Ad
Ad
Ad
Short

Microsoft plans to invest $2.9 billion in Japan to boost the country's AI capabilities, research, and cybersecurity, the company said in a statement. The tech giant will double its hyperscale cloud computing and AI infrastructure, train over 3 million people in AI skills in the next 3 years, establish the first Microsoft Research Asia lab in Japan, and work more closely with the Japanese government on cybersecurity. Microsoft says the investments aim to speed up Japan's digital transformation and AI adoption. The company also wants to support research partnerships with Japanese universities and help policymakers improve cybersecurity for the government, businesses, and society. The news comes as OpenAI recently opened an office in Tokyo. Microsoft is also investing in AI infrastructure in several EU countries.

Ad
Ad
Short

In an interview on X, Musk said that his AI startup, xAI, is currently working on a second version of its Grok model, which he believes will be better than GPT-4. He expects it to be ready in May, followed by new, much more capable models. Musk's AI startup, xAI, recently released Grok 1.5, which, despite some improvements, cannot keep up with the top AI models. Grok 1 is available as open source, but there are smaller, more efficient, and more capable open-source models. Musk also said that he expects AI capabilities to surpass human intelligence by the end of next year, assuming power and hardware availability can keep up with the demands of increasingly powerful technology. Recently, the chip shortage has limited AI progress.

Short

Demis Hassabis, founder of AI lab DeepMind, is reportedly "deeply frustrated" with Google's merger and focus on commercializing AI, insiders tell The Information. The merger of Google Brain and DeepMind has not gone smoothly, and tensions remain. Hassabis even considered leaving Google before the AI units were merged into a new lab. He also reportedly told a colleague that it might be difficult for Google to catch up with OpenAI's video AI Sora. Hassabis reportedly made some organizational changes to restore the importance of pure AI research at Google. Hassabis also complained about the exodus of employees to OpenAI and the coverage of these departures. The pressure to show progress in AI may have led to an exaggeration of the capabilities of DeepMind's systems. But Hassabis remains convinced that artificial general intelligence is within reach.

Ad
Ad
Short

Mustafa Suleyman, co-founder of DeepMind and new CEO of Microsoft AI, announced the opening of a new AI hub in London. The Microsoft AI London Hub will work on advanced language models and their supporting infrastructure, as well as develop tools for foundation models. Jordan Hoffmann, a former Inflection and DeepMind employee, will lead the hub. Microsoft plans to make a long-term investment in the region and hire top AI scientists and engineers, Suelyman says. He, Hoffmann, and about 60 AI experts recently joined Microsoft through its indirect acquisition of UK-based AI startup Inflection AI. Inflection still exists, but is now a skeleton crew offering services for fine-tuning AI models. The move allowed Microsoft to acquire a lot of AI talent in one fell swoop, build a stronger AI presence in Europe, and eliminate a potential OpenAI competitor. Competition authorities are already keeping a close eye on Microsoft because of its massive investments in AI startups like OpenAI.

Google News