Artificial Intelligence: Energy Aspects скачать в хорошем качестве

Artificial Intelligence: Energy Aspects

In this insightful lecture, originally presented at the Technion on December 28, 2025, Leon Kraversky - founder and current managing director of Soltell Systems and Country Representative for Aleasoft Energy Forecasting - explores the intersection of AI and the energy sector from his 15+ years of professional experience in energy innovation, data analytics, and technology innovation, including being an Energy Advisor to the Foreign Ministry of the Republic of Korea, business innovation advisor to the BGN Technology tech transfer company of the Ben Gurion U-ty and innovation advisor to corporates, funds and energy companies. Key topics covered: Evolution of AI: From early concepts in the 1980s and 1990s machine learning, deep learning, neural networks to the modern large language models (LLMs) - clarifying what truly qualifies as AI today. AI in Energy Applications: Real-world examples, including hybrid AI models for long-term and short-term electricity price forecasting (e.g., accurate 16-year Spanish market predictions by AleaSoft), Sensorless soiling detection in solar PV plants (e.g. accurate soiling metrics with data science). AI as Optimization vs. Enablement: Where AI improves efficiency (e.g., forecasting, energy management) and where it's essential (e.g., ultra-fast energy trading and grid management decisions humans can't handle). AI's Impact on Electricity Demand: Analyzing IEA projections (data centers from ~450 TWh in 2025 to over 800 TWh by 2030), drivers from AI growth, and practical power solutions - favoring fast-deployable natural gas, solar + storage over nuclear/hydrogen due to timelines. Mitigating Factors: Efficiency gains in computer chips, localized AI processing on devices (e.g., Google's Gamma model handling 80% of queries on smartphones), and distributed data centers. Job Market & Broader Effects: AI's influence on energy jobs - net increase projected despite automation, energy transition creating millions of new roles in renewables, and overall positive outlook for developed economies. Outline: 00:00 Introduction. 02:30 It is coming! 04:59 What is the meaning of this? 07:43 Data Science and AI. 10:19 Solving real problems in energy. 11:40 Sometimes AI is a must. 13:46 AleaModel by AleaSoft. 14:25 Do we require AI for everything? 16:23 SysMap by Soltell Systems. 17:07 Enablement vs Optimization. 19:42 AI and the energy demand challenge. 21:33 What power source is fast to deploy locally. 25:13 AI and the network bottleneck. 29:11 Chip energy consumption. 30:35 AI impact on the market. 34:01 Where do we go from here? 36:01 Closing remarks. Leon concludes with an optimistic view: AI will drive energy optimization, reduce emissions, and reshape jobs toward more sophisticated roles - all while technological advances temper the feared surge in power demand. Perfect for students, energy professionals, and anyone curious about AI's real-world energy implications. #Artificialintelligence #EnergySector #AIinEnergy #DataCenters #Renewables #EnergyForecasting