My take: over success breeds complacency. Apple knows where its money is coming from. It has carved out an extremely hard to establish hardware production chain via iphone and macbooks, and is able to provide a certain consistent level of quality for it. Software is an afterthought, especially for software that is not in service of this primary hardware revenue source. From a business point of view it literally does not make sense for them to do anything differently.
Of course, disruption is always a possibility. Google was the undisputed AI leader for years, but their reputation as a House of Knowledge was overshadowed by their comfort as a search, cloud, and advertisement business. These are steady services that just need to be reliable to remain profitable, no invention required.
For a while I was surprised by Mircrosoft's signs of life around generative AI by the time OpenAI came about, but it seemed to relapse into complacency too.
I honestly believe there is some unstated law of success, I think there is a "ceiling" to success, at which point it becomes impossible to expand. It has something to do with the correlation between success in complexity. As a business grows more successful, it becomes more tied down to various commitments, constraining its ability to innovate without assumptions. There's a limit to what any given entity can handle.
I like this take.
To add to that, it feels that for a company to maintain their software effectively, there needs to be a certain level of cross-departement knowledge, people who can connect the dots between frontend and backend for example, because usually it is in these transitions between two layers that bugs start to form. I feel like this becomes increasingly more difficult in large, complex company's where every departement is self-contained and there is not much vertical movement amongst colleagues, only horizontal. Which makes it really hard to solve issues that are not solely linked to one part of the process. So success doesn't only breed complacency, it decreases the possibility to do cross-departmental work because all the departments become to big to effectively facilitate this.
For a while I was surprised by Mircrosoft's signs of life around generative AI by the time OpenAI came about, but it seemed to relapse into complacency too.
I honestly believe there is some unstated law of success, I think there is a "ceiling" to success, at which point it becomes impossible to expand. It has something to do with the correlation between success in complexity. As a business grows more successful, it becomes more tied down to various commitments, constraining its ability to innovate without assumptions. There's a limit to what any given entity can handle.