Skip to main content

OpenAI transcribed over a million hours of YouTube videos to train its LLMs, Google engaged in same practice

In order to access more reputable English language-based text on the internet in 2021, OpenAI researchers created a speech recognition tool called Whisper, writes The New York Times. It was designed to transcribe audio from YouTube videos, giving the company a trove of data to train its LLMs.

Read Entire Article


https://www.techspot.com/news/102536-openai-transcribed-over-million-hours-youtube-videos-train.html?utm_source=dlvr.it&utm_medium=blogger

Popular posts from this blog

OpenAI faces ChatGPT probe over data risk issues, damaging reputations

The agency sent a 20-page demand letter to OpenAI for records about how it addresses risks related to its AI models, writes The Washington Post. The Civil Investigative Demand also requests details about the company, its AI model and training, how it deals with privacy and prompt injection attacks, API... Read Entire Article https://www.techspot.com/news/99395-openai-faces-ftc-probe-over-chatgpt-damaging-reputations.html?utm_source=dlvr.it&utm_medium=blogger

Apple's virtual WWDC kicks off today, watch the keynote here

The keynote address starts Monday, June 22 at 1:00pm ET / 10am PT / 6pm BST. You can stream it live here via YouTube or check it out on Apple's website. https://bit.ly/2VbuvXk

Bill Gates says we don't have to worry about AI energy use

Speaking at a London event last week hosted by his Breakthrough Energy venture fund, Gates said, "Let's not go overboard on this." Read Entire Article https://www.techspot.com/news/103617-bill-gates-dont-have-worry-about-ai-energy.html?utm_source=dlvr.it&utm_medium=blogger