Technology

Weekly AI Recap: OpenAI loses CTO & heads of research, Meta unveils Llama 3.2

By Webb Wright, NY Reporter

The Drum

September 27, 2024 | 7 min read

Listen

Listen to article 4 min

Also, The New York Times confirms iPhone designer Jony Ive is working with OpenAI CEO Sam Altman on an AI hardware project.

More OpenAI execs depart

OpenAI chief technology officer Mira Murati announced that she was leaving the company “to create the time and space to do my own exploration,” she wrote in an X post on Wednesday, without further elaborating on what that exploration might entail.

Murati has become a well-known figure in Silicon Valley since being appointed CTO at OpenAI, making frequent media appearances. She was briefly named as the company's interim CEO following Sam Altman’s ouster last year, a debacle that ended with his reinstatement days later.

Later on Wednesday, Bob McGrew and Barret Zoph – OpenAI’s chief research officer along with a VP of research, respectively – also announced via X posts that they are leaving the company, using similarly vague language.

Murati, McGrew and Zoph are the latest OpenAI executives to depart in what’s become something of an exodus among the company’s core team. Ilya Sutskever, for example, one of the company’s cofounders and its former chief scientist, left in May and has since launched his own AI company.

The trio announced their departures just hours before reports broke about OpenAI’s efforts to transition into a for-profit company, a move that would allow it to generate enough revenue to keep pace with its expenses. It’s currently overseen overseen by a nonprofit board, which includes Altman.

Meta leans further into AI & AR

During its annual developer conference on Wednesday, Meta emphasized its ongoing push into AI and augmented reality (AR).

The company debuted Llama 3.2, the latest and most powerful iteration of its open source large language model, which the company has positioned as a viable competitor to proprietary models from other AI industry giants like OpenAI and Google. Two versions of Llama 3.2 – one trained with 11bn and the other with 90bn parameters – are multimodal, capable of interacting both with text and images.

However, the most significant announcements from the event concerned upgrades to the company’s smart glasses, which were developed in partnership with Ray-Ban. In a live demo during his keynote address, company CEO Mark Zuckerberg showed how users can communicate in real-time in two different languages using a built-in translation feature. The AI-powered glasses enable users to perform other useful tasks, like remembering where they parked by simply looking at their car.

Though Meta also revealed some VR-related hardware upgrades, these weren’t the focal point of the event. It’s clear that the company is veering from its previous focus on the metaverse toward AI-powered products that it believes will have a more immediate and tangible appeal for consumers.

Want to go deeper? Ask The Drum

Jony Ive confirms AI hardware ambitions

Ex-Apple iPhone designer Jony Ive confirmed that he’s working with Sam Altman to build a new AI device in a profile published by The New York Times on Saturday. Reports of the collaboration – which are said to be aimed at building the “iPhone of artificial intelligence” – first started to trickle in last fall.

While details around the project are shrouded in mystery, Ive and Altman are expected to raise $1bn for the project from investors by the end of this year, The New York Times reports.

OpenAI unveils Advanced Voice mode

OpenAI has begun rolling out a new, more sophisticated voice feature called Advanced Voice for ChatGPT Plus and Team subscribers.

The upgrade includes five new voices, two male and three female, in a range of accents. Users can also access a text page in their settings to specify how they’d like the model to interact with them, adjusting for factors such as cadence and speed.

The new voice features are not yet available in the EU, the UK, Switzerland, Iceland, Norway or Liechtenstein, OpenAI writes on its website.

Ad targets Silicon Valley execs who support Trump

A new ad from the political action committee The Lincoln Project, comprised of Republicans who oppose Donald Trump, paints tech investors and business leaders who support Trump’s reelection as hypocrites, arguing that these techies almost certainly wouldn’t hire him at any one of their companies.

Some tech industry giants, including Elon Musk and Marc Andreessen (neither of whom are mentioned by name in the new ad), have voiced their support for Trump. His perceived hard-line stance towards China, coupled with his reduction of corporate tax rates during his tenure as president and his commitment to deregulating the AI industry, are all factors that have appealed to some tech moguls.

For more on the latest happenings in AI and other cutting-edge technologies, sign up for The Emerging Tech Briefing newsletter.

Technology

More from Technology

View all

Trending

Industry insights

View all
Add your own content +