Chris Whitlock’s Post

Where's the juice? Are we running out of #energy faster than anticipated? Can our #powergrid sustain the demand of #AI, #clean-tech, and #utilities? Peak Nano finds this article powerful. Amid explosive demand, America is running out of Power https://hubs.la/Q02nKSLL0

Today’s The Washington Post highlights the US’ rising power needs, driven in part by technology, the latter including AI. I am not an AI expert, but my sense from research I do is that tech companies already see this limitation on the horizon and are actively working to find solutions. Check out Liquid AI as an example of what can be done with less computational power. https://lnkd.in/eeaGf-ns

AI is changing the game, not just for data processing but also for energy consumption and sustainability efforts. And now, with the transformation around mobility, the ride is only getting more interesting for the grid. #AI #energyconsumption #sustainability

I found this paper about AI and data centers energy consumption very interesting. https://lnkd.in/eay3fhzJ

With the onslaught of AI, power needs will continue to rise.

Interesting story in The Washington Post about growing demand for electricity, driven - in part - by increasing need for compute power driven by AI technologies. This highlights another overlooked cost of implementing AI. Certainly, AI technology can do great things. But software developers must always keep in mind that the value generated by AI-infused solutions must surpass the costs of making AI solutions a reality. The bar that must be cleared to achieve profitability in the AI realm keeps rising. #artificialintelligence #BusinessExpenses #SoftwareDevelopment

Last week, a major energy conference was dominated almost entirely by one topic: artificial intelligence. Tech leaders and executives ranging from Amazon to Microsoft warned that the main limitation of AI development was finding enough power sources to supply data centers. Why don’t we have enough power? US demand for electricity has taken a nosedive the past few decades. However, that changed as the AI frenzy took hold. Companies became set on building data centers, which are colossal buildings that hold the computing infrastructure that trains large-language models (like ChatGPT). They just keep coming: Earlier this month, Oracle Chairman Larry Ellison claimed that he’s building an AI data center that could fit eight Boeing 747s stacked in front of one another. We need more power According to the Washington Post, these gigantic data centers that train AI systems need “exponentially more power” than traditional data centers, causing utilities to dramatically upgrade their demand projections. https://lnkd.in/gtUTPE4S

Is America running out of power? The thought has stirred up a lot of conversation recently. Data centers and large warehouses of computing structure built to manage a rapidly innovative artificial intelligence landscape, along with the proliferation of crypto-mining are a huge strain on our nation's power grid. According to experts, this situation is critical in several regions of the United States. Read more from The Washington Post ⬇️ #powergrid #electricity #ai #artificialintelligence #cryptomining

An article in the Washington Post caught my eye over the weekend highlighting demand for electricity in the U.S. surging due to the rapid growth of data centers partly fueled by the huge demand in AI. The rapid growth of AI and other computing-intensive applications is contributing to rising energy demand and, by extension, greenhouse gas emissions. Training and running large AI models consumes significant amounts of electricity. A 2019 study by the University of Massachusetts found that training a single large language model can emit as much carbon as five cars over their lifetimes. As AI scales up, this energy footprint will grow. The WP article notes that data centers' share of U.S. electricity consumption could rise from 4% in 2022 to 6% by 2026, driven largely by AI growth. If that power comes from fossil fuels, it will exacerbate global warming. While major tech companies are investing in renewable energy to offset their AI energy consumption, the scale and pace of clean power deployment may not keep up with exponential increases in computing power demand. The AI boom, for all its potential economic and societal benefits, risks locking in more carbon-intensive energy infrastructure at a time when we urgently need to decarbonize. So what should be done ? - Ensuring AI is powered by clean energy to the greatest extent possible, through a combination of tech companies procuring renewables and policies driving grid decarbonization. - Incentivizing the development of more efficient AI models and hardware to reduce computing's energy intensity. This should be done at a governmental policy level. - Considering the lifecycle emissions of AI when making policy and investment decisions. Realizing AI's potential while reining in global warming will require concerted efforts to align the technology's trajectory with climate imperatives. A failure to do so could undermine progress on decarbonization at a critical time. https://lnkd.in/e5Nb9dsy #ai #enterpriseai #ethicalai #greenai

Data centres account for ~1% of GHG emissions - they are power hungry. An increased in data centres is an increase in power requirements. When we are looking towards a decarbonised world, we often think of our current energy requirements. However, we should be thinking of our future energy requirements that new technology requires. This also goes for industries like mining where a company is considering going up the value chain. A company's energy profile will change as new activity are undertaken, technology deployed and resources are progressed.