AI In summary, the CEO of OpenAI, Sam Altman, emphasizes the necessity of a breakthrough in energy production to support the advancement of increasingly powerful AI models that demand substantial energy resources. Altman advocates for renewable energy sources such as nuclear fusion and has personally invested $375 million in Helion Energy, a nuclear fusion startup set to supply energy to Microsoft in the coming years.
The energy consumption of AI models, like the previous GPT-3 system, is staggering, with reports indicating it consumed 936 megawatt hours (MWh), equivalent to the annual energy consumption of about 90 households. As the scale of models increases, the energy requirements will also escalate, necessitating further advancements in energy production.
In another development, Google DeepMind introduces AlphaGeometry, an AI system trained to prove geometric theorems at a level comparable to human mathematics Olympiad gold medalists. This system, comprising a language model and a symbolic deduction engine, showcases AI’s ability to reason logically and discover new knowledge, particularly in solving complex mathematical problems.
However, concerns arise regarding the democratization of healthcare through medical AI chatbots. The World Health Organization (WHO) warns that AI systems developed without diverse training data may exacerbate inequities, especially in underserved regions. The WHO advocates for independent audits of AI technologies to mitigate biases and ensure fair and effective healthcare delivery.
Furthermore, Amazon introduces an experimental AI shopping assistant on its mobile app, allowing users to inquire about specific products. Powered by generative AI, this chatbot provides information extracted from product listings but does not offer product comparisons or pricing history. While the assistant’s capabilities are versatile, users are advised to interpret its responses with caution due to the potential for inaccuracies.
