For the past three years, the biggest names in tech have been in an arms race. OpenAI vs Google vs Anthropic vs Meta vs Mistral. Who has the smartest model? Who scores highest on benchmarks? Whose AI can pass the bar exam? That race isn’t over exactly. But it has become irrelevant. At IBM’s annual Think conference just this week in Boston, one of the most important declarations in tech was made quietly but clearly: the AI model is no longer the main differentiator. What matters now is what you build around it.
What IBM said at Think 2026.
Gabe Goodhart, Chief Architect of AI Open Innovation at IBM, put it directly:
“We’re going to hit a bit of a commodity point. It’s a buyer’s market. You can pick the model that fits your use case just right and be off to the races. The model itself is not going to be the main differentiator.” Gabe Goodhart, IBM Think 2026.
IBM CEO Arvind Krishna went further, declaring that the enterprises pulling ahead in 2026 are “not deploying more AI they’re redesigning how their business operates.”
This is a fundamental shift. And it has real implications for every business owner trying to figure out where AI fits into their future.
What does ‘models are commodities’ actually mean?
Think of it like electricity. In the early days, having electricity was a competitive advantage. Today, every business has electricity. The advantage doesn’t come from having power it comes from what you build and run with that power.
AI models are becoming the same. ChatGPT, Gemini, Claude, Llama they are all remarkably capable. They are all accessible. They are all, increasingly, affordable. Having access to an AI model is no longer special. What you do with it is everything.
And what you do with it according to every major tech company paying attention right now comes down to three things: orchestration, agents, and workflows.
What does this mean for your business right now?
Here is the honest, practical truth for business owners in Kenya and Africa:
You don’t need to be IBM to benefit from this shift. In fact, the shift makes AI more accessible for smaller businesses than ever before. Because if the model is no longer the hard part and you can access world-class AI through affordable APIs then the competitive advantage lies in how smartly you connect it to your business.
What that looks like in practice:
-
Your website that captures leads and automatically sends them into a follow-up workflow.
-
Your customer support system that resolves common queries and escalates complex ones.
-
Your operations dashboard that monitors data and alerts the right person at the right time.
-
Your business processes that run end-to-end without manual intervention.
None of these require you to build an AI model. They require you to have the right digital systems and the right software partner who understands how to connect them.
The AI divide is widening. Which side are you on?
IBM’s theme at Think 2026 wasn’t just about technology. It was about a widening gap. The businesses that are redesigning how they operate around AI systems are pulling dramatically ahead of those still experimenting with a single chatbot.
The model wars kept many business owners on the sidelines, waiting to see who would win. That wait is over. The winners have been declared. Now the real game begins and it’s being played at the level of systems, workflows, and integration.
The businesses that build those systems now even simply, even incrementally will be the ones still standing in five years.
The three things that actually matter now.
Orchestration
Orchestration means connecting multiple AI models, tools, and data sources so they work together as a unified system. One AI handles your customer queries. Another analyses your sales data. Another monitors your operations. Orchestration is the layer that coordinates all of them so they don’t just work individually, they work together intelligently. IBM launched its next-generation watsonx Orchestrate platform at Think 2026 specifically for this purpose.
Agents
We wrote about Agentic AI recently. But now it goes further organisations are no longer deploying one or two agents. They are managing hundreds, sometimes thousands, built by different teams on different platforms. The challenge has shifted from building agents to governing, connecting and scaling them. Who is accountable when an agent makes a wrong decision? How do you ensure consistency? That’s the new frontier.
Workflows
A workflow is a sequence of actions designed to complete a business process. In 2026, AI-powered workflows don’t just automate tasks they reason about what step comes next, adapt when something unexpected happens, and loop in humans only when genuinely necessary. Kevin Chung, Chief Strategy Officer at Writer, described it at IBM Think: ‘AI is shifting from individual usage to team and workflow orchestration coordinating entire workflows, connecting data across departments.