movement constraints. The foundation model provides the base, while additional training and configuration adapt the behavior to that environment. The same principle can support new tasks on the same robot. A system adapted to move containers may later be trained to inspect components without replacing the underlying model. But reuse does not mean plug-and-play. A different robot may have different sensors or physical limits. Safety requirements can also change with the task. The model still needs adaptation and testing before the robot can operate reliably. Every new capability does not have to begin from an empty page. A robot foundation model lets developers reuse what has already been learned and spend their effort on what is different about the next task. #Robotics #FoundationModels

Security complexity becomes a governance risk when a company loses track of who can access its systems and why. @TMobileBusiness Partner

When discussing new energy partnerships, I would keep one question in mind: what will they make possible for the people and businesses that depend on reliable energy every day?

Global energy projects move forward when technical ideas find investment and people build partnerships to deliver them. ADIPEC 2026 will bring the energy industry together in Abu Dhabi to make those connections possible. Join > adipec.com @ADIPECOfficial Partner. Capital can be committed during a meeting, but delivery depends on the work that follows. Technology companies and energy operators need to agree on what can be delivered under real conditions, while investors need confidence that those plans are feasible. ADIPEC 2026 provides a global meeting point for these conversations, connecting the people who can turn opportunities into projects. I would pay particular attention to partnerships that continue after the event, when commitments have to become results. Energy. Intelligence. Impact. #ADIPEC2026 #ADIPEC #Energy #Partnerships