Follow ZDNET: Add us as a preferred source<!–> on Google.
ZDNET’s key takeaways
- Agents will enable our systems and complete our tasks.
- AI will consume traditional software applications.
- Business leaders must address integration and governance.
If you think AI already has a big impact on the workplace, think again. If Boomi CEO Steve Lucas’s prediction is correct, the applications we use to complete our work will disappear in the next few years, as AI agents not only help us to complete tasks but also enable the tools that we need for our work.
“In the not-too-distant future, things that we believe are distinct software categories will be consumed by AI and go away,” Lucas said to ZDNET in a one-to-one conversation at the technology specialist’s recent Boomi World Tour event in London.
Also: 5 ways to prevent your AI strategy from going bust
He envisaged a situation where the applications we know and use today will soon exist only as a logical construct within AI. Employees will use what Lucas calls an AI “experience layer” to communicate verbally, visually, and audibly with traditional systems of record.
So, how will we shift from our traditional way of working to an AI-enabled enterprise during the next few years? Lucas outlined three ways agents will shape the workplace of the future.
1. Billions of agents will exist
Lucas suggested enterprises will evolve like self-driving cars. What once seemed a radical, far-off vision will manifest as an increasingly automated and standardized reality.
“When you first got in a Tesla, and you experimented with self-driving, it was nothing more than cruise control, just slightly more advanced – stop and go in traffic,” he said.
“But self-driving technology has evolved. Over time, the tech recognized stop lights, stop signs, and could bring you to a stop. You get into a Waymo in San Francisco, and the level of self-driving achieved is astounding. So, too, enterprises will become more self-driving.”
Lucas said AI agents are the magic ingredient that will allow companies to transition to automated enterprises.
Also: Is your company spending big on new tech? Here are 5 ways to prove it’s paying off
He previously suggested in May 2024 that thousands of agents would be active inside businesses within two years. Lucas is even more assertive 18 months later: “There will be billions, perhaps trillions, of these things.”
He gave an example of how AI agents might help to ensure data quality, an area of work that’s often enabled by his company’s technology.
“In 30 seconds, I can create an AI agent that qualifies addresses, does address look up with postal codes, and cleans those addresses before they ever get into Boomi,” he said.
“Now, I realize that data quality is more sophisticated than that example. But I submit that use case as a humble example of, ‘If I can use AI to create an agent in 30 seconds, do I need data quality tooling, or do I already have it as part of a broader platform?'”
2. Software will be consumed by AI
Lucas then offered what he referred to as a bonus comment: “I think there is much technology today, and it might not be two years, but it might be three, that will only exist as a logical construct within AI.”
Just as AI agents can help professionals manage data quality concerns, they can also take on other responsibilities. Lucas said the impact of this shift on the software industry and the professionals who fulfill tasks using these tools will be huge.
Also: 5 ways to feed your AI the right business data – and get gold dust, not garbage back
Lucas reflected on the rapid pace of change that has already occurred since the launch of ChatGPT three years ago, suggesting that the AI-enabled transformation we’ve already seen in the workplace can help us understand what might appear to be a radical proposition.
“We’re living in an alternate universe where science fiction is now science fact,” he said. “Think about it for a minute: If AI is sufficiently powerful, which it will be, and I said to AI, ‘I’m a small business, act as my CRM.’ Is it really a stretch for it to be capable of that?”
As with the earlier data quality example, Lucas recognized there would be complexity and nuances to this shift. However, the premise remains the same: Change is coming, quickly.
“The consumer experience I have on my phone, which is that I just talk to ChatGPT, I will have that conversation in the enterprise with an agent, and I’ll say, ‘Look, if I’ve got any expense reports, just approve them,'” he said.
“And I may not even have to say that. The agent might just tell me, ‘I already took care of that work. I approved your expense reports. I checked Bob’s calendar, and he was with those three customers in his report. So, you’re all good.'”
3. System logins will disappear
As agents automate the applications and activities associated with the modern workplace, Lucas said the need to interact with traditional systems, the mainstays of the enterprise IT ecosystem, will begin to dissipate.
“The reality is that two years from now, I will never log into another system,” he said. “AI will just be the experience layer that we have, verbally, visually, and audibly. That’s what’s going to happen.”
Big-ticket systems, such as Salesforce, Workday, and SAP, will become systems of record, almost like the equivalents of mainframes in the modern era: “They are there in the background, very important, doing things, but we will never interact with them.”
Lucas suggested that this high level of automated experience will be dependent on an activation layer between the conversational interface and the systems of record.
Also: AI will cause ‘jobs chaos’ within the next few years, says Gartner – what that means
He said this layer will deal with integration and governance concerns, which have become more prominent as MIT research has demonstrated that 95% of enterprises attempting to harness AI aren’t seeing measurable results in revenue or growth.
The activation layer will help digital and business leaders overcome what Lucas referred to as the challenges to delivering returns from AI projects, integrating with existing applications and data, and tying emerging technologies into existing workflows and processes.
“Companies that want to succeed with AI will have their model or models of choice. They’ll have agentic development platforms, whether they use Boomi, Bedrock from Amazon, or something from ServiceNow,” he said.
“However, they must have the data, application, and workflow glue, and that will be known as an AI activation layer. It’s coming. I see it 1,000 miles away.”
