The two main factors that slowed down the development of artificial intelligence (AI) years ago were the availability of training data and computing capacity. But both obstacles have been largely overcome recently. The use of synthetic data versus real data, originated through deep learning algorithms, is increasing the volume of data available and avoiding biases in the training data. And, additionally, we have witnessed an extraordinary development of computing capacity, unimaginable just a few years ago. But now a new “limiting factor” has been confirmed, which the advisor warned at the last World Economic Forum in Davos OpenAI delegate Sam Altman: energy. Current projections suggest that global investment in data centers (essential infrastructure in the development of AI) will reach $50 billion in 2030. If it materializes, this energy consumption will be equivalent to 21% of the world total, compared to at the current 3%. To address this challenge, a study by Afi suggests that consolidating professional data centers, instead of traditional servers, could reduce their energy demand by up to 60%. AI itself could be the one to find the solution. A recent Microsoft publication points out that the combination of the latest advances in AI and high-performance computing (HPC) has managed to reduce the inorganic materials needed for the development of new batteries from 32 million to only 18 million. This process, which using the traditional scientific method would have taken years to carry out, was completed in a matter of weeks. The impact of AI is not only limited to the technology sector. The case of energy is another line. The volume of fundraising by start-ups in the sector has increased considerably. Helion Energy, an “energy” company that Altman financed with $500 million in 2021, promises significant advances in the use of the first commercial nuclear fusion reactor in 2024. In addition, it has signed agreements to supply energy to Microsoft in 2028. Without However, not only Altman gives financial support to this company. Other relevant personalities in the technological field such as Elon Musk or Moskovitz (co-founder of Facebook) and venture capital entities are betting on the development of this new form of energy. Investment by large technology companies (Google and Meta support nuclear fusion) pushes the boundaries of what is considered “technology” and involves a kind of large-scale vertical integration: generating its own energy to support its data centers. Will the future energy companies be the current technology companies? David Cano and Javier Pino are professors at Afi Escuela. Follow all the information on Economics and Business on Facebook and xor in our weekly newsletter
The Five Day Agenda
The most important economic quotes of the day, with the keys and context to understand their scope. RECEIVE IT IN YOUR EMAIL
Subscribe to continue reading
Read without limits_