In 2016, technology commentator Kevin Kelly published 'The Inevitable,' in which he outlined several expectations for artificial intelligence by the year 2026. Among these, he predicted that 'By 2026, Google’s main product will not be search but AI,' a forecast he traced to a conversation with Google cofounder Larry Page, who told Kelly that even in its early search‑focused years the company was 'really making an AI.' Kelly argued that Google’s vast stream of search queries functions as continuous training data for its machine‑learning systems, so each user interaction improves the underlying AI rather than merely refining search results.
Kelly also foresaw structural changes in the AI industry, anticipating an 'oligarchy of two or three large, general-purpose cloud‑based commercial intelligences' would dominate the market. He linked this expectation to the network effects inherent to cloud computing: as more users engage with a cloud‑AI service, the system becomes smarter, which attracts even more users, creating a self‑reinforcing cycle that tends to marginalize smaller competitors. Complementarily, he described the forthcoming AI as resembling Amazon Web Services—cheap, reliable, industrial‑grade digital smartness that runs behind the scenes and is accessed like a utility.
Regarding the form and accessibility of AI, Kelly contended that 'the bulk of AI will be special purpose software brains' designed for narrow, high‑performance tasks such as language translation or image recognition, rather than exhibiting broad, human‑like capabilities. He further imagined a future where individuals could pose questions 'in conversational tones, any question at all' and receive explanations on topics ranging from sports trivia to cosmological principles, with the answering system learning to articulate what is known and what remains uncertain.