What's Coming Next in AI: My Predictions

Okay I want to be upfront about something: I'm not an AI researcher. I don't have insider knowledge. I'm a person who uses AI tools constantly, reads way too many things about them, and has opinions. So take these predictions for what they are, one person's read on the direction things are heading, based on what I'm seeing and what seems logical to me.

Some of this might be wrong. I could be wrong about the timeline, wrong about the specifics. But I genuinely believe all of it directionally, which is more than I can say for a lot of predictions I read that feel like the person is just saying what sounds impressive.

The Workflow Integration Is Going to Get Much Deeper

Right now, most people use AI as a separate step in their workflow. You go to a tool, you type something, you get something back, you put it into whatever you were doing. That friction is going to disappear.

We're already starting to see it with AI that operates inside your existing tools rather than alongside them, things that can see your screen, understand what you're working on, and help without you having to context-switch. That's going to become the norm. The idea of opening a separate chat window to get writing help is going to feel dated in a couple of years, I think. It'll just be woven into whatever you're already using.

For content creators specifically, I think this means AI that understands your full body of work, your audience, your past performance data, and uses all of that context automatically rather than having you re-explain it every time. I'm honestly excited for that. Rebuilding context from scratch is one of the most tedious parts of my current workflow.

Video Generation Quality Is Going to Cross a Threshold

I said in my tools post that AI video generation isn't quite there yet for my main content. I think that changes. I don't know if it's six months or eighteen months, but the trajectory is pretty clear, the quality improvements have been fast and they're continuing.

When it does cross that threshold for me personally, I think it reshapes a lot of things about content creation. Not eliminates human creators. I don't believe that, but changes what we spend time on. The technical production stuff becomes much less of a bottleneck. What matters more is the concept, the perspective, the specific knowledge or experience that makes something worth watching.

Real talk, I have some anxiety about this. Not in an "AI is going to take everything" way, more in a "the medium I've built skills in is going to change significantly" way. I'm trying to stay curious about it rather than defensive. The creators who figure out what their actual value is and lean into that are going to be fine. The ones who are competing primarily on technical production quality are in a harder spot.

Personalization Is Going to Get Uncomfortable

Here's one where I'm not fully sure how I feel about where we're headed. The personalization capabilities of AI are advancing in ways that are genuinely impressive and also kind of unsettling when I think about them past the surface level.

Right now, personalized content mostly means "you engaged with things like this before, here's more like that." What's coming is more like AI that adapts the way it communicates in real time based on what it knows about you, your communication style, your knowledge level, what emotional register you respond to. That has legitimate useful applications. It also raises real questions about manipulation and what it means for people to receive fundamentally different information presented in ways specifically engineered to resonate with them.

I don't have a neat resolution for this. I just think it's worth thinking about before we're fully in it. The businesses and tools that handle personalization in transparent, trustworthy ways are going to matter more over time.

The Agent Era Is Going to Actually Happen This Time

We've been hearing about AI agents, AI that can take actions autonomously rather than just respond to prompts, for a few years now. It keeps getting announced and then not quite being real. I think it actually becomes real and reliable in the next couple of years.

What I mean specifically: AI that can handle multi-step tasks without constant hand-holding. Research something, compile findings, draft a response, send it, without you approving each individual step. Right now this kind of thing is flaky enough that it's not really practical for things that actually matter. That reliability problem is getting solved.

For online business this is significant. A lot of what takes time in running a small business is multi-step coordination tasks that don't require deep expertise but do require attention and follow-through. If AI agents can reliably handle a category of those, and I think they will, it changes how much one person can manage.

The Skills That Will Matter Are Changing

This is the one I think about most practically because I'm trying to figure out what to actually invest my time in learning.

My honest read: being good at using AI tools is already a valuable skill and stays valuable for a while. But the more durable thing is being genuinely good at the underlying domain, whatever your thing is. Writing, design, business strategy, whatever. AI makes mediocre work easier to produce. It makes great work possible for more people. But it doesn't really generate genuine expertise or distinctive perspective. Those still have to come from somewhere.

The skill of knowing what good looks like, having taste, being able to evaluate outputs and make them better, is going to be really important. AI can produce a lot. Knowing which of the things it produces is actually good requires judgment that still has to come from a human who knows something.

I'm investing time in going deeper on the things I actually know and care about, using AI to produce faster and cover more ground, but not treating AI as a substitute for actually learning. That's my plan. We'll see if I'm right about it.

None of this is certain. I've been wrong about things before and I'll be wrong about things again. But this is genuinely where my head is at right now, and I figured that was worth more than a list of things that sound smart but that I'm not sure I actually believe.