Avoid an asymmetry assumption in customer AI strategy

I'm one of a small band of analysts at Gartner who survey and research CEOs. My primary interest is in how macro trends are shaping their thinking and the implications of that for technology related business change agendas. This year's big issue is of course generative AI. I'm reading many CEO interviews and responses to investor day questions about their views on the matter. It's early days of course. Most CEOs are a little hazy in their responses. They say that there are potential benefits but also risks. Some point to internal efficiency possibilities. However nearly all of them comment about how AI might be used by their company to advance the product or the customer experience  - and that's where I see a troubling asymmetry assumption. Business leaders seem to be assuming that AI under their control will be at the center of customer interaction advances. But what about AI in the hands of the customers themselves? Or AI in the hands of new intermediaries acting more in the interests of the customers than the producers?  ChatGPT, Bard and the rest will place enormous 'end user' power in the the hands of customers. Generative AI will massively improve the ease of learning to code and of generating code. Open source AI projects will make very powerful and increasingly specialized large language models available via APIs. This all means that small startups and even individual developers can readily create pretty powerful 'custobots'. That's our down-to-earth term for a machine customer - a non-human economic actor that obtains goods or services in exchange for payment. These custobots will do shopping, selecting, negotiating and buying on behalf of their human customer owners... and eventually for themselves. In specific niches, custobots have already become a significant market phenomenon, just using more conventional algorithmic designs. For example Ticketbots help buyers secure seats at concert venues and sports events. Sneakerbots help customers get access to small batches of collectible footwear that can be hard to pounce on. Once we add the power of generative AI, the potential of these software bots to morph and multiply is enormous because they can be evolved to interact in complex, natural language dialogs with sellers.. of almost anything. There is a good chance they will be faster, more logical and test more product criteria when they are shopping. Put simply - they could be harder customers to predict and to please, than the humans on whose behalf they are working. So be wary of slipping into an unconscious asymmetry assumption. Do not presume that the value adding application of AI will take place primarily on the side of the producer or vendor. Customers - B2B and B2C - will harness the technology to work on their end of things too. Machine Customers will be a powerful force - in every market in the long term. That's why we wrote a book about it. Amazon link to 'When Machines Become Customers' -  https://a.co/d/3cdMV4S