Moving from AI Assistants to AI Agents

Are we at the start of the next wave of AI’s productivity promise?

Subscribing to the notion that production is defined by labor and capital inputs, I posit that new technologies are often embraced by organizations because of significant productivity gains that could be unlocked: 

Source: Morgan Stanley, “Tech Diffusion: 10 Lessons from 100 Years” (6/1/23)

In today’s age of AI, it’s no surprise to see AI’s productivity promise as a core source of excitement for enterprise adoption, as such an uplift could drive meaningful margin expansion. Such productivity gains are often quoted as a tangible source of enterprise ROI in many AI startup pitches. In fact, McKinsey purports that the value of AI’s impact on labor productivity represents a bigger opportunity than revenue-generating use cases:

Source: McKinsey & Company, “The economic potential of generative AI” (June 2023)

Furthermore, the current AI paradigm centers on generative and creative capabilities, which could impact white collar task automation in ways not previously fathomed during prior AI waves. These conditions create a compelling “Why Now” moment for enterprise AI adoption:

Source: Goldman Sachs, “The Potentially Large Effects of Artificial Intelligence on Economic Growth”, (3/26/23)

Academics have studied the impact of such potential productivity gains in today’s AI revolution. But this impact is not just theoretical — we’ve seen real-life evidence of tangible productivity gains from GenAI adoption over the past year across different high-level tasks:

Source: Morgan Stanley, “AI Guidebook: Fourth Edition” (1/23/24)

Upon closer inspection, one will notice that most of these gains have been concentrated in use cases where AI is leveraged as an “assistant”, often in situations where the user interacts with AI through a chatbot interface to retrieve information, develop a work output, or answer questions based on prompts. It makes sense that enterprises are embracing this sort of user interface since it mirrors the ChatGPT interface that consumers have become accustomed to. 

But this is perhaps just the first wave of AI’s productivity promise, as a bigger wave — AI “agents” — is about to sweep us off our feet. Unlike a simple “assistant”, AI agents go one step further to demonstrate capabilities and advanced reasoning that enable them to perform tasks on behalf of users. For example, in a consumer setting: instead of simply asking an AI to generate recommendations for a vacation itinerary, imagine the AI being able to execute on these recommendations by making reservations for you. In an enterprise example: instead of having to prompt an AI co-pilot to complete lines of code, imagine an AI acting like a real-life developer to autonomously plan, write, and deploy, code into production.

Aaron Levie, CEO of Box, very eloquently articulated the power of this agent movement in his post from earlier this week:

The next big breakthrough in AI is AI Agents. This is when AI goes from being used as an assistant to chat with, to using AI to accomplish complete tasks that a human might otherwise have to perform. This moves AI from being a "read-only" operation to fundamentally a "read/write" operation. Ultimately, this brings us much closer to the full promise of AI, in particular in the enterprise, where AI can begin to complete any part of a workflow, and we're already seeing examples today of Agents that write entire software applications or respond to customer support tickets. 

As Andrew Ng highlighted in a recent talk, many people had a “ChatGPT” moment when they tried out ChatGPT and it significantly exceeded their wildest expectations, but perhaps a bigger wow moment is awaiting us as we experience the immense power of AI agents. I believe our jaws will truly drop then and the productivity floodgates will be unleashed. Just in the past several weeks, I’ve seen huge leaps in innovation on this front, from multi-agent systems to hybrid agentic workflows that can complete enterprise tasks with a level of reliability that was unfathomable a few months ago. I’m extremely excited to see how this trend of AI agents unfolds and how enterprises will react to this new productivity frontier.

This post and the information presented are intended for informational purposes only. The views expressed herein are the author’s alone and do not constitute an offer to sell, or a recommendation to purchase, or a solicitation of an offer to buy, any security, nor a recommendation for any investment product or service. While certain information contained herein has been obtained from sources believed to be reliable, neither the author nor any of her employers or their affiliates have independently verified this information, and its accuracy and completeness cannot be guaranteed. Accordingly, no representation or warranty, express or implied, is made as to, and no reliance should be placed on the fairness, accuracy, timeliness or completeness of this information. The author and all employers and their affiliated persons assume no liability for this information and no obligation to update the information or analysis contained herein in the future.