AI Is Moving Fast. Adoption Isn’t. The Expectation Gap Is Getting Dangerous

Unfiltered with Matt & Nige is back 🎙️ And Episode 05 starts in classic Unfiltered fashion: the boys still don’t fully nail the intro.For this episode Matt is in New York to see Nigel. Nigel is in Basingstoke.They immediately go off on a tangent about tracking airplanes, boarding lofts, shelving projects, Hornby train sets, and what retirement apparently looks like.Once they find the thread, the real theme of the episode appears…AI is accelerating at a massive pace… but adoption and change aren’t keeping up.TLDR: the expectation gap is widening.Boards think “AI is the answer”. Teams are still trying to make the basics work. And that gap is getting bigger, faster.This week on Unfiltered the two talk all things:* the widening gap between AI capability and real adoption (in big companies and in normal life)* why the tools are everywhere… but the “so what do I actually do with this?” problem is still real* AI talking to AI and whether we’re creating more noise than signal* the confidence problem: incomplete answers delivered with maximum certainty (and why it drives Matt mad)* the future of advice vs guidance (financial advice, brokers, advisors… what gets augmented vs replaced)* plus a surprising hero: pen and paperGetting a little deeper… the episode is really about trust.Not “trust the AI”.Trust in the whole system: the data going in, the outputs coming out, and what happens when people start building decisions on top of the wrong foundation.The adoption gap: “AI is the answer… what’s the question?”Nigel’s point was sharp: execs are excited, boards are excited, everyone’s saying “replace legacy systems”… but the 10,000 to 30,000 people inside the organisation aren’t there yet. Because change management isn’t solved by a model release.And the business doesn’t move at the same pace as the tooling.Matt took it wider: it’s true for individuals too.The case studies are everywhere, the promise is everywhere, but it still takes time to return to the tools, test features, build habits, and work out what fits your life. And even then… you’re not always ready to hand over the keys.Pen, paper, and the “Limitless” brain stateThis was one of the best bits: Matt’s back in notebook mode.Write notes everywhere, snap them, dump them into AI, turn them into action.Simple, old school, effective.Nigel’s similar with writing: he doesn’t use AI to write his newsletter because writing is how he does his thinking.The takeaway:AI can accelerate output, but it doesn’t replace the thinking part.And sometimes the thinking part starts with… a pen.AI talking to AI (and the rise of “friction”)This section was basically modern life in one sentence:Matt’s in a dispute with a car hire company. They’re sending AI generated letters. He’s replying with AI generated letters. Everyone’s confident. Nobody’s solving anything.Nigel’s broadband provider is charging him for kit they’ve already confirmed they received. 30 minutes on the phone. Process broken. Total friction.Btw, if you’re still tuning in at this point, you’re doing well.The confidence problem (the bit that really annoys Matt)Matt’s rant here is extremely relatable:AI can give you an incomplete answer with full confidence, plus the weird flattery layer, and it creates a bigger downstream problem because people build on top of it.Nigel’s extension was important: if wrong insights keep feeding systems, you end up with more false information, more noise, and it gets harder to find the needle in the haystack.And that’s where “market intelligence” stakes go up.Bad inputs don’t just waste time, they distort decisions.Advice vs guidance: brokers, advisors, and what gets augmented firstThey then pulled it into financial advice and the “guidance vs advice” line.How far can AI go in comparing options, explaining, narrowing, and helping you decide - at a fraction of the cost?Same question as insurance, really:Brokers and advisors may not disappear overnight… but the simple stuff gets augmented first, then compressed.And then the nuance: it’s not only trust, it’s recourse.If something goes wrong, who do you go back to?The “raise the floor, raise the ceiling” bitThis episode also had a strong thread about learning: the floor is rising, everyone’s leveling up, and curiosity is becoming the advantage.Not just using the tools… but using them with the right questions:“show me”, “prove it”, “unpack it”, “what’s missing”, “what would change your mind”.That’s how you avoid fake certainty.And that’s how you make the tools genuinely useful.The human ending (because obviously)They wrap in full Unfiltered chaos: headphones dying, random batteries, coughing across 6,000 miles, travel and Substack plans.Also: the idea that AirPods should be insured as a utility subscription 🤷The big takeawayAI is moving fast.Adoption isn’t.And the gap between what leaders expect and what organisations can safely deliver is widening.The winners won’t be the ones shouting “AI first”.They’ll be the ones who can turn capability into adoption - without creating more noise, more friction, or more fake certainty.Subscribe for more Unfiltered every week 🎙️ This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit unfilteredwithmattnige.substack.com