In software at 8×8, employees operate Anthropic’s Claude software to compose emails, analyze customer feedback and write code, but so far the growing reliance on an AI-powered chatbot isn’t a problem for the finance team. While other Silicon Valley companies such as Meta, Uber and Salesforce have publicly expressed concerns about the rising costs of generative AI tools and in some cases have begun to implement usage restrictions, 8×8 says it is in the red.
The company estimates it has saved about $5 million in annual costs over the past 18 months by canceling subscriptions to dozens of software and educational tools it deemed unnecessary, in part because Claude could provide similar capabilities. So far, Claude’s annual 8×8 bill is “well below” that amount, says Joel Neeb, the company’s director of transformation and business operations.
Neeb expects the savings and costs will eventually even out as the 8×8 solution encourages more workers to operate AI and incorporates the technology into more convoluted work. For now, though, there’s still a huge gap, which “makes my CFO happy,” he tells WIRED. He declined to provide exact total spending on generative AI.
As companies collectively spend hundreds of millions of dollars on AI tools for coding, marketing and customer service, a modern obsession has emerged in the tech industry: “tokenomics,” or managing the rising costs of using artificial intelligence. (Tokens represent the amount of content analyzed and generated by the AI model.)
Last month, the CEO of Royal Bank of Canada revealed that token usage had increased by 500 per cent in the last six months. At Cisco, a third of employees operate an internal AI chatbot every day, so “token usage is getting crazy,” CEO Chuck Robbins said on the earnings call. According to the company’s CEO, Spenser Skates, some top engineers at analytics software company Amplitude “spend thousands of dollars a month or more on tokens.” Aaron Levine, CEO of Box, said, “The conversation about token budgeting has definitely become one of the most important,” and “animatedtopics.
About 300 companies responded to questions or concerns about AI tokens during earnings calls or public discussions with financial analysts in April or May, according to a review of WIRED transcripts by data provider AlphaStreet. That’s a small fraction of the thousands of connections made during that period, but just 93 companies mentioned the “token” in April and May a year ago.
Executives at several companies have said they are developing or looking to purchase systems to help monitor token usage and select the cheapest model for a given prompt. Others said they were still trying to strike a balance between hiring more people and increasing token budgets to achieve their goals.
Software is rarely cheap, but the latest generation of AI tools are putting extraordinary strain on C-suites for a variety of reasons. Prices are constantly changing. New models, which are more powerful and more expensive than the previous ones, are released every month. Engaging entire organizations in new ways of working has been a challenge, so AI-powered productivity gains in one team can lead to bottlenecks in another.
20 percent
That said, some companies are still encouraging employees to use AI more often without having to worry about the card. In April, Long Island, New York-based sportswear brand Baseball Lifestyle 101, which expects to generate $250 million in sales this year, told about 50 of its top executives to spend the equivalent of about 20 percent of their salary on AI tokens each month.
Bill Rom, co-founder and chief strategy officer of Baseball Lifestyle 101, tells WIRED that costs will likely top $100,000 a month by the end of the year, but it’s already paying off. Claude recently helped secure a $1 million order by finding that the retailer was running out of stock of certain sizes of its popular ice cream shorts. “That’s a day and a half of work that can now be done in an hour or two, and could earn me an eight-figure extra income within 12 months,” Rom says.