Most users aren’t going to manage API keys, know that that even means, or accept the friction.
The problem with the current model is that there is a high barrier to justifying the user pays essentially a 2nd/3rd subscription for ultimately the same AI intelligence layer. And so you cannot currently make an economically successful small use case app based on AI without somehow restricting users use of AI. I don't think AI companies are incentivized to fix this.
A capital intensive, low margin business. The dream of every company.
I think, the reality is, as models become more competitive they are becoming commodities. There's really no reason an app has to be built on GPT, or Gemini. It makes much more sense for apps to be "model agnostic" and let their customers choose which models to use.
I think, if OpenAI sticks to just trying to make their own apps for everything, they will be outrun. People will make apps outside of their ecosystem and will just use them as an API dumb pipe, regardless of if OpenAI wants that. And if they don't want that and restrict it, then their models will fall to the wayside as more competitive models which DO allow that take their place.
They're in a bind here, which is probably why we are seeing this announcement. OpenAI can see the writing on the wall for them.
The perfect "killer app" for AI would kill most software products and SaaS as we know them. The code doing the useful part would still be there, but stripped off branding, customer funnels and other traps, upsell channels, etc. As a user, I'd be more than happy to see it (at least as long as the AI frontend part was well-developed for power users); obviously, product owners hate this.
Due to technical and social limitations, most apps are also limited in what they can do, this naturally shapes and bounds them and their UIs, forming user-facing software products.
Intelligence of the kind supplied by SOTA LLMs, is able to both subsume the UI, by taking much broader context of the user and the use case into account, distilling it down to minimal interaction patterns for a specific user and situation, and also blur the boundaries of products, by connecting and chaining them on the fly. This kills the marketing channel (UI) and trims the organizational structure itself (product), by turning a large SaaS into a bunch of API endpoints for AI runtime to call.
Of course, this is the ideal. I doubt it'll materialize, or if it does, that it'll survive for long, because there's half a software industry's worth of middlemen under risk of being cut out, and thus with a reason to fight it.
I assume the fall off there will be 99% of users though, the way it works today.
But this theoretically allows multiple applications to plugin into ChatGPT/claude/gemini and work together.
If someone adds zillow and… vanguard, your LLM can call both through mcp and help you plan a home buy
Maybe a 'connect with OpenAI' button so the service can charge a fee, while allowing a bring your own token type hybrid.