Should I set up a personal AI agent to help with my daily tasks?
– Search for help
As a general rule, I think relying on any kind of automation in everyday life is dangerous when taken to extremes, and potentially alienating even when used in moderation, especially in personal interactions. An AI agent that organizes my to-do list and collects online links for further reading? The myth of an AI agent that automatically texts my parents every week with a quick life update? Terrible
However, the strongest argument for not using artificial intelligence generative tools in your daily routine is the environmental impact of these models during training and output generation. With all that in mind, I dug through the WIRED archives, published at the glorious dawn of this mess we call the Internet, to find more historical context for your question. After a bit of searching, I came back convinced that you probably use AI agents every day.
The idea of artificial intelligence agents, or god forbid “Agent AI”, is the current buzzword for every tech leader trying to promote their latest ventures. But the concept of an automated assistant dedicated to completing software tasks is not a new idea. Much of the discourse surrounding “software agents” in the 1990s mirrors the current conversation in Silicon Valley, where tech company leaders now promise an influx of AI-powered productive agents trained to do online work on our behalf.
A WIRED interview with MIT professor Patti Mace, originally published in 1995, says: “One problem I see is people asking who is responsible for the actions of an agent. “Buying something you don’t want will raise a lot of interesting issues for you, but I’m convinced we can’t live without them.”
I caught up with Maes in early January to see how his views on AI agents have changed over the years. He’s as optimistic as ever about the potential of personal automation, but he’s convinced that “too naive” engineers don’t spend enough time addressing the complexities of human-computer interaction. In fact, he says, their recklessness could create another artificial intelligence winter.
“The way these systems are built now, they’re optimized from a technical standpoint, from an engineering standpoint,” he says. But, they are not at all optimal for human design issues. He focuses on how AI agents, despite improvements in the underlying models, are still easily fooled or resort to biased assumptions. And a misplaced trust causes users to trust answers generated by AI tools when they shouldn’t.
To better understand other potential pitfalls for personal AI agents, let’s break this nebulous term into two distinct categories: those that feed you and those that represent you.
Feeders are algorithms with data about your habits and tastes that sift through chunks of information to find what’s relevant to you. Sounds familiar, right? Any social media recommendation engine that populates a timeline with custom posts or an incessant ad tracker that shows me that mushroom gum for the thousandth time on Instagram can be considered a personal AI agent. As another example from a 1990s interview, Mace mentioned a news aggregator that was well-organized to return the articles he wanted. It looks like my Google News landing page.
