AWS, Salesforce Execs See a ‘Fully AI World’ With ‘Limitless Labor’

Executives from AWS, Salesforce and Oracle believe that artificial intelligence (AI) agents will upend the world of work in ways people today might not expect.

Consider the following potential future scenarios:

• You can talk directly to a software program to get what you need instead of clicking through menus and settings.
• An AI agent will take care of a customer calling into a call center from start to finish.
• Software integration will not be needed anymore because the AI agent will just click through the other software like a human would.

Two years after the ChatGPT moment, “we have entered an even more transformative phase with generative AI,” said Swami Sivasubramanian, vice president of agentic AI at AWS. “The emergence of agentic AI applications really represents a fundamental shift in its capability.”

Unlike earlier generative AI capabilities, AI agents can reason, plan and complete tasks on behalf of humans or systems. They can research, pay bills, manage enterprise applications and even decompose high-level objectives into executable steps, Sivasubramanian said.

Down the road, he envisions a “fully AI world” where AI agents make decisions, automate workflows and coordinate by themselves with less human oversight. “But this evolution is not a roadmap for the future. In fact, most of it is already happening today.”

AI agents are helping biotech company Genentech automate drug research, saving nearly five years of work, according to Sivasubramanian. Credit ratings company Moody’s has used them to shorten its credit risk reporting process from a week to less than an hour.

“This is exactly the kind of breakthrough in productivity and scale you’re going to be able to achieve with customers adopting agentic AI,” Sivasubramanian said at last week’s HumanX conference in Las Vegas.

Read more: Deloitte Begins Rollout of AI Agents for Finance and Other Business Functions

Adam Evans, executive vices president and general manager of Salesforce AI, said the company’s AI agent platform, Agentforce, has seen “overwhelming demand” in the fourth quarter — its first full quarter of selling the service.

More than 5,000 customers have signed up, which is “incredible early demand,” Evans said at a separate discussion at the HumanX conference. As more companies use AI agents, he said this becomes a “limitless labor” pool.

Salesforce itself uses agents. For instance, an AI agent helps customers who have questions about the Salesforce platform, their accounts and similar queries. Evans said the AI agents resolved all but 3% of the cases, which are then escalated to human representatives.

Evans sees the next level for customer service comprising of AI agents that can do more than just answer and refer queries. “Just like a brand ambassador, you’re going to have an AI agent that knows you as a customer, that can help sell, help understand your account, can escalate when necessary, and just kind of be a touchpoint the whole way,” he said.

Evans cited three types of AI agents:

• External agent that interacts with the customers through chat, email and voice.
• Internal agent that interacts with employees through channels like Slack, helping them in their workflows and the like.
• Background agent that works behind the scenes to do things like look at sales leads in the CRM, do research and analyze client communication.

Evans cited examples from early adopters like Wiley Publishing, which reported a 40% increase in customer satisfaction due to AI-driven customer service agents. Pfizer has started using AI agents to assist in life sciences as well. “We’re seeing customers in regulated industries as well” using agents, he said.

Salesforce has structured its business model to accommodate this shift, introducing flexible pricing that allows customers to transition between traditional seat-based pricing and AI consumption-based models.

Read more: OpenAI Launches New Tools for Building AI Agents

Miranda Nash, group vice president at Oracle AI, sees AI agents becoming the primary interaction layer between users and enterprise software, eliminating the need for complex interfaces.

“Let’s stop adapting ourselves to computers and make them adapt to us,” Nash said.

That means users might not have to click through the menus of Oracle software to navigate it. “In the future, we really believe you can just ask Oracle questions, and that leads to, behind the scenes, a rather complex, potentially multi-agent workflow,” Nash said during a presentation at the HumanX conference in Las Vegas.

Today, Oracle embeds AI agents across HR, finance, supply chain, and operations, enabling departments to automate their own processes. For HR, it means AI-driven hiring processes; for finance, it’s AI optimizing accounts payable; for supply chains, it’s automated logistics management. Every role is being enhanced by AI, she said.

Nash has the same view as Evans and Sivasubramanian that AI agents will fundamentally reshape employment. But all three downplayed fears of mass job cuts.

“There’s definitely going to be some displacement and there’s going to be more opportunities,” Evans said.

Evans highlighted Salesforce’s own experience: Employees previously tasked with repetitive customer support roles are now being moved into higher-value customer success and sales positions.

Sivasubramanian said AI agents are enabling companies to streamline workflows, reduce training times and free employees from mundane tasks.

He said AWS’s internal AI deployment saved Amazon itself 4,500 developer years’ worth of work and over $250 million in capital expenses. “I think that’s pretty incredible,” Sivasubramanian said.

Said Nash: “The only option now is to get in the cloud, get to agents, and meet the future of this world of work.”

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

The post AWS, Salesforce Execs See a ‘Fully AI World’ With ‘Limitless Labor’ appeared first on PYMNTS.com.