Welcome to the latest edition of FindBiometrics’ AI update. Here’s the latest big news on the shifting landscape of AI and identity technology:
AMD held a launch event for its MI300 AI chips, which are being pitched as serious rivals to Nvidia’s dominant offerings. Whereas the MI300A is meant mainly at supporting the complex work of supercomputers, the MI300X will compete more directly with Nvidia’s H100 chip and the forthcoming H200 chip. AMD says the MI300X is 10-20 percent better than the H100 at inference, and that it generally offers a better performance on certain applications. It should be available in Q1 of 2024.
Google has launched a new AI model, Gemini, claiming that it beats every other AI system in 30 of 32 standard industry benchmarks. A simplified version has been integrated into the company’s Bard chatbot, while the most advanced version is still being tested for safety concerns. Gemini is multimodal, having been trained on text, video, and audio content.
Vast Data raised $118 million in a Series E funding round, bringing its total valuation to $9.1 billion. The New York-based startup offers a large-scale data storage platform aimed at supporting AI workloads, and has benefitted from surging interest in recent months.
Elon Musk’s AI startup, X.AI, is looking to raise $1 billion in investment funding, and has so far raised $135 million in pursuit of that goal. Prior to the funding round, the startup’s capital had been provided entirely by Musk himself. The startup has begun rolling out its flagship chatbot, Grok, to premium subscribers of Musk’s social media platform, X.
G42, the United Arab Emirates’ leading AI company, says it will phase out China-made hardware as a result of US sanctions. Speaking to the FT, CEO Peng Xiao said his firm is “in a position where we have to make a choice,” adding, “We cannot work with both sides. We can’t.”
Meta and IBM jointly launched a coalition dedicated to open source AI technologies. Over 50 organizations have joined the “AI Alliance”, including Intel, Oracle, the National Science Foundation, and Cornell University. They plan to pool their resources to develop solutions that won’t compel businesses to rely on a single vendor, such as OpenAI.
Daon has launched xSentinel, a voice deepfake detection tool, as part of its AI.X platform. Primarily aimed at voice communication channels such as call centers, xSentinel is designed to scan for various cues that a given voice is AI-generated within seconds of the start of a call. Daon first launched its AI.X platform in June of this year.
The California Privacy Protection Agency (CPPA) is preparing to hold deliberations on its draft regulations for Automated Decision-Making Technologies (ADMT). Issued near the end of last month, the draft regulations are aimed at lining up the CCPA’s rules and enforcement actions with the Consumer Privacy Protection Act. ADMT are essentially defined as technologies that process data to execute decisions or aid human decision-making.
Cali Group is getting ready to open what it calls the world’s first fully autonomous restaurant. The “CaliExpress by Flippy” burger joint will feature Miso Robotics’ “Flippy” robot chef, and will allow customers to pay using PopID’s biometric payments system. Cali Group had earlier experimented with both technologies at its CaliBurger restaurants in 2020, partly as an effort to mitigate the spread of COVID-19.
The chatbot’s take: We asked ChatGPT about the AI Alliance’s prospects:
–
December 7, 2023 – by Alex Perala
Follow Us