Tech giants are expanding into new, uncharted territory - something we’ve seen before with cloud computing. For some, like Amazon, Google and Microsoft, building a public cloud was a natural and logical move and they executed fairly well and have captured the majority market share. Others, like IBM, bought their way in, but could not integrate a foreign product and produce a complete and competitive offering. With regard to the cloud, it seems like the builders won.
Currently, Microsoft and Amazon are buying their way in with huge investments in OpenAI and Anthropic. Meanwhile Meta, Google are focusing on internal development, supplemented by acquiring smaller players to produce their own first-party AI platforms.
All the previously mentioned companies are great at writing software, but with AI the hardware component can’t be overstated. Mass market applications running on commodity hardware will serve the general purpose use cases, but the real innovation will come from whomever delivers the tightest integration of proprietary hardware running purpose-written operating systems and applications. Today, Apple does this very well (arguably the best) when it comes to consumer-level general purpose computing. Over time they’ve (mostly) built a platform in the broadest sense of the word: an expansive and integrated hardware/software ecosystem, designed and developed under one roof.
Everyone’s throwing money at AI, but as with the cloud, not all strategies will stick. IBM’s cloud misstep is a cautionary tale—will AI have its own flops?