Acquiring Agency
“Acquiring Agency” – a cartoon that illustrates the difference between AI assistants and AI agents.
Earlier this year, Ryan J. Salva, senior director of product at Google and an ex-GitHub Copilot leader, said he’s come to ‘hate’ the word ‘agents. “I think that our industry overuses the term ‘agent’ to the point where it is almost nonsensical.”
AI assistants first came to general awareness with IBM Watson, which gained fame by winning Jeopardy! in 2011, and then Siri and Alexa popularized AI assistants for everyday use.
These assistants helped us by following commands and responding to requests, acting as a reactive tool for tasks like searching, scheduling or automation. While the world of AI agents can also be traced back to the 1970s, it was DeepMind’s Alpha Go beating of the Go world champion in March 2016 that put it on the map.
With the arrival of ChatGPT, AI models could now understand complex prompts, adapt to context and generate dynamic responses, making AI agents more versatile and responsive.
We are now at the point where these ‘agents’ can autonomously make transactions and adapt their tasks to suit an overall objective working with other agents in a swarm.
A simple way to think about this is that AI assistants are task specific, while AI agents are on a path to agency, where they think more independently.
Ultimately, AI agents will replace all AI assistants, but it will be a human ‘trust’ or ‘comfort’ factor that will determine the pace of this change.
13
0
1
Sources:
Lakshmi Varanasi (Mar 20, 2025) – AI agents are all the rage. But no one can agree on what they do. – Business Insider
Jodie Cook (Mar 18, 2025) – AI Agents Explained In Simple Terms Anyone Can Understand – Forbes
