I've been thinking the same since GPT3 too, and since ChatGPT, and since Claude and... But here I am, still paying for ChatGPT Pro because it's literally has the best model you can get access to for a fixed price each month, and none of the others so far come close. I still use Anthropic's and Google's models to compare/validate against, because I assumed at one point they'd surpass OpenAI, but so far they haven't. This all makes me believe less and less each day that it'll actually be commoditized.
I think OpenAI having the best model still isn't enough. The AI marketplace isn't really in a race to the top, it's in a race to the mass market middle. If Gemini is good enough for the majority of people to complete the majority of tasks they want to then market effects and bundling can get an already dominant company like Google to take over the market. And that's without considering the integration possibilities, e.g. Gmail and Google Docs.
That doesn't mean everyone will use Gemini. As a software engineer I prefer Claude Code and will pay good money for it. I'm sure there will be plenty of other specialisms that will have preferred models. But OpenAI's valuations are based on the idea that it's going to be everywhere, for everything, all the time. And I'm skeptical. ChatGPT Pro is a $200 a month product. That's not a mass market proposition.
> good enough for the majority of people to complete the majority of tasks
It will never be this. There is always the expectation of being able to do more things.
"Log into my work email and deal with all of them whilst I have a bath".
"Start a company for me to earn some extra weekend cash by washing peoples driveways. Find and hire some people to do the actual washing"
"Find a nice house for me by a lake, negotiate a good price and buy it (get a mortgage if necessary) then book all the removals services and find me a new job nearby".
If 3 or 4 competitors can all provide a mostly identical product, isn't that a commodity? That is essentially the case right now, with the different companies playing around with UI, integrations and business model.
If all models were equal then sure. But for professionals who use these to solve complex problems and need correctness above all? The models and weights are not equal and interchangeable.
I use them every day for coding and Gemini 3 pro, Opus 4.5, and GPT 5.1 (haven’t tried 5.2 yet) are basically identical in terms of ability. Opus 4.5 has a slight edge in my personal experience so far.
That's interesting. I think besides the hard facts (like numnber of possible tokens before throtteling happens) the perceived quality differs from use case to use case. In that sense, ChatGPT is my daily driver, in terms of helping with coding problems or debugging Claude feels far superior to me. And when it comes to ideation and creativity (product idea validation, etc.) Gemini surpasses both of the other in my opinion.
But what could possibly be the secret sauce? Whatever it is, eventually enough engineers will move between orgs to get that stuff cross-pollinated.
Certainly there’s little to suggest that it has much to do with Altman’s leadership or a culture of engineering excellence/care that has been specifically fostered at OpenAI in a way that isn’t present at Facebook or especially at Google.
I've been thinking the same since GPT3 too, and since ChatGPT, and since Claude and... But here I am, still paying for ChatGPT Pro because it's literally has the best model you can get access to for a fixed price each month, and none of the others so far come close. I still use Anthropic's and Google's models to compare/validate against, because I assumed at one point they'd surpass OpenAI, but so far they haven't. This all makes me believe less and less each day that it'll actually be commoditized.