Eighteen months ago, it was easy to be a skeptic, I certainly was! The AI landscape felt dominated by parlor tricks, hallucinating chatbots, six-fingered AI hands, and models that could pass a bar exam but couldnt reliably tell time on an analog clock. But as we sit in mid-2026, the narrative has fundamentally shifted. The industry has moved from generating words to executing complex workflows. The gap between what AI can mimic and what it can actually do has closed. If you arent an advocate by now, you simply arent paying attention.
The Last 18 Months: From “Glorified Autocomplete” to “Wait, Did It Just Think?”
- AI Learned to Think (and Sweat): For a long time, AI was basically the world’s most advanced autocomplete – it just guessed the next logical word. But recently, we entered the era of reasoning models. Now, when you ask a complex math or logic question, the AI actually takes a pause to think about it before answering.
- The Six-Fingered Hand Problem is Dead: Eighteen months ago, AI image generators were notorious for creating terrifying human hands with too many knuckles. Today? We have text-to-video AI. You can type A cinematic drone shot of a golden retriever snowboarding on Mars, and ten seconds later, you have a clip that looks like it cost £10 million to produce.
- The Rise of Vibe Coding: You no longer need to know what a Python script or a C++ library is to build software. We’ve entered an era developers are jokingly calling vibe coding. You just aggressively describe what you want an app to do in plain English, and the AI builds it.
- Chatbots Grew Hands (aka AI Agents): The old AI could give you a recipe for lasagna. The new AI (called an Agent) can look up the recipe, see what’s in your smart fridge, add the missing ingredients to your grocery delivery cart, and check out. They’ve moved from just talking to actually doing.
The Next 18 Months: Get Ready for the Clingy Bot Era
So, what happens between now and the end of 2027? Grab some popcorn (and maybe upgrade your home Wi-Fi), because things are about to get weirdly personal. Here are some of my predictions for the next 18 months.
- You Are Now a Middle Manager for Bots: The big corporate buzzword for the next year is Multi-Agent Workflows. This means you won’t just use one AI. You’ll have a team of them. You’ll give a project to your Manager Bot, who will delegate the research to the Research Bot, who hands the data to the Writer Bot, who gives it to the Fact-Checker Bot. Your new job will basically be sipping coffee while making sure your digital employees arent slacking off.
- Proactive AI (It’s Going to Get Clingy): Right now, AI sits in a little box and waits for you to type a prompt. In the next 18 months, AI will become proactive. It will start reaching out to you first. Imagine waking up to a notification: Hey, I noticed your morning meeting got moved, so I pushed your alarm back 45 minutes, ordered your usual coffee, and drafted an apology to your boss for that typo you made yesterday. Helpful? Yes. Slightly invasive? Also yes.
- The Great Electricity Panic: All this thinking requires an absurd amount of power. The AI industry is building massive data centers that consume energy like it’s going out of style. Over the next year, expect to see a lot of dramatic news headlines about tech companies desperately trying to build their own nuclear reactors just to keep their chatbots running.
- AGI Rumors Will Reach Fever Pitch: AGI (Artificial General Intelligence) is the holy grail -an AI that is as smart as a human across every single subject. We aren’t there yet, but tech CEOs will spend the next 18 months vaguely hinting that its right around the corner to keep investor money flowing.
The Bottom Line: The robots aren’t taking over the world in a dramatic sci-fi laser battle. Instead, they’re just going to slowly take over our spreadsheets, our video editing, and our calendar invites until we literally cant remember how we survived without them.