10 Things You've Always Wanted to Know About AI but Didn't Want to Sound Silly for Asking

We've all been there, nodding along convincingly in the conversation, one side of the brain screaming "Look smart, look smart!!!" whilst the other side is delving into deep hippocampal territory frantically running around going "What's a technical term I can use to pretend I know what's going on?"

This happens a lot when people start talking about AI these days: enthusiasm on the outside, mild panic from within. If you are worried about not being up-to-date with the latest technologies, relax — most others are in the same boat.

And fear not, here's a list of 10 things you can read in under 5 minutes which will have you humming the tune of "We Are The Champions" (or at least managing a confident nod at the next dinner party).

#10 Veo 3

You may have noticed an influx of modern takes on historical events floating around on YouTube Shorts or TikTok recently. That's all thanks to Veo 3, Google's latest video generator that leapfrogs the competition (until, of course, they get leapfrogged themselves). Crazy stuff, but unless you are in the actual business of video and content creation, it's a good-to-know thing only.

#9 MCP (Model Context Protocol)

The easiest way to explain MCP is: imagine AI as a brilliant but isolated brain in a jar. MCP is the nervous system that connects it to actual tools and data sources. It's really exciting in the AI world because it means your AI assistant can finally book that restaurant reservation, check your calendar, or pull real data instead of hallucinating statistics like an undergraduate padding out an essay. Unlikely to come up in general conversation, but you'll gain instant street-cred if you can drop "MCP" seamlessly into it.

#8 AI Agents

Not the Matrix kind. These are AI systems that can actually do things rather than just chat about them. They plan, execute, and adapt — booking flights, managing inventories, or debugging code while you sleep. Great for specific, structured tasks, but usually overhyped on the socials, with little empirical evidence of organisations replacing real employees with AI Agents.

#7 NotebookLM

Google's research assistant that turns your documents into a personalized AI expert. Upload your files, and suddenly you have a PhD student who's actually read everything and can explain it back to you. The audio overview feature, which generates surprisingly natural podcast-style discussions about your content, has become the unexpected hit.

#6 Vibe Marketing

Great tools for the non-marketers, as all you need to do is feed rough ideas to AI to get polished marketing campaigns in return. No more agonizing over the perfect Instagram caption or email subject line — just describe the vibe you're after and watch the machine conjure something that sounds vaguely human but somehow more enthusiastic about your product launch than you are.

#5 AI Slop

The digital equivalent of processed cheese — AI-generated content that floods the internet with the nutritional value of cardboard. Those LinkedIn posts about leadership written by people who've never led anything? AI slop. The product reviews that read like they were written by aliens attempting human emotion? Slop. It's the dark matter of the internet — everywhere, yet nobody admits to creating it.

#4 n8n

The workflow automation platform that's become the Swiss Army knife of the no-code movement. Our prediction is that n8n or other automation tools will become a standard part of the vocabulary within the next few years. It connects your apps like a hyperactive digital plumber, moving data between services with the kind of efficiency that makes middle managers nervous about their job security.

#3 Vibe-coding

One of the most popular terms in AI in 2025, these are basically tools any non-technical person can use to create a website, app, calculator, or anything that tickles their fancy. Vibe-code tools use natural language models so you can basically just put in a request like "Make me an app that calculates estimated calories of foods based on the picture I take of it" and it will go away and do its thing. Think Cursor, Windsurf, or Replit Agent — the democracy of coding has arrived, and it's wearing flip-flops.

#2 Claude Code

Anthropic's command-line tool that lets developers delegate entire coding projects to AI. If your company has a decently sized IT development team, chances are: a) they're currently already using Claude Code to do most of their work and are just chilling for the remaining 6 hours of their day, or b) plans are underway to revolutionise how software development projects are executed, using AI-coding tools as the backbone for doing the "grunt" coding work.

#1 Constitutional AI and Safety Debates

The thing everyone's actually talking about but nobody wants to admit they don't fully understand. As AI models grow more powerful, the question of how to keep them aligned with human values has gone from academic exercise to urgent priority. It is honestly a massive subject and this post cannot do it justice. When someone inevitably mentions "the robots are out to kill us" at a dinner conversation, just reply, deadpan, "Don't be silly, we're safe. Our kids, however…"