Salesforce today published an AI Fluency Playbook, a guide the provider of software-as-a-service (SaaS) applications hopes will be used to accelerate adoption of artificial intelligence (AI) agents.
The overall goal is to enable organizations to more confidently assign tasks to AI agents, says Ruth Hickin, vice president of workforce innovation for Salesforce. In addition to concerns about job losses, many workers still simply lack the skills needed to work confidently with AI, she adds.
“Engagement is the foundation for AI fluency,” says Hickin. “It’s really critical businesses overcome the initial barrier of fear and confidence.”
The AI Fluency Playbook is based on how Salesforce has been able to employ AI agents to automate its own internal workflows. The company claims 85% of its employees now feel confident using AI tools to drive productivity in their daily work, representing a 16% increase year over year.
The Agentforce in Slack tool, for example, is credited with saving Salesforce employees over 500,000 hours, while an Engagement Agent helped manage 190,000 sales leads, the company claims. A Service Agent handled more than two million support requests for the customer service team.
After initially touting workforce reductions, Salesforce CEO Marc Benioff has since admitted the company was over confident in terms of the degree to which AI would automate internal processes. Nevertheless, the Salesforce CEO remains a staunch advocate of digital business transformation enabled by AI.
In the meantime, most organizations continue to experiment with AI. While there are clear opportunities to reduce the amount of time required to perform specific tasks, the frameworks needed to orchestrate and govern AI agents have not yet been widely adopted.
In general, there needs to be more focus on the people and processes to realize the potential of an incredible technology, says Dan O’Brien, president and COO of the Futurum Group.
“The early innings of the AI revolution have focused so much on the technology, but where we see companies really having the most success with AI is in rethinking their processes for the AI-era,” says O’Brien. “It needs to be more about redesigning around AI than integrating AI into existing processes.”
It’s not clear to what degree organizations are proactively training employees to take advantage of AI technologies. In some cases, organizations have made available formal training programs, while in other instances business leaders are mostly hoping that the rank and file will discover how best to, for example, make use of AI agents on their own. The challenge with the latter approach is that it often results in a lot of trial and error, with successes often difficult to replicate across the organization as a set of best practices.
Ultimately, it’s not a question of whether AI will be adopted so much as it is now to what degree. There are still numerous reliability issues that organizations will need to address before a set of probabilistic generative AI technologies can be relied on within the context of what are often deterministic processes that need to be completed the same way every time. Most organizations have also yet to fully come to terms with the total cost of using generative AI tools at scale.
In the meantime, however, organizations would be well advised to develop a strategy now versus hoping that somehow a thousand flowers might one day magically bloom in a way that actually benefits the business.
