Newsletter
Highlights:
Nvidia and Pfizer invest in Israeli medical AI tech firm CytoReason to advance research.
China ramps up censorship of large language models in accordance with socialist regime policies.
Trump’s allies seek to launch the Manhattan Projects to put “America First in AI” and remove regulations.
Credits to REUTERS/Tyrone Siu
Business
Nvidia, Pfizer lead $80 mln funding for Israeli medical AI tech firm CytoReason (Reuters)
Nvidia and Pfizer are among some of the investors to raise $80 million in a private funding round for CytoReason, an Israeli company using AI to develop disease models. The company aims to expand its application of its models into additional indications and grow its proprietary molecular and clinical data. "The rapid expansion of new technologies, like artificial intelligence, holds tremendous potential to help transform what is possible in human health," said Mikael Dolsten, Pfizer's chief scientific officer.
Anthropic launches $100 million AI fund with Menlo Ventures, ramping up competition with OpenAI (CNBC)
Anthropic is creating a $100 million fund with Menlo Ventures, modeled after the iFund 2008 partnership between Apple and Kleiner Perkins. The fund is to back early-stage startups and get them using Anthropic’s artificial intellignce technology to gain an edge over OpenAI. Menlo will put up the cash to invest in the startups while Anthropic will give founders $25,000 in credits towards its large language models. Anthropic is also offering coaching, quarterly meetups, and other resources without taking any financial stake in the startups.
Credits to Getty Images
Politics
China deploys censors to create socialist AI (FT)
Chinese government officials are testing artificial intelligence companies’ large language models to ensure their systems “embody core socialist values”, in accordance with the country’s stance on censorship. The Cyberspace Administration of China (CAC) has forced large tech companies such as ByteDance and Alibaba to take part in a mandatory government review of their AI models.
Britain's new government aims to regulate most powerful AI models (Reuters)
Britain’s new Labour government has said it seeks to effectively regulate artificial intelligence models, but they have yet to propose any specific laws. On Wednesday, King Charles set out newly-elected Prime Minister Keir Starmer’s legislative agenda in a speech to open the new session of parliament. Starmer has promised to establish appropriate legislation on “the most powerful artificial intelligence models” but is taking a cautious approach on rolling them out.
Trump allies draft AI order to launch ‘Manhattan Projects’ for defense (WashingtonPost)
Former president Donald Trump’s allies are drafting an AI executive order in the form of “Manhattan Projects” to develop military technology and immediately review “unnecessary and burdensome regulations” - a move that is favorable to Silicon Valley investors and companies. The framework would also create “industry-led” agencies to evaluate AI models and secure systems from foreign adversaries.
Credits to The Verge/Allison Johnson
Technology
Samsung’s new image-generating AI tool is a little too good (The Verge)
Samsung’s new Galaxy Z Fold 6 comes with a novelty AI tool called “sketch to image.” The user presents a rough sketch and the generative AI handles the rest to create something more detailed from a handful of styles like “3D cartoon” or “illustration.” What’s particularly impressive is that it’s capable of generating photorealistic images. The writer drew a comically large sketch of a bee over field of flowers, expecting it to generate a large in-focus bee. But what it generated, was a normal size bee with the appropriate blur to match the background as if the photo was simply captured at the perfect moment (See image above.)
Want to spot a deepfake? Look for the stars in their eyes (RAS)
New research at the Royal Astronomical Society suggests that AI-generated fakes can be spotted by analyzing human eyes in the same way that astronomers study pictures of galaxies. The work, by University of Hull MSc student Adejumoke Owolabi, focuses on the way light reflects off the eyeball. Fake images often lack consistency in the reflections between the left and right eye, whereas real images show the same directional reflections in both. It’s not a “silver bullet for detecting fake images” says Professor Pimbblet, an astrophysicist at the University of Hull, but “this method provides us with a basis, a plan of attack, in the arms race to detect deepfakes.”
New ways to get creative with Microsoft Designer, powered by AI (Microsoft)
Microsoft announced that the Designer app is now generally available through Microsoft Copilot, with new features to help you create with the power of AI. With a Copilot Pro subscription, users can use AI to generate images or designs right into Word or Powerpoint. Microsoft Designer is also available as a free mobile app and in Microsoft Photos.
Credits to Spotify
Entertainment
Spotify launches a new voice and language for its AI DJ (The Verge)
Spotify is offering a new Spanish-language voice option for its AI DJ feature, originally launched February 2023. Users will be able to switch between Livi, the female spanish-speaking voice, or X, the male English-speaking voice. The Spotify AI DJ is similar to a radio announcer, providing commentary and context to a daily rotating selection of playlists. At the moment, the AI DJ is only available to Spotify Premium members.
