OpenAI CEO Sam Altman and several other company executives hosted an Ask Me Anything (AMA) session on Thursday. The session was hosted on the social media platform Reddit and users were asked to ask questions about the AI company’s products such as ChatGPT or general queries about artificial intelligence (AI) and artificial general intelligence (AGI). During the session, Altman said that GPT-5 will not be released this year, however, the company plans to offer “some very good versions” before the end of 2024.
OpenAI staff is hosting an AMA on Reddit
AMA a class It is hosted on the ChatGPT subreddit. Calling it “our Reddit launch,” several executives including Altman, OpenAI CPO Kevin Weil, Senior Vice President of Research Mark Chen, Vice President of Engineering Srinivas Narayanan, and Chief Scientist Jakub Pachocki participated in the Q&A post. Official X (formerly Twitter) handle for OpenAI as well to publish About Reddit AMA.
Asked about the timeline for a release of GPT-5 or its equivalent, Altman said: “We have some very good releases coming later this year! However, there is nothing we will call gpt-5. This seems on par with what I have confirmed Multiple reports have most people expecting OpenAI to release its next major model sometime in 2025.
Another user asked what value SearchGPT or ChatGPT provides, and Altman said he finds it a faster and easier way to access information. He also stressed that the web search function would be more useful for complex searches. “I’m also looking forward to a future where a search query can return a dynamically customized web page in response,” he added.
Weil also answered many questions from users. Someone asked about the delay in Sora, and an OpenAI program manager said the delay was due to the extra time it took to perfect the model, get the integrity and impersonation right, and the need to scale the compute. But he did not reveal its release date.
Weil also highlighted that the “o” series AI models, such as the GPT-4o and o1-preview, will become a mainstay in the company’s lineup and will appear even after the release of the GPT-5. In addition, he also revealed that ChatGPT’s advanced voice mode can be modified to add a singing voice to the AI.
One user suggested: “Can we get a hands-free way to end a voice chat with ChatGPT? Using the action button shortcut on iPhone, but having to manually dismiss the voice chat every time.” Will seemed to like the idea and said: “I love this idea. Participate with the team now!
OpenAI SVP or Research Mike Chen also answered an important user question about AI hallucinations. Explaining why hallucinations caused by AI models have not completely disappeared, he described it as a fundamentally difficult problem. This is because AI models learn from human-written texts, and humans can often make errors, which are then added to the underlying dataset of large language models (LLMs).
“Our models are getting better at citing, basing their answers on reliable sources, and we also think RL will help treat hallucinations as well — when we can programmatically check whether models are hallucinating, we can reward them for not doing so,” Chen added. .
News Source – www.gadgets360.com
