Artificial Intelligence Predictions for 2026

My prediction for AI in 2026 is that it's just gonna keep growing. I don't think it's some kind of bubble. Some of the things people are trying with LLMs won't work out, but plenty will.

We Are In The Trough Of Disillusionment

We are here: trough

LLMs Are Not Your Friend

LLMs, the tech that is now synonymous with AI, is a technology for figuring out the next best thing to say. I think of it as a digital "yes man". You can't talk to them like they're your friend and get a real relationship like it's with a human being. The AI Girlfriend apps are way creepy.

AI Will Not Replace Software Engineers

Have you ever heard the saying that there are two parts to every software project? The first 90%, and the second 90%? AI is great at getting that first 90% done, and also making that second 90% even more arduous.

I've seen enough AI code slop to be confident that my software engineering skills will be needed for some time, even with lots of fun "vibe coded" software getting churned out. I'm interested to see how much the "get your vibe-coded app ready for production" industry grows.

AI Is Really Helpful For Writing Code

Having said that ChatGPT isn't going to replace software engineers, I also want to share that I find ChatGPT very useful for software engineering. Recently, I needed to write a function that would return a TimeSpan until the next time a job should run, and there were very specific business rules around it. I detailed those rules to ChatGPT and it gave me a function that worked correctly. After reviewing and testing the code as I would any other software going in to a production system, it became "my code" rather than "vibe code". And it made my life a bit easier.

You Can't Really Trust What AI Business People Or News Articles Say

Because there is so damn much money invested in companies with Large Language Model technology, the leaders of those businesses need to keep their investors happy. At least, happy with their investments. As such, any shortcomings or problems will be hidden as much as possible, and any successes will be hyped up. Because this is the next exciting big thing, news articles are going to hype things up as much as they can, in both positive and negative perspectives, because that gets the clicks and the traffic. That's the world we live in.

There's usually some truth to the matters at hand, but more often than not, things are made out to be more exciting than they really are.

Some News

Just for reference of recent events when I look back at this.

Anthropic, owners of the Claude programming IDE, announced on November 11, 2025 that it would be investing 50 Billion Dollars in AI infrastructure.

Nvidia, a computer chip maker, announced on September 22, 2025 that it would be investing Up To 100 Billion Dollars in OpenAI.

OpenAI, the company that owns ChatGPT, has reached a valuation of 500 Billion Dollars as of this news from October 2, 2025.

I saw a bunch of headlines about how the Big Short guy had bet against AI markets, but now I can't find any articles with specific numbers. It leads me to think that those headlines were more click-bait than real evidence that we're in an AI bubble where a bunch of money is going to evaporate one day. It looks to me more like a couple of people just arguing on the internet.