'AI Can’t Fix a Business That Doesn’t Understand Itself' - Akili AI CEO

The artificial-intelligence boom has arrived in boardroom discussions with an awkward translation problem. Much of the public conversation about AI revolves around chatbots like Claude, ChatGPT, and Gemini. However, Akili AI founder and CEO Simon Bransfield-Garth believes AI is like a jet engine. A powerful propulsion tool that is unhelpful if it is not integrated in a befitting manner with other components of the jet. 

Inside businesses, the more consequential uses of AI can be considerably less glamorous. It can be used in processing a loan application, interpreting an insurance contract, or making sure a customer does not wait several days for an answer. And that gap between the AI people hear about and the AI organisations can actually use is becoming one of the central obstacles to adoption.

“The popular narrative on AI is driven by leading AI companies in California.  They're going to talk about using AI for sequencing new genomes, everybody talks about the frontier models. But the real world is not like that. The real world is full of tedious, boring problems where you want AI to eliminate the drudgery of those tedious, boring problems,” Bransfield-Garth says in an interview with The Kenyan Wall Street. 

There is a limit to what AI can sensibly fix and a poorly understood process cannot simply be automated into competence.

Akili AI, an AI solutions provider, stepped in to ensure that businesses integrate the technology into their existing operations. The company's experience suggests that the difficult part of corporate AI is often not acquiring the technology. It is figuring out what the technology is supposed to do.

The company's approach begins with the business itself rather than the machine. Instead of asking where AI can be inserted, the firm starts by identifying a problem, the organisation's promise to its customers, and the point at which the business is failing to keep that promise. AI is then considered as a means of closing the gap.

That distinction matters because businesses do not generally suffer from a shortage of software or other tech deficiencies. They often suffer from slow processes, fragmented information, and work that is tedious enough to consume employees' time but important enough that it cannot be simply ignored.