Brian Bryson – MIT Technology Review

From Pilots to Production: AI Is Now a Business Problem

The theme of MIT Technology Review’s most recent AI conference was deliberately chosen: pilots to production. It reflects exactly where the enterprise AI conversation has moved, from what can AI do to where it belongs and how do we integrate it at scale. Bryson observed a notably wider spectrum of attendees at this year’s event than in previous years, ranging from organizations still in the early stages of understanding AI’s potential to those already running sophisticated deployments across their core business functions. What united them was a shared recognition that the technology itself is no longer the primary challenge. “AI itself is no longer the story,” Bryson said. “It’s integration. It’s pulling AI into the fundamental operating system of the business.”

That shift has a meaningful implication for how senior leaders should be thinking about their AI strategy. The organizations making the most progress are not the ones chasing the latest models or debating which large language model to standardize on. They are the ones focused on workflows, on how AI can accelerate the time from idea to implementation, how it changes the way teams build and deliver, and how it affects customer experience, decision making, and trust. The executives Bryson brings to MIT Technology Review’s stages are increasingly comfortable with that framing. “The more it becomes a business problem,” he said, “the more comfortable executives are with how to manage it, because it is now just a foundational change management problem.”

What Leading Organizations Are Actually Doing

Two examples stood out from MIT Technology Review’s recent conference as illustrations of what effective AI integration looks like at scale. The first is Walmart, the world’s largest private employer. Over the past two to three years, Walmart has trained approximately 1.7 million associates on AI, deploying an agentic system that supports merchants, associates, and internal staff across its operations. The speed and scale of that deployment reflects two core principles that Donna Morris, Walmart’s chief people officer, articulated at the event: at Walmart, AI is people-led and tech-powered, and the definition of value is anything that improves the customer experience. Those two guardrails kept the organization focused on the business problem rather than the technology, and enabled a deployment that most organizations would consider impossible at that scale.

The second example is ServiceNow, which reduced a four-day IT service process to approximately eight seconds using agentic AI, and redeployed the people previously performing that process to higher-value work. The lesson from both examples is the same. The organizations achieving the most meaningful results are not using AI to automate tasks in isolation. They are redesigning workflows from the ground up, starting with a clear definition of the value they are trying to deliver and working backward to determine where AI belongs in that process. “It’s not just slapping a chatbot on your operations,” Bryson said. “It’s really thinking about how you can use the power of AI to accelerate anything and change the way you work.”

The Agentic Shift and the 404 Payment Problem

The next major wave of AI adoption is agentic, and it is arriving faster than most organizations are prepared for. Bryson points to two dimensions of this shift that enterprise leaders need to start thinking about now. The first is the emergence of agent-to-agent commerce. As consumers increasingly deploy personal AI agents to act on their behalf — researching products, making purchases, managing tasks, organizations will need to design their digital experiences for two distinct audiences: human users and the agents acting on their behalf. A presentation at MIT Technology Review’s conference by Peter Smart, chief design experience officer at Fantasy, illustrated what this looks like in practice: a retail website redesigned specifically to communicate with a purchasing agent, guiding it efficiently to the right product without the visual and navigational elements designed for human browsing. The implication for enterprise leaders is significant. Public-facing digital infrastructure will increasingly need to support both a human interface and an agent interface simultaneously.

The second dimension is what Bryson calls the 404 payment problem,  a reference to a revenue model built into the original HTTP specification that was never widely implemented. As AI agents absorb and interact with content across the internet at exponentially greater volume than human users ever did, the current model of free content access becomes economically unsustainable for content creators and publishers. A microtransaction-based compensation model, where agents submit payment as part of each data request, is a logical solution, and Bryson expects it to emerge within the next few years. “The volume of transactions with agents is now exponentially greater than the volume of transactions with humans,” he said. “There is going to need to be a compensation model that works.”

The Convergence of AI, Quantum, and Energy

The most consequential technology story of the next decade is not AI in isolation. It is the convergence of AI, quantum computing, and energy, and the intersection points where those three forces amplify each other. Bryson and the MIT Technology Review team have organized their upcoming MTech Future conference around exactly that thesis, examining how breakthroughs in one domain are accelerating progress in the others. AI is being used in materials science to develop new battery technologies and improve energy efficiency. Quantum computing is accelerating the modeling capabilities that underpin both AI research and energy system design. And the energy demands of AI infrastructure are driving investment in new power generation and distribution technologies at a scale that is reshaping the global infrastructure buildout.

For enterprise leaders and investors, Bryson’s view is that the money will be made at the convergence points, the places where two or three of these forces intersect and create possibilities that none of them could produce independently. “Understanding the interconnectedness and convergence of technologies, and how one thing is going to shape another, that is what dictates the speed of flow,” he said. “The ability to see those intersection points, those convergence points where things are moving, that is where the value is going to be.” Speed for speed’s sake, in his view, is a trap. The organizations that win will be the ones that slow down enough to understand the convergence before moving fast to capitalize on it.

MIT’s Strategic Repositioning and the Personal Reboot

MIT is not standing still while these forces reshape the world. Under president Sally Kornbluth, the institute has identified six strategic priorities, AI, quantum, energy, climate, human health, and a cross-disciplinary focus on how these domains interact. This fall, MIT will open its doors to the world through an institute-wide event called Future Fest, designed to promote the value of the scientific approach and the importance of focusing on the core technologies that will move the world forward. MIT Technology Review’s upcoming conference is directly tied to that initiative, with AI, quantum, and energy as its central focus.

For Bryson personally, the past year brought a reboot of a different kind. Nearly a year ago, he underwent spinal fusion surgery — a procedure that removed a disc between two vertebrae and left him unable to take a step the morning after the operation, despite having entered the surgery in the best physical condition of his life. The recovery required rebuilding his entire physical foundation — core strength, hip flexors, muscle memory, from scratch, relearning movements his body had compensated for years. “The solution was to be slow, to be resilient, to accept that things had changed, and to build new strengths,” he said. The parallel to the business advice he gives every day was not lost on him. Whether rebuilding a body or rebuilding an organization, the principle is the same: speed for its own sake leads to failure. Understanding the new foundation, adapting to the new reality, and building deliberately from the ground up is what produces lasting results.