Welcome to the newest edition of FindBiometrics’ AI update. Here’s the latest big news on the shifting landscape of AI and identity technology:
OpenAI has revealed Sora, its first text-to-video generative AI system. Other AI startups have already made some pioneering efforts in this space. Sora appears to be way ahead of them, but OpenAI says it’s working with “red teamers” specializing in areas such as misinformation before proceeding with a wide launch.
OpenAI’s revenues have now reached $2 billion on an annualized basis, as of December. Its annualized revenues had reportedly only just reached $1.3 billion in October, suggesting considerable pace. OpenAI’s management are aiming to double the $2 billion figure in 2025.
Google is rolling out an upgraded version of its Gemini AI platform, dubbed “Gemini 1.5 Pro”. It can process more information in a single query: Up to an hour of video, 11 hours of audio, and documents amounting to more than 700,000 words. Cloud customers and developers will be the first customers to get access to it.
Together AI is looking to raise $100 million in a funding round that is expected to bring its valuation to over $1 billion. The 20-month-old startup offers customers access to cloud-based infrastructure for AI training. Salesforce Ventures is reportedly leading the funding round.
Lambda, another startup offering cloud computing services for AI training, has raised $320 million in a new funding round that featured multiple new investors including SK Telecom and Super Micro Computer Inc. Lambda was founded in 2012, and is now valued at $1.5 billion.
G42, the United Arab Emirates-based AI company, has sold its stakes in ByteDance and other Chinese companies to appease partners based in the US. The move comes after CEO Peng Xiao had acknowledged in December that his company “cannot work with both sides” in the escalating tech feud between the US and China. G42’s American partners include Microsoft and OpenAI.
Hugging Face has announced a new “”Provenance, Watermarking and Deepfake Detection” collection of AI training tools. The company is known for its open-source library of pre-trained AI models. The new collection was announced by Margaret Mitchell, Hugging Face’s chief ethics researcher, who warned about “the rise of AI-generated ‘fake’ human content”.
The Australian government has appointed a 12-member expert committee to identify the riskiest applications of AI technology. In announcing the group, Industry and Science Minister Ed Husic noted that officials in the European Union have identified risky AI use cases as they have hammered out the pending AI Act, including the risks of “biometric identification and categorization of people”.
The chatbot’s take: We asked for some insight about OpenAI’s branding efforts.
–
February 15, 2024 – by Alex Perala
Follow Us