AI Transforming the Workplace: From Repetitive Tasks to Multiplayer AI
AI has rapidly transformed the workplace, taking over the repetitive, time-consuming work that once ate into people’s days. But according to Gabriel Hubert, CEO and co-founder of AI company Dust, the workplace is moving to another level, with a significant shift toward multiplayer AI.
The Move to Multiplayer AI
Instead of individual employees using isolated tools, multiplayer AI involves agents being shared and used collaboratively across different departments. These agents learn from a company’s data, allowing teams to automate processes and share workflows.
In an interview with Sifted, Hubert explains how businesses can move from single-player AI—where value and productivity are created by the individual—to multiplayer AI used collaboratively across teams.
The typical way AI has been used in many businesses is this: an employee opens a chatbot, sticks it in the context they need for a task, sends a prompt, gets a response, and uses that information to complete a task or move on. While this can increase individual productivity, it has its limits, according to Hubert. “One person learns a better way of doing something, but that improvement doesn’t necessarily propagate,” he says. “The individual moves faster, while the company continues to work in much the same way.”
In companies operating with a multiplayer AI model, agents can “start participating in the same workflows as other people and other agents,” he adds. “They can farm out work, reuse what another team has learned, and contribute to a shared system rather than starting from scratch each time.”
Multiplayer AI agents essentially act as “digital teammates”: users can “@” mention a particular agent in a collaborative workspace, ask it to perform a task, then assign it to another agent. The important thing is that the workflow is shared, improves over time, and becomes available to the rest of the team.
For example, a “blog writer” agent can generate a blog post, then forward it to a “LinkedIn” agent to write social copy based on the shared context. Another area where Hubert has seen multiplayer AI implemented effectively is in sales.
“Traditionally, a salesperson spends 30 minutes researching information, deciding whether a prospect is relevant, and updating a CRM,” he explains. “Different representatives will inevitably do things differently.” In a multiplayer workflow, the sales rep can ‘@’ mention an agent who gathers the relevant data, applies the company’s qualification criteria, updates the CRM, and routes the lead. The important thing is that the workflow is shared, improves over time, and is available to the rest of the team.
As agents evolve, they “need to be able to do things, not just retrieve and summarize information,” Hubert adds.
Governance, Security, and Scaling
As organizations move from single-player AI to multiplayer AI, traditional IT governance models are not always fit for purpose. Across the EU, 55% of large companies use AI, but only a quarter of these companies believe their governance models are fully equipped to handle implementation, according to data from Smarsh, a data and intelligence platform.
This can lead to “shadow AI” within an organization: the unauthorized use of AI tools or features by employees, in a way that is not governed by a central IT team. For example, if an agent has access to a company’s Google Drive, any user interacting with that agent may accidentally come across confidential records.
An agent inherits access to data from the space it was built on and this access remains the same no matter who uses it. At Dust, administrators manage access control for AI agents, says Hubert. These administrators decide which data sources are connected and how this information is organized.
“Some spaces are open to the entire company while others are reserved for specific teams or individuals. Collaborators can create and use agents based on the spaces they have access to.” An agent inherits access to the data of the space it was built on and that access remains the same no matter who uses it. An administrator can allow the entire company to use an agent even when most employees cannot directly see the underlying data.
At Dust, if a user does not have access to specific data, the platform prevents the agent from providing this data to them. “A company’s operational context and knowledge should belong to the company. It should not be locked into a template vendor, a one-to-one conversation, or a series of disconnected applications,” adds Hubert.
Keep a Human in the Loop
One of the biggest barriers to widespread use of AI in the workplace is that individual employees are often ready before their organization. A study conducted by Microsoft shows that organizational factors such as manager support, talent practices, and culture generate more than twice the AI impact of individual employee efforts alone.
At Dust, the generalization of multiplayer AI relies on “AI operators.” This person is “the most important new role in the company,” Hubert says. “What they share is a different way of looking at work. Instead of asking how AI can help with a task, they ask whether the process should still exist in its current form now that AI is available.”
As agents absorb more execution, judgment becomes more precious. “The best operators make a ‘to-do list,’ asking after each boring task, ‘How do I never have to do this again?’”
By 2027, Hubert hopes that organizations will no longer be wondering whether to use agents, but rather how to manage their workforce. “Once a company has many agents participating in workflows, it needs to understand what they are doing, who is responsible for them, and whether their decisions remain reliable,” he explains.
He also suspects that a larger challenge will remain, one that will require human input. “As agents absorb more execution, judgment becomes more valuable. The agents themselves are not the entire asset. The asset is the loop between the agents, the context the company has, and the people who continue to improve both.”
For more detailed insights, visit the original article here.
“`
