AI News This Week: Everything You Need to Know

Why Trust Techopedia
Key Takeaways

  • Audible invites narrators to create AI voice models, offering royalties and control.
  • Meta admitted using public data from all Australian adults to train AI amid tightening privacy laws.
  • OpenAI's o1 model excels in complex math and science but lacks some GPT-4o features.

This week’s headlines highlight SAG-AFTRA’s new AI agreement with game studios, Elon Musk’s denial of a Tesla-xAI licensing deal, and Adobe’s AI video tools coming to Premiere Pro.

Here’s what you need to know.

Video Game Actors Secure AI Agreement With 80 Major Game Studios

SAG-AFTRA secured AI protections and wage improvements for performers following a strike, reaching agreements with 80 video game studios. However, strikes against major publishers like Disney and Warner Bros. continue as the union pushes for broader safeguards against AI misuse.

Nvidia’s AI Chips Cheaper to Rent in China Than US, Despite Ban

Despite export restrictions, Nvidia’s AI chips are cheaper to rent in China, at $6 per hour compared to $10 in the U.S.. The chips are smuggled into China, and Nvidia faces a U.S. antitrust investigation over its AI chip market dominance.

Tesla Will Not Licence xAI Models, Says Elon Musk

Elon Musk denied reports that Tesla would license xAI’s technology for future revenue, stating there is “no need to license anything.” However, he recently mentioned plans to discuss a potential $5 billion investment in xAI with Tesla’s board.

Audible Invites Human Narrators to Create AI Voice Models for Audiobooks

Amazon’s Audible invites select narrators to create AI voice models for audiobooks, allowing them to earn royalties and control which titles use their voices. The beta trial, limited to the U.S., involves refining AI-generated audiobooks and compensating narrators per title.

"Virtual Voices" have made book discovery on Audible.com nearly impossible
byu/MPPreads inaudible

Amazon to Invest £8B in UK, Continues Cloud and AI Expansion

Amazon Web Services (AWS) will invest £8 billion in UK data centers over four years, creating 14,000 jobs annually. The investment is expected to add £14 billion to the UK economy by 2028, amid growing AI demand and competition from Microsoft and Google Cloud.

Taylor Swift Endorses Kamala Harris Following AI Misinformation Incident

Taylor Swift endorsed Kamala Harris for the 2024 U.S. presidential election after AI-generated images falsely linked her to Donald Trump’s campaign. Swift emphasized the need for transparency to combat AI-driven misinformation ahead of the election.

Study Reveals AI Models Hidden Linguistic Bias Against African American English

A recent study found that large language models like ChatGPT and T5 exhibit racial bias against speakers of African American English (AAE), associating them with negative traits and less prestigious jobs compared to Standard American English (SAE) speakers. This bias raises concerns about AI’s use in hiring and law enforcement.

AI-Powered M&A: UBS Tool Analyzes 300K Companies in Seconds

UBS’s new AI tool analyzes over 300,000 companies in seconds to improve M&A deals, generate buy-side ideas, and identify potential buyers and activist targets. Despite its effectiveness, UBS faces compliance and data security challenges.

Mistral Unveils Its First Multimodal AI Model

Mistral’s Pixtral 12B is its first multimodal AI model for text and image processing. Based on the Nemo 12B model, it includes a vision adapter for tasks like image captioning and classification. It’s free for research and is available on GitHub and Hugging Face, but requires a commercial license.

Adobe Will Bring AI-Generated Video to Premiere Pro This Year

Adobe will add AI video features to Premiere Pro and the web by the end of 2024, including tools for extending and generating clips with prompts and images. The company is also enhancing protections for stock content creators and expanding moderation efforts.

Google’s AI Notebook Now Creates Podcasts to Explain Your Sources

Google’s NotebookLM now features an experiment called Audio Overview, where AI hosts create realistic podcast summaries of the source material. While providing a natural-sounding overview, the tool has limitations in accuracy and availability and currently only supports English.

Amazon Set to Integrate Ads Into Its AI Chatbot Rufus

Amazon will add sponsored ads for U.S. users to its AI shopping assistant, Rufus. While ads will be contextually relevant, advertisers won’t receive performance metrics, potentially affecting their ability to gauge effectiveness.

Chinese AI Models Trail US by Six Months

Chinese AI models trail U.S. counterparts by six to nine months but are quickly catching up. Major Chinese firms like Alibaba and Tencent are investing heavily in AI, with expectations to surpass the U.S. in AI adoption soon.

OpenAI’s New O1 AI Model Can Tackle Difficult Math Problems

OpenAI’s new o1 model, launched yesterday, is designed to tackle complex math, coding, and science problems with performance comparable to PhD students. While it excels in these areas, it is still a preview model and lacks some features of GPT-4o, such as file uploads and web browsing.