My Two Predictions for AI in 2026 (and they're Bad)

2025-12-20

Let me make two quick predictions for AI in 2026, and both of them are bad.

#1: "Human-Generated Content" Disclaimer

Filtering real human-generated content from AI-generated content will become a huge challenge in 2026.

This is already evident in places like LinkedIn. A couple of years ago LinkedIn was mostly "signal-vs-noise", but in 2025 it became "noise-vs-better noise". I think the same will apply everywhere soon — from blogs, vlogs, podcasts, to even news articles.

In 2026, people will be forced to add a disclaimer with every content that they create, like: "Disclaimer: This content is human-generated" to highlight authenticity. It will be fun to watch this.

I already started putting a notice in some places that says not every em dash comes from ChatGPT, because I use a lot of em dashes in my writing for many years, and now it is an indicator of AI-generated content.

#2: AI-Everywhere for No Apparent Reason

It was blockchain a few years ago. Everyone wanted to put blockchain in their business, even it didn't make any sense at all.

Now it's AI.

I just read an interesting article (that inspired me to write this post) about AI being used everywhere and to prepare for that stupid world. In the article, the author emphasises the fact that "even if it is completely stupid, AI will be everywhere, get used to it!" by showcasing a video interview with an employee at Anthropic, discussing the infamous AI-powered vending machine experiment.

I can see this coming in 2026: "Grab our AI-cooked sandwich with AI-picked ingredients, built on a multi-model agentic toaster"

AI is extremely useful and powerful, it will change the world, and elevate the human race to new heights, but AI isn't a silver bullet for everything. Some things are better left alone.