OpenAI CEO Admits that the Company is Not Actively Training GPT-5

OpenAI

OpenAI CEO admits that the company is not actively training GPT-5 on improving the newest abilities

OpenAI CEO admits that the company is not actively training GPT-5 are agreed with portions of the letter, such as the need to ensure AI models are safe and consistent with human values. According to CEO Sam Altman, OpenAI is not training a fifth version of its generative pre-trained transformer (GPT) and is instead focused on improving the capabilities of its newest GPT-4 model.

ChatGPT Bots are Spamming Reddit: The Reddit forums are being inundated with comments created by ChatGPT and uploaded by bot accounts, leaving moderators unable to keep up with the mounting volume of spam. These spam bots frequently promote sexual material, illicit narcotics, and shady sellers’ items. The bot problem was already severe, and Reddit’s automatic anti-spam algorithms provided little assistance, and by the time they did, it was too late, and the bot’s presence had usually accomplished its goal. Several hundred bot accounts have purportedly been banned from Reddit, but they continue to grow on the site, uploading AI-generated material and flooding automatic systems that flag questionable information.

OpenAI Launches Bug Bounty Program: The objective of OpenAI is to develop artificial intelligence systems that benefit everyone. To that end, we spend extensively on research and engineering to assure the safety and security of our AI systems. “However, we understand that, as with any complex technology, vulnerabilities and flaws can emerge,” the business stated in a statement. The bug bounty system does not cover prompts that circumvent content controls meant to make OpenAI’s models create inappropriate material, a method known as “jailbreaking.”

Musk Supplies Twitter Engineers with Shiny New GPUs to Build Generative AI: Musk recently signed an open letter calling for a six-month moratorium on training huge AI models due to concerns that the technology would grow too advanced to regulate. Despite this, he is still developing AI and adding new features to his microblogging site.

The post OpenAI CEO Admits that the Company is Not Actively Training GPT-5 appeared first on Analytics Insight.

Follow us on Twitter, Facebook
0 0 votes
Article Rating
Subscribe
Notify of
guest
0 comments
Oldest
New Most Voted
Inline Feedbacks
View all comments

Latest stories

You might also like...