OpenAI Computer-Using Agent
openai.comI genuinely wonder if this kind of use case is going to completely upend UX testing. There's something dynamic, perhaps poorly so, that would change the way prescriptive, painfully locked UX tests work...
Depends on the pricing & speed. I think you'd use this to create scripts that would run frequently, then on failure you'd have this evaluate if there were changes and if so create a new script.
It would depend on the reliability of the model to produce deterministic results. So far that’s been pretty tough.
Well, looks like they doubled down on the whole "computer use/ing" language, which is a total misnomer.
Now pardon me while I body use to get some coffee.
If I order their $200 a month plan can I use this today?
The answer is yes, but I misunderstood when reading this and thought it actually could use my computer when it just runs its own browser.
ByteDance dropped an open source computer using AI today as well.
I thought this was funny from the UI-TARS ReadME:
``` ## Note - Use Chinese in `Thought` part. - Summarize your next action (with its target element) in one sentence in `Thought` part.
## User Instruction """
```
Can it play a game?
How about a nice game of chess?