🇺🇸▼
Artificial intelligence systems use a huge amount of electricity. Major tech companies are building large new data centers to run these programs. To power these massive facilities, companies are turning to new nuclear reactors. These next-generation designs promise to be safer and cheaper than older plants. They are a great choice for the energy-heavy facilities needed for advanced AI. We recently held a discussion for subscribers about these important topics. The talk focused on huge data centers and the future of nuclear power. Both subjects are on the MIT Technology Review list of ten Breakthrough Technologies for 2026. The link between AI growth and nuclear energy is growing stronger every day.
Demis Hassabis is the CEO of Google DeepMind. He recently responded to a viral post with a short comment. He said, "This is embarrassing." His remark was about a post by an OpenAI research scientist named Sébastien Bubeck. Bubeck claimed two mathematicians used the new GPT-5 model to solve ten hard math problems. This online exchange in mid-October shows how people exaggerate what AI can do. It shows how social media spreads unrealistic claims quickly. This fuels a culture of hype that hides the true limits of the technology. While AI is advancing fast, false stories create confusion. People need to understand the real limits of these systems.
Extreme heat is a severe problem for many nations. During the summer of 2025, heat waves caused power failures in North America, Europe, and the Middle East. As climate change increases the need for air conditioning, electrical systems face huge strain. A concept called radiative cooling could provide a solution. Special paints and coatings can scatter sunlight and release heat into space. These materials do not need extra energy to cool things down. They offer a sustainable way to lower indoor temperatures. By reflecting the sun and letting heat escape, these new materials can help cities survive heat waves. They can do this without overloading power plants.