Unreal Intelligence

Unreal Intelligence

Download on the App Store

All Unreal Intelligence episodes

  • Pay Me Now or Pay Me Later
    Somewhere in your company right now, an AI agent is writing code nobody is reading. That's not a technology problem — it's an invoice you're signing without knowing the amount.
    This episode is about the three bills coming due across the entire industry: the AI-generated code that looks cheap until you have to maintain it, the "clean" data nobody actually agrees on, and the accountability vacuum that opens up the moment you treat a tool like it can own a decision.
    Karl walks through the real cost of each — from GitClear's data showing copy-pasted code surging while refactoring collapses, to the hundred-times cost difference between LLMs and classical ML on structured tasks — and gives you one concrete Monday-morning move before these invoices start compounding interest.
    Knight Capital. Zillow. Klarna. McDonald's. The pattern is the same every time.
    You can pay now or pay later. Later is where the zeros get added.
    Subscribe so you don't miss what's next.
    23 min
  • Power Tools on a Sinking Ship: How Corporations Are Misusing AI

    Everyone’s buying AI to fix email overload — and nobody’s asking why we have this many emails in the first place. This week I’m going after the most expensive open secret in corporate tech: most enterprise AI deployment is theater. Companies are pouring billions into licenses pointed at surface-level text work while the broken processes underneath keep generating the noise. MIT’s research says 95% of GenAI pilots fail to deliver ROI. The models aren’t the problem. The thinking is.  I walk through the two human failure modes wrecking AI initiatives — overtrust and refusal. On the overtrust side: the Air Canada chatbot ruling that made “the AI said it” legally your problem, the lawyers sanctioned for ChatGPT-invented case law, the Bard demo error that wiped $100 billion

    off Alphabet, and why hallucination is an architectural feature you manage, not a bug you wait out. On the refusal side: the skeptics in your org who won’t touch the tool — and why losing your most experienced people from the AI loop creates the worst possible risk distribution.  Then the fix: treat AI like a smart, fast, occasionally-wrong junior co-worker embedded in redesigned workflows — never an oracle bolted onto broken processes. I close with a four-step Monday-morning playbook: one measured pain point, one named human owner, real outcome tracking, and a gentle on-ramp for your skeptics. Process first, AI second, supervision always. 

    25 min
  • Who am I to argue?
    A senior director just handed her go-to-market strategy to a chatbot. Didn't push back once. When Karl asked why, she said seven words that should worry every leader using AI right now: "Who am I to argue with it?"
    That's the real crisis. Not hallucinations. Not job loss. Deference. Smart, experienced people outsourcing their judgment because the model sounds more confident than they feel.
    This episode makes the case that deference is the single biggest threat to value creation in enterprise AI — and it's already inside your org.
    You'll walk away with:
    - Why confident-sounding output is a feature of how these models are trained, not evidence they're correct
    - The "argument rule" — a one-line process change that measurably improves AI output quality without slowing anyone down
    - The inversion drill: how to get the model to critique its own work before you ship anything that matters
    Your judgment is the asset. The model is the variable. Don't let that flip.
    New episodes weekly. Subscribe wherever you listen.
    26 min
  • The 5% Problem: Why Your AI Agents Aren’t Paying Off
    Ninety-five percent of corporate AI projects are returning basically nothing. Not a little — nothing. And the MIT research behind that number isn't an indictment of the technology. It's an indictment of how companies pick, scope, and measure AI work.
    Karl's position: the failure is organizational design wearing an AI costume. The models are fine. The deployment isn't.
    You'll walk out with a six-step checklist for building agents that actually pay off, a clear framework for why "find the dollar before the use case" changes your prioritization completely, and an honest look at why broad mandates like "transform customer service with AI" aren't projects — they're wishes.
    If you've got a pilot that looked good in the demo and went quiet in real work, this episode is the postmortem you didn't know you needed.
    Follow Unreal Intelligence so you don't miss the next one.
    24 min
  • Your AI Ban Is Your Biggest Security Risk
    A bank blocked ChatGPT at the firewall, sent the all-staff email, and felt good about it. Six months later, forty percent of their knowledge workers were using it anyway — on personal phones, pasting in whatever they needed, completely invisible to the security team. The ban didn't stop the AI. It stopped the company from seeing it.
    That's the argument Karl makes in this episode: a blanket AI ban isn't a control. It's a blindfold. And it may be the single biggest security gap in your company right now.
    You'll come away knowing exactly how to pull a shadow usage report and what to do with the number you find. You'll understand what enterprise AI contracts actually give you that a personal account never will — and why that difference matters legally, not just technically. And you'll get five concrete moves to replace your ban with something that creates real visibility and real governance.
    More usage, less risk. That's not a contradiction. It's the whole point.
    If you've got a ban in place — or you're about to put one in — this one's worth your commute.
    28 min