Until recently, Gemini was capable of opening and processing links. I personally witnessed it successfully extracting information from websites numerous times. However, it appears that all of its models now return an error when presented with a link. This is a regression from its previous functionality.
1 Like
these comanies are using us as free testing, and free codeBases inputs, they restrict all featrues that have any sort of a value, we are feeding techno monsters that are stilling our intelectual properties!
You are losing zero intellectual property rights if you consistently apply the simple approach: In AI Studio prompts, only use publicly available information to design prompts; only use API in pay-as-you-go mode when sending data to the model that you do not wish Google will use in future model training.
On free testing, I also initially had the thought that developers were getting side-hustled to also be free QA for Google. But that is incompatible with the benign neglect / equal opportunity disregard for any feedback developers have; no organization would so thoroughly ignore their QA department if in fact developers are QA on the side.
1 Like
people are puting whole code bases into GEMINI, they are feeding the monoster, and they are getting lobotomised left wooken useless manipulator
The real answer will be open source ai, gemini and chatgpt are a way to learn how to use it, but unless and until you self host, you got nothing. That is not unique to ai … that is true of pretty much everything.