More companies are adopting ‘AI-first’ strategies. Here’s how it could impact the environment

As artificial intelligence gets smarter, a growing number of companies are increasing its implementation in their operations or more heavily promoting their own AI offerings. The buzzword for this is “AI first.”

Duolingo is among the latest to adopt an AI-first approach. The company’s CEO, Luis von Ahn, announced the change in an all-hands email Monday, saying it would stop using contractors to do work AI can handle and only increase head count when teams have maximized all possible automation.

“The way we work is fundamentally shifting. AI is becoming the default starting point,” said Duolingo’s Chief Engineering Officer Natalie Glance in an internal Slack message she shared on LinkedIn. “Start with AI for every task. No matter how small, try using an AI tool first. It won’t always be faster or better at first—but that’s how you build skill. Don’t give up if the first result is wrong.” 

Von Ahn, in his email, said the AI-first approach was already paying dividends, helping the company with its content creation process. “Without AI, it would take us decades to scale our content to more learners,” he wrote. 

Earlier this month, Shopify’s CEO Tobi Lütke told workers at that company that using AI was now a “fundamental expectation” in daily tasks. “Our task here at Shopify is to make our software unquestionably the best canvas on which to develop the best businesses of the future,” he wrote. “We do this by keeping everyone cutting-edge and bringing all the best tools to bear. . . . For that we need to be absolutely ahead.”

AI’s rise in business has been forecast for years, of course. But as more companies make it a priority, there are other impacts to be considered.

A scientific paper released by Cornell University late last year titled, “The Unpaid Toll: Quantifying the Public Health Impact of AI,” said the pollution from data centers powering the AI industry could lead to up to 1,300 premature deaths each year by 2030. It further estimated that public health costs related to the air pollution those centers put out are already at $20 billion per year.

Data centers are nothing new. They’ve been around since the 1940s, when the University of Pennsylvania built one to support the first general-purpose digital computer, the ENIAC. But as generative AI has grown, so too has the demand for newer, more powerful centers.

The power requirements of data centers in North America increased from 2,688 megawatts at the end of 2022 to 5,341 megawatts at the end of 2023, according to MIT. And demand is only growing. (Energy Secretary Chris Wright, in February, called for more nuclear power plants to meet the growing demands of AI companies.)

“The demand for new data centers cannot be met in a sustainable way,” said Noman Bashir, a Computing and Climate Impact Fellow at MIT’s Climate and Sustainability Consortium. “The pace at which companies are building new data centers means the bulk of the electricity to power them must come from fossil fuel-based power plants.”

This is all occurring as concerns about the environment have been deemphasized at many Big Tech firms. Companies like Walmart, Siemens, and Apple all opted against signing an open letter earlier this year reaffirming commitment to the Paris Agreement. (Duolingo, which released an environmental statement last March, did not reply to questions about how the AI-first approach might impact the company’s environmental footprint.) 

Meanwhile, the The President administration has dismantled dozens of climate programs in its first 100 days. And the Environmental Protection Agency (EPA) is considering overturning previous findings that classify greenhouse gas pollution as harmful, which could impact its ability to regulate carbon emissions.

By 2030, Cornell forecasts, the public health burden of AI data centers will be double that of the U.S. steelmaking industry. And it could be on par with all of the cars, buses, and trucks in California.

Shopify and Duolingo are hardly the only companies adopting an AI-first approach. Many companies large and small are racing to incorporate AI into all levels of their services and workflows. Financial services firm Lettuce leans into AI to assist with tax solutions. Findigs lets property managers use AI to screen rental applicants. And a real estate brokerage in Portugal is using an AI interactive real estate agent, which has already booked $100 million in sales.

In the grand scheme, though, corporate use of AI is still in its infancy. ServiceNow’s Enterprise AI Maturity Index last year measured AI maturity at 4,500 businesses in 21 countries on a scale of zero to 100. The average score was 44, with only one in six companies topping 50.

Part of what’s keeping that score low is the newness of the technology. Another factor is cost. (Does using AI, especially one that’s developed in house, actually save money given the cost of data centers, for instance?)

But in the coming months and years, more companies are likely to move to an AI-first approach. And that will likely increase emissions, pumping more CO2 and pollution into the atmosphere, raising even more health concerns. 

View the full article