Should I create a personalized AI broker to assist me in my daily tasks?
—Searching for Assistance
I believe that using any form of technology in your daily life is risky when taken seriously and possibly upsetting even when used in moderation, especially when it comes to personal relationships. An AI broker that compiles my task lists and compiles online resources for reading later? Fabulous. An AI adviser that sends my parents weekly fast updates on their lives? Horrific.
The effects these models continue to have on the environment during training and production generation is the strongest argument against never incorporating more generative AI tools into your everyday routine, though. With all of that in mind, I dug through WIRED’s library, published during the beautiful sun of this chaos we call the web, to get more historical perspective for your problem. You’re probably currently using AI brokers every day, I think after a little browsing.
The idea of AI brokers, or God-forbid “agentic AI”, is the latest buzzword du day for every software leader who’s trying to buzz their new investments. However, the idea of an automated associate working on applications jobs is not particularly novel. In Silicon Valley, where frontrunners at technology companies then promise an approaching flood of relational AI-powered agents trained to perform website chores on our behalf, the discussion around” software agents” in the 1990s is a stark echo of the present discussion.
A 1995 WIRED meeting with MIT doctor Pattie Maes, which was previously published in 1995, states that “one of the issues I see is that people will question who is accountable for the actions of an agent.” ” Specifically when officials spend too much time on a machine or make a purchase on their behalf,” the statement continues. Brokers will bring up a lot of intriguing problems, but I’m convinced that we won’t be able to survive without them.
I contacted Maes in January to find out how her view on AI officials has evolved over time. She’s as cheerful as ever about the potential for personal technology, but she’s convinced that “extremely stupid” designers are not spending enough time addressing the challenges of human-computer relationships. In fact, she says, their recklessness could induce another AI winter.
” The way these systems are built, right now, they’re optimized from a technical point of view, an engineering point of view”, she says. ” But, they’re not at all optimized for human-design issues”. She focuses on how, despite improvements to the underlying models, AI agents are still susceptible to being deceived or made up, despite advances in technology. And a lost sense of security encourages users to answer questions generated by AI tools when they shouldn’t.
Let’s divide the term into two distinct categories: those that feed you and those who represent you, to better understand other potential pitfalls for personal AI agents.
Feeding agents are algorithms that use data from your habits and tastes to comb through vast amounts of data to determine what’s important to you. Sounds familiar, right? Any social media recommendation engine that fills a timeline with specific posts or incessant ad tracking that shows me those mushroom gummies on Instagram for the thousandth time could be regarded as a personal AI agent. As another example from the ‘ 90s interview, Maes mentioned a news-gathering agent fine-tuned to bring back the articles she wanted. That resembles my Google News home page.
As for AI agents who represent you, they are just like algorithms, but they were created to mimic your behavior and act on your behalf when you interact online. Imagine having a digital twin who can call a restaurant to reserve you a table, attend video meetings for you, and write thank-you notes to your wedding guests. Additionally, generative AI tools that mimic famous people and interact with their fans are included.
I would steer clear of using any of these AI replicas, regardless of whether I used one or interacted with one. The importance of the digital world has overtaken that of the physical world, and screens are now largely responsible for our quality of life. Let’s leave the automation to the tedious tasks, even if the technology is eventually good enough to send an AI agent to FaceTime my grandma or let it loose in the group chat to riff with my friends doesn’t feel like a laughable idea. We should show up as ourselves in online interactions. It’s increasingly all we have left.
