Companies pour cash into AI despite climate risk; Lululemon sales slow

Companies, VCs pour money into AI as environmental, energy costs soar

What you probably already know: The big tech companies are pouring money into artificial intelligence in a race to capture the future opportunities many believe generative AI will bring. The risk of underinvestment, Google/Alphabet’s Sundar Pinchai said, is far greater than the risk of overinvesting. Google, Meta, Microsoft and Amazon invested a staggering combined $52.9 billion in AI in the last quarter, the most ever, and continue to up their spend.

Why? Even though there’s little enterprise adoption of genAI so far and many of the early startups aren’t yet profitable, the race is on and no one wants to get left behind. Venture capitalists have pumped $64.1 billion into AI companies so far this year, nearly a third of all VC investments for the year, according to PitchBook. That’s drawing scrutiny from environmental groups, which warn the increased use of AI could have disastrous consequences for climate change. Goldman Sachs produced a report recently that showed that, by 2030, there will be a 160% increase in demand for power for AI apps.

What it means: Demand for data centers is growing as companies fall over themselves to develop practical uses for large language models and other AI applications. Microsoft has doubled the number of data centers since early 2020, and Google has increased its data centers by 80%. States are starting to push back as demand for electricity to power the data centers outstrips supply in some areas. Demand for power for data centers is now nine times higher than it was in 2015.

What happens now? There are some who suggest that AI might actually help solve its own environmental issues by more quickly identifying efficient ways it can do the work it needs to do, while also being deployed to help solve other environmental challenges. For example, AI is already being used to help detect crop diseases before they’re visible to farmers and prevent crop loss, while simultaneously reducing the amount of pesticides necessary for growing common crops. AI is also reliant on the humans who provide the data the models use to learn — provide too large of a data set, and it will take the system more effort to learn, and thus use more energy in the process. Smart inputs can help reduce the environmental impact.