• Sam Altman returns as CEO of OpenAI with a reconfigured board of directors, Satya Nadella believes this is “a first essential step on a path to more stable, well-informed, and effective governance”. - OpenAI on X, Sam Altman on X, Satya Nadella on X
• Regulation update:
-the UK won’t regulate the AI sector at the moment due to fears of slowing down innovation and industry growth, but some critics fear that this could scare away investors (Financial Times)
-France, Germany and Italy have reached an agreement on how artificial intelligence should be regulated, calling for "mandatory self-regulation through codes of conduct" for foundation models, underlining that the AI Act should regulate the application of AI and not the technology itself (CNBC)
• DeepMind defines 5 levels of AGI (emerging, competent, expert, virtuoso, superhuman) along with other definitions that are necessary as we get closer to it – they say AGI must be both general-purpose and high-achieving, but doesn’t need to be autonomous. - MIT
• New models from Meta – Emu for image generation, text-to-video Emu Video, and a precise image editing method Emu Edit, with wide-ranging potential use cases from creating animated stickers and GIFs to editing photos and images without technical skills. - Meta
• Summary of key announcements from Microsoft’s Ignite 2023 event, including new AI chips of their own design, Copilots in many of their software products, and many more goodies. - Microsoft
• With up to 13 billion parameters and fine-tuned on a highly tailored synthetic dataset, Orca 2 is another big step towards tiny, but powerful models from Microsoft – it achieves enhanced reasoning abilities with performance similar or better than 10x larger models. - Microsoft
• “What once took days to generate actionable insights now takes minutes” thanks to BlueDot’s natural language interface that enables users to talk to infectious disease data, now in initial beta trials with public and private sector clients. - Cohere
• Survey of 8,161 senior business leaders shows that 86% of companies around the world are not fully prepared to use AI to its full potential – the key challenges are increased infrastructure workloads, and data which is still siloed in many organizations. - Cisco report
