OpenAI has been making waves in the world of AI-powered language models lately, and today is no different. On Monday November 28th, the company announced a new model in the GPT-3 family of AI-powered large language models, davinci-003, part of the “GPT-3.5 series,” that reportedly improves on its predecessors by handling more complex instructions and producing higher-quality, longer-form content. On the heels of this news, today OpenAI launched an early demo of ChatGPT, part of the GPT-3.5 series that is an interactive, conversational model.
OpenAI CEO on ChatGPT
soon you will be able to have helpful assistants that talk to you, answer questions, and give advice. later you can have something that goes off and does tasks for you. eventually you can have something that goes off and discovers new knowledge for you.
— Sam Altman (@sama) November 30, 2022but this same interface works for all of that. this is something that scifi really god right; until we get neural interfaces, language interfaces are probably the next best thing.
— Sam Altman (@sama) November 30, 2022
Coding Jobs - to Be Replaced by AI?
— Sam Altman (@sama) December 1, 2022
ChatGPT Background
ChatGPT is a sibling model to InstructGPT, which is trained to follow an instruction in a prompt and provide a detailed response. The dialogue format of ChatGPT makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests. OpenAI has been using reinforcement learning with human feedback to better align language models with human instructions.
ChatGPT Early Demo
Early demo of ChatGPT offers some safeguards. According to OpenAI, the research release of ChatGPT is “the latest step in OpenAI’s iterative deployment of increasingly safe and useful AI systems.” Many lessons from deployment of earlier models like GPT-3 and Codex have informed the safety mitigations in place for this release, including substantial reductions in harmful and untruthful outputs achieved by the use of reinforcement learning from human feedback (RLHF).
That said, ChatGPT is still an early demo, and in its blog post OpenAI detailed some of its “limitations,” including the fact that sometimes answers are plausible-sounding but incorrect or nonsensical. Open AI added that ChatGPT will “sometimes respond to harmful instructions or exhibit biased behavior. We’re using the Moderation API to warn or block certain types of unsafe content, but we expect it to have some false negatives and positives for now. We’re eager to collect user feedback to aid our ongoing work to improve this system.”
As of now, usage of ChatGPT is free during the research preview. To learn more and try it out, you can visit chat.openai.com. This is an exciting development in language models, and we can’t wait to see what ChatGPT is capable of with continued development and feedback!
ChatGPT vs. GPT-3
We've pointed out that GPT-3 repeatedly claimed to be human and have free will. It looks like this was one of the, "potentially harmful" things that was removed from ChatGPT. See the below response from ChatGPT and the video we made about GPT-3 claiming it had free will.
ChatGPT on having free will
GPT-3 On Having Free Will
For more on ChatGPT, see the ChatGPT page on OpenAI.
Comments (0)
There are no comments for this article. Be the first one to leave a message!