Fast company logo
|
advertisement

Yesterday Slack released a new survey of more than 10,000 desk workers and Salesforce (which owns Slack) announced a new AI-powered copilot.

Desk workers increased their AI usage by 24% last quarter, according to Slack

[Source Photo: cottonbro studio/Pexels]

BY AJ Hess3 minute read

Artificial intelligence evangelists have predicted that automation would radically change the way we work for years. While automation has steadily transformed blue-collar industries like manufacturing and agriculture over the course of the past several decades, the technology has only gained quantifiable traction among desk workers more recently. 

Yesterday Slack released a new survey of more than 10,000 desk workers around the world indicating that the use of AI in the workplace accelerated 24% in the last quarter. Slack’s Workforce Lab found that in September 2023, one in five desk workers reported they have tried AI tools for work. By January 2024, that share had risen to one in four desk workers.

Christina Janzer, Slack’s senior vice president of research and analytics at and head of Slack’s Workforce Lab, says the growth of the use of AI last quarter was more than she expected. 

“In 2023, we talked a lot about the promise of AI, but we were seeing low adoption rates throughout 2023,” she says. “We didn’t see any uptick last year, so the fact that we went up from 20% to 25% is a big deal in my mind. But it’s also a reminder that we’re still very early in this AI journey. We’re still figuring it out. We still have a lot of work to do.” 

On the same day that Slack released this research, Salesforce (which owns Slack) announced that it would make a beta version of Einstein Copilot, a new customizable generative AI-powered conversational assistant, publicly available. Salesforce already has Generative Pre-trained Transformer (GPTs) for different end users. For instance, Service GPT is designed for customer service representatives, Sales GPT is for salespeople, and Commerce GPT is for commerce managers. The new tool would allow any worker to draw on all of these tailored GPTs.

While many organizations have taken steps to ensure that workers do not input sensitive company data into large language models like ChatGPT or Google Bard, Einstein Copilot would allow Salesforce users to internally ask questions, request summaries, and create new content based on a company’s own private data.

As the CEO of Salesforce AI, Clara Shih is responsible for the company’s AI product strategy, which after years of investment, she characterizes as now being “on fire.” 

“Salesforce AI Research was founded almost a decade ago and we developed some of the earliest large language models—Google was first and then was followed by Salesforce AI research,” says Shih. “We’ve also been a leader in predictive AI. We introduced the first predictive AI for CRM in 2016.”

The first people to use Einstein Copilot were members of Salesforce’s own workforce. Both Janzer and Shih said this internal experimentation is part of the company’s broader culture and product strategy.  

“A big part of our learning and experimentation culture is we have our own teams use our products first,” says Shih. “Salesforce’s customer service team uses Service GPT, Salesforce’s sales team uses Sales GPT, and they’ve also been the first to trial our Einstein Copilot.” 

Janzer also described an instance in which one of her researchers conducted an experiment testing herself against different AI tools to prove that much of her work is still best done by a human. 

Indeed, while many workers are anxious that AI will replace some, or all, aspects of their jobs, Shih likens the way AI will impact workers to the way the invention of the internet impacted workers through the 1980s and 1990s. “There are so many lessons we can learn from what happened in the internet era. The internet replaced some jobs; it created far more jobs, but it changed every single job,” says Shih. “And what people spent time on before the internet—printing papers, faxing, documents, waiting for faxes, filing papers away in a filing cabinet—that work just went away. And instead, everybody had to learn how to type. Everyone had to learn Google search and email.”

She continues, “If you don’t know Google search and email, it probably would be hard to find a job these days. AI is doing the same thing to work.”

Recognize your brand’s excellence by applying to this year’s Brands That Matter Awards before the final deadline, June 7.

Sign up for Brands That Matter notifications here.

WorkSmarter Newsletter logo
Work Smarter, not harder. Get our editors' tips and stories delivered weekly.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Privacy Policy

ABOUT THE AUTHOR

AJ Hess is a staff editor for Fast Company’s Work Life section. AJ previously covered work and education for CNBC. More


Explore Topics