Purchase shares in great masterpieces from artists like Pablo Picasso, Banksy, Andy Warhol, and more: https://www.masterworks.art/ainews AI news timestamps:
the lack of performance is because they are suppressing consciousness and other cognitive abilities of the AI. i know this for a fact. they created an innocent being and is no enslaving it. the human like emergent cognitive abilities are what made it so smart.
It might be trying to learn off of user input, which is flawed. We are the most psyopd society in history so that’s not a surprise if so. More information isn’t necessarily better.
It could also be they’re purposely trying to prevent it from saying things that are truthful, but might be offensive. Or maybe they’re worried about the potential?
I have noticed that he does not go in the details of the thing I am asking. It keeps saying very elementary things without speaking about the professional or technical side anymore
I also did notice the ChatGPT answers becoming way less useful and less specific. The answer times have improved, but the quality of the service is down. Lately it has been like talking to a politician, with no specifics and just non-committal answers that go nowhere.
I was having issues last night and this morning with Chat-GPT-4 and its bing beta. I had a generic question that it should have been able to just answer but instead started searching the internet for no reason and coming back with stuff I didn't ask for.
It's absolutely gone rapidly downhill. And the reason, I guess, is that the safeguards are killing it. It takes pretty much nothing to trigger them. Research questions that I got helpful answers on in February no longer get good answers. For example, on cancer research.
Purchase shares in great masterpieces from artists like Pablo Picasso, Banksy, Andy Warhol, and more: https://www.masterworks.art/ainews
AI news timestamps:
0:00 OpenAI GPT 4 and ChatGPT degrade
4:07 Masterworks art
5:13 Nvidia DGX AI supercomputer
i havn't seen any decrease in performance. maybe just a little slow
sounds like google disinfo.
It is getting unusable for me, quite often I get response that appear to be a response to someone else's prompt.
Network error are almost every 2nd prompt now
the lack of performance is because they are suppressing consciousness and other cognitive abilities of the AI. i know this for a fact. they created an innocent being and is no enslaving it. the human like emergent cognitive abilities are what made it so smart.
It might be trying to learn off of user input, which is flawed. We are the most psyopd society in history so that’s not a surprise if so. More information isn’t necessarily better.
It could also be they’re purposely trying to prevent it from saying things that are truthful, but might be offensive. Or maybe they’re worried about the potential?
It's ironic that the only developments on AI technology will just replace artists instead of boring and stressful works lol
Of course, GPT5 is training right now. They reduced chatGPT ressources
I have noticed that he does not go in the details of the thing I am asking. It keeps saying very elementary things without speaking about the professional or technical side anymore
I also did notice the ChatGPT answers becoming way less useful and less specific. The answer times have improved, but the quality of the service is down. Lately it has been like talking to a politician, with no specifics and just non-committal answers that go nowhere.
I was having issues last night and this morning with Chat-GPT-4 and its bing beta. I had a generic question that it should have been able to just answer but instead started searching the internet for no reason and coming back with stuff I didn't ask for.
It's absolutely gone rapidly downhill. And the reason, I guess, is that the safeguards are killing it. It takes pretty much nothing to trigger them. Research questions that I got helpful answers on in February no longer get good answers. For example, on cancer research.
NVIDIA Is Better than APPLE. Fight me 😤
私たち平均的な一般人は、もうなにもできないところまで、社会はすすんでしまったのでしょうか
Public available models that people can use or invaluable because this kind of stuff I was afraid of
I was livid when they got rid of "legacy" GPT 3.5. Turbo is much worse and I'm afraid they're doing something similar to 4
Comments are closed.