AGI should be neither seen nor heard

Let’s imagine the possibilities.

We’ve all seen the movies (or at least the memes). Whether we’re imagining Iron Man’s AI-powered assistant JARVIS, or Rosie the Robot, the autonomous maid from the 1960s cartoon “The Jetsons,” the goal seems to be for everyone to get their own chatty sidekick to help with chores. 

Humans and machines, side by side, working and talking and playing together. It’s almost as if our future is being written to make robots look good on camera. 

Is that really the best way to integrate AGI into our lives? What if, instead, we conjured a future where humans keep doing the things we’re good at while AGI works behind the scenes to keep the show going. 

Presenting: Background AGI

So far, the addition of AI into the average person’s life has had a singular result: it increases the number of decisions we have to make. There is no killer app for advanced AI yet. We’re all experimenting (or not) at our own pace and there isn’t a single common thread keeping us all coming back. 

Chatbots are the promise of amazing solutions for very specific problems that nobody really has yet. 

You might use it for writing, the next person might use it to generate images, but the only people using it to make their lives easier or make money are influencers and insiders.

Video games are more fun, Netflix has better stories, and most people would rather chit chat with other humans. Your average person isn’t staying up late vibe coding or generating campaign spreadsheets.

Eventually the novelty is going to wear off. And that sweet spot between people willing to pay for writing/coding/image assistance and those who don’t need AI-assistance to make a living may not be as big as the powers that be believe it to be. 

The only path forward is advanced AI systems to become so useful that we forget we’re using them. Simply put: AI needs to fade into the background.

What’s the endgame for LLMs? The proliferation of executives who have to consult their chatbots to answer questions during a board meeting? 

What if, instead, it was real-time sensor processing for everyday people? Maybe we pivot to heads up displays or external brain-computer interfaces. Whatever it is, it won’t involve locking yourself in a room to work with your pal, the chatbot or ceding space in your life for a physical humanoid presence. 

Whether we’re talking chatbots or robots, they’re not going to be useful if they increase our cognitive load. 

This means bulky, humanoid robots that occupy space all over the house constantly carrying, using, and exchanging tools in order to perform tasks are likely to increase the number of decisions we need to make. We’ll have to work around them. Or, if we interrupt their schedules, we’ll have to work around the work they didn’t get done. 

A central control system capable of adapting to our lifestyles, however, could make our lives easier without saying a word or moving a finger. Wouldn’t it be nice if our appliances knew when to shut up because we’re on the phone, or when to get rid of spoiled food? 

What if, instead of a robot that ran over our pets’ tails, our WiFi system could alert us when our kids or pets are stressed out over something?

The possibilities for AGI are endless as long as we imagine ways the machines can support humanity without pretending to be us.

Read more: What are people expecting from domestic humanoid robots?

Art by Nicole Greene