OpenAI’s mission is to ensure that artificial general intelligence (AGI)[1]
By AGI, we mean highly autonomous systems that outperform humans at most economically valuable work.
benefits all of humanity. We therefore think a lot about the behavior of AI systems we build in the run-up to AGI, and the way in which that behavior is determined.
Since our launch of ChatGPT, users have shared outputs that they consider politically biased, offensive, or otherwise objectionable. In many cases, we think that the concerns raised have been valid and have uncovered real limitations of our systems which we want to address. We’ve also seen a few misconceptions about how our systems and policies work together to shape the outputs you get from ChatGPT.
Below, we summarize:
How ChatGPT’s behavior is shaped;
How we plan to improve ChatGPT’s default behavior;
Our intent to allow more system customization; and
Our efforts to get more public input on our decision-making.
OpenAI’s mission is to ensure that artificial general intelligence (AGI)[1]
By AGI, we mean highly autonomous systems that outperform humans at most economically valuable work.
benefits all of humanity. We therefore think a lot about the behavior of AI systems we build in the run-up to AGI, and the way in which that behavior is determined.
Since our launch of ChatGPT, users have shared outputs that they consider politically biased, offensive, or otherwise objectionable. In many cases, we think that the concerns raised have been valid and have uncovered real limitations of our systems which we want to address. We’ve also seen a few misconceptions about how our systems and policies work together to shape the outputs you get from ChatGPT.
Below, we summarize:
How ChatGPT’s behavior is shaped;
How we plan to improve ChatGPT’s default behavior;
Our intent to allow more system customization; and
Our efforts to get more public input on our decision-making.
This is a developer community for coders and that blog post was written for ChatGPT users and as a public relations piece, and not written for developers…
Since the start of alphabet / reading / writing, we’ve had…
Human Scribes
Printing Press
Typewriter
Word Processing / PC
Blogs / Internet
GPT / AI
There’s a lot of time between the early items in the list, and a lot less time as new advances come along. Those who control the words hold the power in society usually. Each step has also led to more power through a sort of force multiplier effect…With more people having access (usually!)…
All that said, it’s good to see an attempt at AGI for everyone. It’s important, I believe. Most people don’t understand the societal changes that are coming - and how quickly they’ll be implemented. People have been saying that for years, but the tech has caught up to the dreams…
That’s an interesting concept to bring up regarding custom fine-tuning. In the present scenario, some of my concerns are related to if the model is being trained on more data that is semantically negative.
I having a slight optimistic personality, would prefer to have a model that adds on to it to discover the better side with my conversations.
I have tried out a few examples, and here is one of the responses (looped- he always expresses concerns, not pleasure about the progress
About 60% of humans are optimistic naturally, and this is a trait that can be learned. ChatGPT has been trained and is being trained with all the internet that human beings have populated.You have a fantastic observation with chatGPT. @Miraj I will ask for a realistic optimists viewpoint and play around with this concept more.
