Artificial intelligence is rapidly transforming marketing and the customer experience, as well as industry and operational processes overall. In fact, 64% of marketers say AI will be critically important to their marketing success over the next 12 months, according to the Marketing Artificial Intelligence Institute, with 98% of marketers saying they are already personally using AI in some way.
All this is creating a massive amount of unimagined new capabilities for us, both personally and professionally. Now that we are at a point where we can achieve much of what AI promises, the question is—at what cost?
That’s a question that wasn’t on the radar some 70 years ago, when Alan Turing first explored the mathematical possibility of artificial intelligence. Its practicality had to wait for the advanced computer capabilities that followed. Now, because of the cloud and access to an immense amount of computing power, we're seeing AI come to life based on these old theories.
But as a result, many unresolved challenges facing AI usage are just becoming apparent. Two of the most important facing marketers seeking to capitalize on AI are its immense appetite for energy and the resultant impact on the environment.
Failing to address both may throttle AI-driven campaigns, pending technological breakthroughs not yet on the horizon.
AI everywhere, and for everyone
We’re fast approaching a time where AI is everywhere. We already are, or will soon be, using AI on our phones, computers, thermostats, appliances, watches and more, and the potential energy consumption of this technology is massive. As the use of AI continues to grow, enterprise AI leader Nvidia is set to ship 1.5 million AI server units per year by 2027. Running at full capacity, we’re talking at least 85.4 terawatt-hours of extra electricity consumption a year, even in the face pressures to reduce our carbon footprint.
Then there’s the issue of water withdrawn or consumed for electricity generation and cooling the servers. The University of California, Riverside, and the University of Texas, Arlington, report that the global AI demand may be accountable for 4.2 billion to 6.6 billion cubic meters of water withdrawal in 2027, more than the total annual water withdrawal of Denmark or half of the United Kingdom. All this diverts resources needed for a rapidly growing population and aging water infrastructures
Sucking up their own share of power are AI-powered advertising campaigns. Marketers are eyeing the possibility of campaigns that personalize ads and digital experiences for millions of people. Today, we have decent metrics on conversions and cost per click, but energy increasingly will have to be accounted for in business cases and media plans. If your return on ad spend and CPMs are warped because of AI’s enormous appetite for energy, and your budget is challenged by reaching that last cohort of a million more people, a full-blown AI campaign could become cost-prohibitive, or impact the sustainability numbers you’re reporting to Wall Street.
It appears that regulators are beginning to notice. Pending legislation from the House AI Caucus would require standards to measure and report the full range of AI’s environmental impacts, and to create a framework for AI developers to report on those impacts.
Just like legislation aimed to protect consumer privacy, this is another reason for marketers to pay attention. Many departments, including marketing, already must report on their overall impact on the brand’s net-zero goals. Now consider the environmental impact of media strategies and buys using generative AI to personalize ad units for potentially millions of individuals.
Theory to reality, and beyond
But all is not doom and gloom. The theory-to-practicability road is evident in a wide variety of other fields. In genomics, where the first gene sequencing was achieved 50 years ago, we’re now seeing stupendous real-world breakthroughs in health care. The practical uses of quantum computing are coming into focus as a potential means of solving complex problems faster and more economically. Meanwhile, scientists are gradually expanding the capacity for fusion energy production that promises to create an emissions-free source of power that does not rely on fossil fuels.
There’s no reason why future innovations in AI generation and use cannot address its current challenges of excessive energy consumption and a massive carbon footprint.
Marketers aware of all this can enjoy a potent differentiator. For example, they might ask AI not only to launch personalized campaigns but also to control for environmental impact as part of overall ad spend. Perhaps energy usage per click, per usage or per conversion can be added to current common metrics, all signaling a brand’s demonstrable commitment to energy savings.
Say you need to reach this many consumers with a conversion rate of X, but also reduce impact on the environment by Y. AI might then provide an overall energy spend threshold for the campaign, similar to overall equations of reach, conversion, cost per click, cost per conversion and the like—all while ensuring your ads reach the right people at the right time with a highly personalized experience.
Developing cost-effective AI-driven campaigns, still hungry for energy and with sustainability challenges, will remain important factors as AI becomes ubiquitous. This is a journey that companies like Avanade can help navigate.
Avanade has spent the past two decades working with more than 5,000 clients across the world. Right now, the most successful organizations are adopting technology in bold new ways to reinvent, reimagine and redefine themselves for the future. Digitization and a focus on AI and hyper-connectivity continue to create new opportunities for growth, allowing organizations to experiment through physical, digital and immersive technologies and connected devices.
There’s no doubt that AI will drive future meaningful evolutions in many fields, and businesses can do so in a responsible manner by partnering with experts. When—not if—we get to a point where there is readily accessible, reliable, portable, sustainable power, it’s worth considering the immense possibilities that artificial intelligence will unlock—for the general good, well beyond the next AI campaign.
Join Joe Kehoe and Jeremy Kolpak at SXSW in Austin, Texas, on March 8, 4:00 p.m.-5:00 p.m. CST, where they will be discussing “Beyond the Horizon: Unleashing the Power of Future AI.”