What is the realistic next big step for AI?

AI is rapidly advancing at a rate many industries have never seen before - and it's one that many people in the tech scene struggle to understand just how insane this pacing is. In just a few short years we have gone from models with knowledge caps that made frequent errors, had easily detectable nuances and were expensive, to today where they can speak, convey emotion, use audio, video and files, browse the web, and even use your computer with agents, and even AI that can think with Chain of Thought.

Many perceive the next big step as AGI or Artificial Generative Intelligence - AI that can truly think and teach for itself, even on things it was never trained on or prompted to do.

This is a monumental task. It is on the horizon, but I don't think it's the next big step in the larger picture. Even if it does happen, it will be expensive for a long time before it becomes feasible for even some of the most tech-fluent of us.

I think the next step is taking all of the amazing things that these companies have created, and taking them mainstream in a way that doesn't feel intrusive. Slowly introducing background AI agents, and chat tools that are really helpful to a task and don't feel obtrusive like how they were. Teach people that AI is a genuinely helpful resource if used well. Because at the moment while AI feels natural and even vital to some of the more tech-fluent, many of the general population barely even know what ChatGPT is or what it can do, and I guarantee if you asked them what Claude or Perplexity are, they'd look at you confused.

This is also the reason I have faith in Apple Intelligence. It may be SUPER gimmicky now, but when we hit the point of not needing to brag about AI and can move to the subtle features that improve UX in really clever ways, which is the long term plan, that's where it will truly shine.

I wish I could say the same for Microsoft with Copilot, but their marketing and forced rebrand have been so aggressive they leave me with little hope that it will scale far beyond gimmicks and Recall.

Google have interesting leverage given their breadth of services - it all weighs on if they choose to advance past a tacked in chatbot and showy Gemini bits and make it more subtle and integrated.

All in al I think the true winner of the AI race will be whoever can appeal to the genuine consumer and realise that AI doesn't have to be a gimmick or party trick - and it shouldn't be advertised as one.