ChatGPT the app is not even in the same ballpark as the Chinese ones.
OpenAI treats it almost like personalized social media. It saves everything, so it can pull all sorts of stuff into the context. A bunch of ancillary services are front-and-center. In other words, is sycophantic engagement-maxing. And what its actually doing under the hood is opaque.
The Chinese services I’ve played with are more “utilitarian.” They have some great agent harnesses and tools, but generally its presented as a utility more than a “personal world” like ChatGPT, and what it’s using/doing is crystal clear.
In other words, from a service perspective, I find this claim silly. They’re diverging, greatly.
I’d dispute it for the underlying LLM, too, but on different grounds.
All of that wannabe AGI horseshit is preventing these things from specialization and becoming more efficient and useful IMO.
In my recent usage of LLMs I’ve come to think of them as a type of eye-poppingly expensive, inefficient interpreter.
Losing a lot of the bullshit could make them efficient and inexpensive. To an extent, that’s what Anthropic has already been doing by making models that are mostly code focused.
ChatGPT the app is not even in the same ballpark as the Chinese ones.
OpenAI treats it almost like personalized social media. It saves everything, so it can pull all sorts of stuff into the context. A bunch of ancillary services are front-and-center. In other words, is sycophantic engagement-maxing. And what its actually doing under the hood is opaque.
The Chinese services I’ve played with are more “utilitarian.” They have some great agent harnesses and tools, but generally its presented as a utility more than a “personal world” like ChatGPT, and what it’s using/doing is crystal clear.
In other words, from a service perspective, I find this claim silly. They’re diverging, greatly.
I’d dispute it for the underlying LLM, too, but on different grounds.
All of that wannabe AGI horseshit is preventing these things from specialization and becoming more efficient and useful IMO.
In my recent usage of LLMs I’ve come to think of them as a type of eye-poppingly expensive, inefficient interpreter.
Losing a lot of the bullshit could make them efficient and inexpensive. To an extent, that’s what Anthropic has already been doing by making models that are mostly code focused.
Just because the those models don’t show / use your chat history doesn’t mean they can’t store it. More a warning than a counter argument though.