AI AND ENERGY USE: A GROWING CONCERN

Over the last few months, we’ve written extensively about Artificial Intelligence. Mostly we’ve been focusing on the usual issues; its benefits, hidden bias and the potential impact on the workforce.

But in a recent conversation with someone who is attuned to all things environmental, we realized that we’ve been leaving out a very important issue that’s surfaced over the last couple of years.

But it’s escaped our attention nearly entirely. In fact in an IBM Center for the Business of Government AI roundtable we wrote about, it didn’t come up once. Nor is it mentioned in the National League of Cities report,  “AI in Cities.”

Here’s the deal: AI has the potential to use incredible amounts of electricity, and at a time when global warming is no longer a matter of debate in the serious scientific community, this feels to us like the kind of emerging problem which soon enough will be in the headlines everywhere. So, though this column can hardly be described as a scoop, maybe it’ll be an eye-opener for some.

Though the major threats do not appear to come from the use of AI to generate simple text, when it comes to more complex tasks, like generating images, some researchers have found that creating one image on a powerful AI model can use as much energy as fully charging a cell phone. This is put into perspective in an article in Medium, that found that as of the end of 2024, “AI creates about 34 million images every single day. That’s 1.4 million images an hour. Roughly 23,333 images every single minute.”

And that’s right now. The growth in the use of AI to create images, though unpredictable, is bound to be incredibly fast for as far into the future as anyone can see.

As we started to dig a little deeper into this topic, we were particularly jolted by an article in Harvard Business Review that stated that “the global AI energy demand (is) projected to exponentially increase to at least 10 times the current level and exceed the annual electricity consumption of a small country like Belgium by 2026. In the United States, the rapidly growing AI demand is poised to drive data center energy consumption to about 6% of the nation’s total electricity usage in 2026, adding further pressure on grid infrastructures and highlighting the urgent need for sustainable solutions to support continued AI advancement.”

This isn’t just theoretical. In early April,  a Reuters survey of 13 major U.S. electric utility earnings transcripts “found nearly half have received inquiries from data center companies for volumes of power that would exceed their peak demand or existing generation capacity - that's everything they supply to homes and businesses – a metric that reflects the sheer size of oncoming data center needs.”

To be fair, some futurists believe that Artificial Intelligence, itself, will help to figure out ways to save the energy it consumes. A November 2023 report by the California Government Operations Agency indicated that “Incorporating GenAI in government can drive environmental sustainability by optimizing resource allocation, maximizing energy efficiency and demand flexibility, and promoting eco-friendly policies. For instance, this technology can enhance operational efficiency, decrease paper usage and waste, and support environmentally conscious governance. Notably, stakeholders also highlighted the need for reducing environmental impacts of GenAI use and ensuring environmental costs are equitably distributed.”

We’ve noticed that many cities, counties, states and the federal government all appear to be focused on seeing who can make the most use of AI as quickly as they can. In fact, there appears to be something of an international race going on between the U.S. and China to see who will dominate in AI.

We understand that the potential of AI is so breathtakingly large that we don’t expect environmental concerns to slow it down.

But like other genies that can’t be put back in the bottle, we’d argue that governments everywhere take this into account as they find new and better uses for a technology with such enormous potential. “In the 2024 legislative session, at least 45 states, Puerto Rico, the Virgin Islands and Washington, D.C., introduced AI bills.” according to the National Conference of State Legislatures. We thumbed through the NCSL’s list of these bills and couldn’t find one that mentioned environmental dangers. That has to change.

#StateandLocalGovernmentGenerativeAIPolicyandManagement #StateArtificialIntelligenceManagement #AIandGlobalWarning #StateEnergyManagement #StateandLocalGovernmentAIManagementandPolicy  #StateandLocalGovernmentArtificialIntelligenceManagement #AIandEnvironmentalConcerns #StateGovernmentEnergyUseAndArtificialIntelligence #CityGovernmentArtificialIntelligenceEnergyUse #CityGovernmentEnergyManagement #NLCAIinCitiesReport  #CityGovernmentEnergyPerformanceManagement #ArtificialIntelligenceEnergyUse #StateandLocalEnergyUseManagementAndAI #StateGovernmentEnergyPerformanceManagement #CountyArtificialIntelligenceEnergyUse  #NCSLArtificialIntelligenceBills #CaliforniaGovernmentOperationsCenter#StateGovernmentGenAIManagement #CountyGovernmentEnvironmentManagementandPolicy #StateGovernmentGenAIUse #StateGovernmentEnvirnmentManagementandPolicy #NationalLeagueOfCities #IBMCenterBusinessOfGovernment #BandGReport #BarrettandGreeneInc