I started this on LinkedIn and quickly shot past the character limit, so I made it a small blog post. Okay, so here we go:
Over the next year, natural language processing is going to get much better. Models will become more capable of changing themselves, improving their own work and handling kinds of communication that feel difficult to imagine right now.
That’s the easy prediction.
Within five years, I think personal AI agents will become normal. They’ll manage schedules, communications, purchases, research and a growing pile of daily decisions. For people who can afford good ones, life may get noticeably easier.
That last part matters.
The best agents will be expensive to build and run. They’ll have more context, better tools and access to better models. We already accept enormous differences in housing, transportation, education and healthcare based on what someone can afford. AI could become another divide, except this one helps make decisions across all the others.
Some people will have an intelligence working beside them all day. Other people will get a cheap chatbot trying to sell them something.
I also think we’ll give AI agents legal authority.
Power of attorney sounds wild, but consider what we already ask software to do. It moves money, executes trades, signs documents and applies rules to decisions that shape people’s lives. An agent that understands contracts, legal precedent and someone’s stated wishes could eventually act within carefully defined boundaries on their behalf.
The hard part won’t be whether the AI can read the documents. It’ll be deciding when the system has authority, who is responsible when it gets something wrong and how a human takes control back.
(and here is where people will say that I am crazy)
10 Years from Now
Ten years out, I think AI may be able to create personalized realities convincing enough to change our perception of time.
Imagine putting on a headset—or whatever replaces the headset—and taking what feels like a week-long vacation while only a few hours pass outside. The system knows what relaxes you, what frightens you, which people you want there and what kind of place feels real to you. It builds the experience as you move through it.
Virtual reality, augmented reality and AI all meet in one environment that responds to your mind as much as your body.
What happens when that environment feels better than the life you return to?
What happens to work, relationships or memory when someone can experience days inside a machine during an afternoon? If the AI understands a person well enough to create a world specifically for them, it can probably make that world very difficult to leave.
Call it AI-driven alternate reality, time-perception manipulation or the first practical step toward living inside a simulation. I don’t know what name will stick.
This one is much more speculative. It’s also the prediction I can’t quite shake.
By Matthew Williamson · Originally published on Clevyr · November 13, 2023 — read the original →