InvestmentNews’ Post

Compute for AI and crypto uses a ton of energy. But not yet as much as I would have thought. “ChatGPT, the most popular generative AI application, uses only as much electricity as 17,000 U.S. households — out of a total 131.3 million households. That’s only about 0.12% as much as Bitcoin uses. Even if Google used generative AI for every search query, Alex de Vries estimates that the technology would use only about 20% as much electricity as Bitcoin currently uses. But that’s just today. AI scaling is exponential; today’s models take about 100 million times as much compute to train as models in 2012, when the AI boom got started.” From Noahpinion

As Bitcoin, AI, and EV's continue to demand higher Electricity, the need for worldwide infrastructure continues to expand. Who's on your Engineering team? #Conco #Engineering #PowerPlants #Bitcoin https://lnkd.in/gH3zug8s

I knew that #AI demanded a lot of energy but I didn't know there would be literally a #AIpower struggle between AI and #Bitcoin that is going to disrupt everything including #renewable energy and #StrandedEnergy. See my latest blogpost: https://lnkd.in/gyXra69Z

Crypto and AI Set to Double Energy Consumption by 2026, Data Centers Struggling to Keep Up Key Insights: ⚡️The International Energy Agency (IEA) predicts a 30% increase in crypto energy use and a tenfold surge in AI consumption by 2026. ⚡️Data centers, fueled by AI and crypto mining, could demand up to 1,050 TWh of electricity, equivalent to a country’s demand, like Germany or Sweden. ⚡️The US, hosting a third of the world’s data centers, anticipates a 6% increase in electricity consumption by 2026. ⚡️Ireland, with 82 data centers and low corporate taxes, may see a third of its electricity demand driven by data centers by 2026. ⚡️The IEA expects a 40% rise in crypto energy demand by 2026, despite some success stories in reducing environmental footprints. 🌐💡🛡️ Cryptoverse Feed | https://lnkd.in/emsC-DXd

🌍 AI’s Carbon Footprint: A Growing Concern 🌍 A recent C-suite survey reveals that 78% of executives recognize AI’s significant carbon footprint and the urgent need for sustainable AI practices. With AI’s rapid advancement, emissions from major tech companies have surged. Global electricity demand is expected to rise by 3.4% annually until 2026, driven by economic growth and the electrification of homes and transport. Data centers, AI, and cryptocurrency sectors are projected to double their electricity consumption by 2026. As AI computing consumes substantial energy, the question remains: Do its benefits outweigh its environmental impacts, or is it harming our climate? 🤔 #Sustainability #AI #ClimateChange #TechForGood #GreenTech https://lnkd.in/gF2GqBmn

Crusoe Energy Systems is making waves with a groundbreaking $3.4 billion joint venture to build a state-of-the-art AI data center in Abilene, Texas. Originally known for transforming flared gas into cryptocurrency power, Crusoe’s shift to AI infrastructure marks a pivotal turn in its journey. Imagine a facility nearly the size of 20 football fields, housing up to 100,000 GPUs, designed explicitly for heavy AI workloads! This project not only highlights the soaring demand for AI computing but also promises to inject nearly $1 billion into the local economy, creating numerous jobs while prioritizing sustainability. With a Fortune 100 company already on board as a tenant, the confidence in Crusoe’s vision is palpable. As we witness this AI data center boom, the question remains: Are we ready for the implications of such rapid technological evolution? Let’s discuss!

[New Podcast]: Power Markets, Pirates, and Energy Backed Money w/ Andrew Myers (co-founder, Satoshi Energy) https://lnkd.in/gPT_dQDR Andrew and I sit down for a fun conversation while overlooking the San Francisco Bay. There are a few sound/wind issues with my mic (mostly towards the end), but hopefully nothing too disruptive. We focus a lot on energy, but touch on a number of interesting ideas from the edge, including: - The declining cost of solar and batteries - The commodification of solar and Bitcoin miners - The interconnection process in ERCOT - Opening new markets as the cost of solar continues to decline - Bitcoin Mining vs. AI compute markets (price and uptime are two major factors) - Reforming legacy power markets vs building new ones with microgrids (micromarkets enable microgrids) - What a power market does: 1) match supply and demand to create price at each node and 2) serve as clearing house - #Nostr as the communication protocol for decentralized energy markets? - Pirates once ruled the world because they were free to experience different modes of life. Do they still? - Andrew’s vagabond/pirate ethos and lifestyle experiments - Andrew’s other business idea: a tech enabled private military as an insurance company - How does humanity evolve into the hivemind? - Patience as the ultimate virtue If you're feeling the show, please listen on Fountain and stream some sats! Full discussion on Stacker News: https://lnkd.in/gukyTR29

Bitcoin uses as much electricity as a small nation. AI will use even more. All EV vehicle fleet = produce and distribute 50% more electricity. Our current grid can’t handle bad weather events. That’s why we need investment in researching new energy technologies.

AI is poised to threaten the electrical grid - but not in the way you might be thinking. Think about the amount of energy that data miners needed to mine bitcoin. Now, think at the very least 3 times that power need. The amount of power needed for data centers focused on AI is going to change the electricity dynamic and we’re only in its infancy. Add EV’s to the equation and our electrical grid comes up short - way short. “It is expected that data centers will triple their power consumption over 2022 levels, up to 390 terawatt hours.” We need to look at new ways to add generation - #accessabsolutelyanywhere.