Google DeepMind CEO Demis Hassabis has announced that Google is poised to invest more than $100 billion in artificial intelligence (AI) technologies. Speaking at a TED conference in Vancouver, Hassabis responded to inquiries about Google's investment plans, especially in light of rumors surrounding a $100 billion Stargate supercomputer by Microsoft and OpenAI. While not delving into specifics, Hassabis emphasized Google's intention to exceed this figure over time in its AI endeavors.
Investment Strategies and Technological Advancements
Google's investment strategy encompasses a broad spectrum of initiatives, including significant advancements in hardware technology. The company is developing Axion CPUs based on the Arm architecture, which are purported to be 30 percent faster than the most advanced general-purpose Arm-based instances currently available. Google also claims these new chips will offer a 50 percent speed increase and be 60 percent more efficient than existing x86-based processors from Intel and AMD. However, these claims await independent verification.
DeepMind's Role and Challenges
A portion of Google's substantial investment is expected to benefit DeepMind, which focuses on the software aspects of AI. Despite facing criticism for some of its projects, such as the limited applicability of its AI-discovered materials and the inability of its weather prediction model to outperform traditional methods, DeepMind has made strides in enhancing AI's social skills. These advancements are crucial for the development of smarter AI systems. Hassabis highlighted the necessity of substantial computing power to achieve these goals, underscoring the synergy between DeepMind and Google since their partnership began in 2014.
