Also, Adobe launches an app for creators to protect them from deepfakes and scraping from AI bots.
The CEO and director of DeepMind shared a Nobel Prize in Chemistry this week for their work on AlphaFold. / Adobe Stock
Pioneering AI researcher Geoffrey Hinton and Princeton physicist John Hopfield were awarded the Nobel Prize in Physics on Tuesday for their early work on neural networks, the technology that laid the foundation for large language models.
Hinton is also a former Google researcher who left his post earlier this year to raise alarms about the dangers of AI.
On Wednesday, the Nobel Prizes for Chemistry went to Demis Hassabis and John Jumper – the CEO and director of DeepMind, respectively – for their work on AlphaFold and to University of Washington biochemist David Baker, who created a novel protein through computer software.
Advertisement
At a tech event in Taiwan on Thursday, Advanced Micro Devices (AMD) revealed the Instinct MI325X, a new GPU intended to rival Nvidia’s currently industry-leading chips. Production of the chips will start by the end of this year, according to CNBC.
While Nvidia has dominated the market for AI chips in recent years – raking in $30bn in revenue in the second quarter of this year – it has been facing pressure from a growing number of rivals, most notably AMD.
On Tuesday, Adobe announced the launch of a new app, Adobe Content Authenticity, designed to protect creators against the unauthorized use of their content and to ensure appropriate attribution.
Through the app, creators can quickly apply ‘content credentials‘ – attached metadata that Adobe describes as “a ‘nutrition label’ for digital content” – to their digital IP. They can signal, for example, if they don’t want their content to be used to train generative AI models. Content credentials are also intended to prevent the non-consensual spread of AI-generated copies of creators‘ work or deepfakes of their likenesses.
Advertisement
AI was a core focus at Advertising Week New York this week. Several talks and panels explored the technology’s growing role in the marketing sector. Like last year, much of the rhetoric revolved around the technology‘s capacity to support marketers in their work and to free them up from mundane tasks, while the possibility of job loss caused by automation was largely denounced or unaddressed.
Advertisement
A week ago, Meta unveiled Movie Gen, a suite of AI models that can – among other things – produce personalized video clips with synchronized sound from still images and a text prompt. The models, which are being primarily marketed to filmmakers, also allow users to edit video clips by adding AI-generated elements or backgrounds.
Like OpenAI’s Sora, Movie Gen has not yet been released to the public, but marketing experts contacted by The Drum earlier this week are already expressing excitement about the tool’s potential.
For more on the latest happenings in AI, web3 and other cutting-edge technologies, sign up for The Emerging Tech Briefing newsletter.