AI, Large Action Models and the future of the digital workplace
For decades, the digital workplace (DWP) has been primarily viewed as a tool for facilitating business operations. While innovations in unified collaboration, communication, service desks, and other areas have been substantial, they’ve largely served as enablers rather than transformative forces. This perspective is shifting dramatically, as emerging advancements such as Large Action Models (LAMs) have the potential to embed AI innovations at the very core of business operations, fundamentally transforming how businesses function.
What is a Large Action Model (LAM)?
Until now, Gen AI has functioned primarily as an intelligent assistant or, at best, a teacher, handling routine tasks such as note-taking, drafting emails, or summarizing documentation. While it has been effective in freeing up time for more strategic or creative endeavors, it has stopped short of actually performing the tasks.
Enter LAMs. A Large Action Model is more than just an assistant—it’s an autonomous agent capable of executing tasks. Just as Large Language Models (LLMs) sift through vast amounts of textual data to interpret meaning and intent, LAMs are designed to analyze, understand, and replicate actions. Rather than merely providing a list of steps needed to complete a task, LAMs can directly interact with digital or physical systems to carry out tasks to completion based on the user’s instructions.
Potential use cases for LAMs
In the near future, we might see a virtual travel agent powered by Gen AI. Picture this: you provide an AI agent with your destination preferences, travel dates, and other parameters. The AI then generates a shortlist of options, detailing the pros and cons of each, including accommodations, amenities, and activities. You could even use a VR headset for a virtual tour. For example, if you’re British, you might gravitate towards an all-inclusive resort to minimize decision-making.
Once you’ve chosen your destination, a Large Action Model (LAM) takes over. Beyond finding the best flight deals, it can book tickets and make reservations for you, using your stored credit card details and profile information. It considers your budget, schedule, room preferences, dietary needs, and other factors.
In procurement, a LAM could draft a tender based on your business needs, identify suppliers, issue an RFP, and assess bids. It might also automate contract reviews, negotiate prices, and submit purchase orders.
Looking further ahead, LAMs could handle physical tasks, such as managing factory operations to adjust production in real time. In healthcare, they could automate routine tasks like preparing documentation, monitoring patients, and managing care plans, allowing clinicians to focus on more complex aspects of care.
The key limitations will be the quality of model training, the safeguards in place, and the level of trust we place in these systems.