Weiquan Lin/Moment via Getty Images
Follow ZDNET:
Add us as a preferred source on Google.
ZDNET’s Key Takeaways
Artificial intelligence agents will become the driving force that enables our digital systems and performs the professional tasks we rely upon every day. Traditional software applications, as we currently understand them, are predicted to be absorbed and eventually replaced by these autonomous intelligent agents. In this evolving landscape, business leaders must confront the challenges of integration, governance, and the responsible deployment of AI-driven strategies.
If you believe that artificial intelligence already exerts a profound influence on the modern workplace, it may be time to reconsider that assumption. According to Steve Lucas, the CEO of Boomi, the transformation brought about by AI is only in its early stages. He foresees that many of the applications we now use to manage daily workflows will ultimately vanish within the next few years. In their place, intelligent AI-based entities—called AI agents—will not only assist humans in accomplishing their duties but will also supply and activate the digital tools required to complete them.
Lucas shared this vision with ZDNET during an in-depth conversation at Boomi’s recent World Tour event in London, emphasizing that the boundary between software categories will blur beyond recognition. What we currently categorize as distinct systems or specialized applications will soon be entirely subsumed and function as elements within an overarching AI framework.
He envisioned a near-future scenario in which the familiar landscape of applications—such as spreadsheets, CRMs, and other tools—will survive purely as logical constructs within a comprehensive AI environment. Workers will interact with this ecosystem through what Lucas calls an “AI experience layer,” a dynamic interface permitting communication that is conversational, visual, and auditory, seamlessly bridging human input with complex systems of record.
The question naturally arises: how will enterprises and employees transition from today’s familiar, application-centric workflows to a workplace fundamentally governed by AI agents? Lucas provided a structured perspective by describing three transformative dynamics through which AI agents will redefine the nature of work.
1. Billions of Agents Will Exist
Lucas compared the evolution of enterprise technology to the development of autonomous vehicles. What began as an audacious concept—vehicles that could drive themselves—has gradually transitioned into a functional reality. He likened early experiments with Tesla’s self-driving capabilities, which once resembled a more advanced version of cruise control, to the initial stages of enterprise automation. Over time, those systems gained the ability to interpret traffic lights, road signs, and environmental cues—eventually culminating in fully self-driving systems like Waymo’s vehicles in San Francisco. In a similar vein, Lucas predicted that corporations will become increasingly “self-driving” in their operations.
At the core of this transformation, he argued, lies the AI agent. These entities, he suggested, are the “magic ingredient” that will enable organizations to progress toward fully automated enterprises. Earlier in May 2024, Lucas had estimated that thousands of intelligent agents could be operating within a single business in just two years. Eighteen months later, his vision has expanded dramatically: he now foresees billions, even trillions, of such agents functioning autonomously across industries.
Lucas illustrated this concept by referencing the task of maintaining data quality—a fundamental yet tedious operational requirement. “In half a minute,” he explained, “I can configure an AI agent capable of verifying addresses, performing postal code lookups, and cleansing that information before it reaches Boomi’s system.” While he acknowledged that real-world data quality management is far more complex, the example served to demonstrate a critical point: if AI can produce such an agent instantaneously, then conventional data quality tools might soon be unnecessary, as their functionality would be integrated directly into broader AI-driven platforms.
2. Software Will Be Consumed by AI
Building upon this paradigm shift, Lucas offered what he described as a “bonus insight”: much of today’s advanced software will cease to exist as independent products within a few short years. Instead, applications will survive as embedded logical entities within expansive AI systems. The implication, he argued, is that software as we know it—discrete, purpose-built, user-operated—will give way to contextual capabilities embedded within intelligent frameworks.
Just as AI agents can ensure the integrity of data, they will likewise assume responsibilities across a wide range of enterprise functions. This development will have dramatic consequences for both the software industry and the professionals who rely upon it. Reflecting on the sweeping transformation catalyzed by the launch of ChatGPT only three years prior, Lucas observed that today’s rapidly evolving digital environment resembles an alternate reality where science fiction has transformed into demonstrable fact. “If AI possesses sufficient power,” he posited, “why shouldn’t it be able to function as a small business’s CRM system upon request?”
He was quick to acknowledge that this shift would introduce complexities—AI-driven operations will encounter nuances and edge cases just as traditional systems do—but the foundational idea remains immutable: this transformation is both inevitable and imminent. Lucas further described a forthcoming workplace experience mirroring the conversational ease of interacting with ChatGPT on a phone. “Soon, I’ll simply converse with an AI agent within the enterprise,” he said. “I might request approval of expense reports—or the agent may proactively inform me that the task has already been completed. It will have verified reports, cross-checked related calendars, and confirmed all necessary details—seamlessly and instantaneously.”
3. System Logins Will Disappear
As AI agents grow capable of managing and automating complex workflows, the need for employees to manually log into traditional enterprise systems will gradually fade. Lucas predicted that, within two years, he personally would never again need to log into a system in the conventional sense. Instead, an intelligent, multimodal experience layer—capable of receiving voice, image, and text input—will mediate all interaction.
Key systems such as Salesforce, Workday, and SAP will continue to perform crucial back-end functions, maintaining their roles as systems of record. However, they will become invisible to end users, much like mainframes quietly powering much of the digital infrastructure today. Users will engage solely through the AI layer, with the underlying platforms operating silently in the background.
For such an ecosystem to function flawlessly, Lucas explained, enterprises will require an activation layer—an intermediary that governs the interaction between conversational AI interfaces and foundational systems of record. This activation layer will not only facilitate integration but also ensure governance, compliance, and coordination across multifaceted environments.
Lucas emphasized that this mechanism is essential, particularly as research conducted by MIT has indicated that 95% of organizations implementing AI initiatives fail to achieve measurable gains in revenue or growth. The activation layer, therefore, will function as the structural “glue” connecting data, applications, and workflows to the emerging agentic ecosystem. Through it, businesses can effectively navigate the practical challenges of correlating new AI capabilities with legacy systems and existing operational models.
He concluded with a pragmatic vision: companies that succeed in this new era will rely on their own preferred AI models, supported by agent development frameworks—whether through Boomi, Amazon’s Bedrock, or ServiceNow’s tools. Yet the connecting foundation must be an AI activation layer, the essential bridge between intelligent dialogue and functional execution. “It’s coming,” Lucas asserted. “I can see it 1,000 miles away.”
Sourse: https://www.zdnet.com/article/how-ai-will-eat-your-software-3-ways-agents-are-reshaping-tomorrows-workplace/
