Google will ‘make things right’ if OpenAI abuses YouTube for AI

Alphabet CEO Sundar Pichai said Google would “fix things” if it decided MicrosoftPowered by OpenAI, it relies on YouTube content to train an AI model that can generate videos.

The comments, in an interview Tuesday with CNBC’s Deirdre Bossa, come after OpenAI chief technology officer Mira Muratti told the Wall Street Journal in March that it wasn’t sure whether the YouTube videos were part of the training data for the company’s Sora model introduced earlier in the year.

Muratti said OpenAI used publicly available data and licensed data. New York Times later reported that OpenAI has transcribed over a million hours of YouTube videos.

Asked if Google would sue OpenAI if the startup violated the search company’s terms of service, Pichai offered no details.

“Look, I think that’s a question for them to answer,” Pichai said. “I have nothing to add. We have clear terms of service. And so, you know, I think usually with these things we engage with the companies and make sure they understand our terms of service. And we’ll fix it outside.”

Pichai said Google has processes in place to find out if OpenAI has broken the rules. Newspapers such as The New York Times have already targeted OpenAI for allegedly infringing copyright law and training models on their articles.

Pichai’s interview followed a keynote to developers at Google’s I/O conference, where executives announced new AI models, including one called Veo that can compose synthetic videos. Those who want to get early access will need to get approval from Google.

OpenAI preempted Google’s event on Monday. The company revealed an AI model called GPT-4o and showed how users of its ChatGPT mobile app will be able to have realistic voice conversations, interrupting the AI ​​assistant to analyze what appears in front of the smartphone’s camera. On Tuesday, Google showed similar upcoming capabilities.

“I don’t think they’ve sent their demo to their users yet,” Pichai said of OpenAI. “I don’t think it’s available in the product.”

OpenAI said in a blog post on Monday that customers of its ChatGPT Plus subscriptions will be able to try an early version of the new voice mode in the coming weeks. Pichai said the multimedia chat capabilities of Google’s Project Astra will appear in its Gemini chatbot later this year.

“We have a clear sense of how to approach it and we’re going to do it right,” Pichai said.

Google has cut the cost of serving AI models in web searches by 80% since previewing it last year, relying on its custom tensor processors (TPUs) and of Nvidia popular GPUs, he said. Google said during the keynote that it is starting to show its AI reviews in search results for all US users

In June, An apple will hold its Worldwide Developers Conference in Cupertino, California. Bloomberg reported in March that Apple was discussing the idea of ​​adding Gemini to the iPhone. Pichai told Bosa that Google has enjoyed a “great partnership with Apple over the years.” A Google expert witness said in court last November that the company gives Apple 36 percent of its search advertising revenue from the Safari browser.

“We’ve focused on delivering great experiences for the Apple ecosystem,” Pichai said. “It’s something we take very seriously and I’m sure – we have a lot of ways to make sure our products are affordable. We’re seeing AI previews today have been a popular feature in iOS when we’ve tested, and so it’s set to continue — including Gemini. We’re going to keep working to get that out there.”

WATCHING: Alphabet CEO on report that OpenAI trained GPT-4 on YouTube: We have clear terms of service