Videos — Page 4

Lightly edited for clarity.

So why hasn’t AI made productivity skyrocket overnight? It’s made creating emails, research, code cheap, but the hard part is connecting the dots, coordination across teams and systems and decisions.

More output floods in. There’s more drafts, more ideas, more analysis, more code. But someone has to prioritize it, someone has to integrate it, someone has to decide what matters.

And that coordination is expensive. It slows everything down if you don’t have the right system.

So you have an engineering team, right, and they’re using AI to turn out code. They’re likely producing much more code than they were before, but now you have more code to manage.

And if reviewing, merging, testing, and shipping it still moves at the old pace, you haven’t removed your bottleneck, you’ve shifted it downstream.

So if two different companies are using the exact same AI tools, and one thrives while the other stalls, the difference is who coordinates better.

Lightly edited for clarity.

When information is everywhere, attention limits what you can actually act on.

AI generates endless insights. It’s a divergence engine. But while you can do anything you want, you can’t do everything you want, and without a system to filter and prioritize and route information, you get noise, not clarity.

So if you’re responsible for a brand or a website, hopefully you’ve started looking at GEO analytics. And if you haven’t or you don’t know how, I have a whole team of people that would love to talk to you about it.

So you looked at your analytics, you used AI to analyze it, and it surfaces 50 different opportunities for personalized content, different industries, different intents, different customer journeys. Could you build all of them? Probably. Should you?

Winning isn’t just about producing more, it’s about deciding what actually deserves your attention.

And when people say judgment is a key human skill in the age of AI, that’s what they mean.

Lightly edited for clarity.

AI capability isn’t the same as AI readiness.

As AI gets more powerful, governance becomes the key constraint.

The organizations that can manage risk well, deploy faster and scale better.

And it’s not really about the tools.

It’s about ensuring security, reliability, and compliance, because if you can’t audit it and monitor it, then you can’t scale it.

Governance doesn’t slow you down, it’s what lets you move fast safely.

Lightly edited for clarity.

So when everyone has powerful tools, what actually sets you apart?

When execution is easy, your competitive advantage shifts to the systems that organize and coordinate that execution.

If anyone can produce research and analysis and code, capability gaps shrink, and the advantage moves to how you integrate workflow systems, feedback loops, and decision-making processes.

Those systems are harder to copy, and that’s where the competitive advantage is.

If we think about two competing startups, they’re both using Claude Code. They’re both using the same skills. Who wins?

It’s the one with the better process.

Lightly edited for clarity.

Technology changes constantly, and the people who succeed are the ones who learn the fastest. That is the ultimate competitive advantage. AI lets you experiment faster. You can try ideas, test approaches, ask questions, iterate.

But learning only happens if you actually try something. See what happens and learn from it. And here’s the easiest way to see this for yourself. Think of something you don’t know how to do, something useful or not.

Open a model, explain what you’re trying to accomplish, give it some context, and start working through it. Do that consistently, and you’ll learn faster than the environment changes. And learning is a huge advantage.

Lightly edited for clarity.

Automation does something sneaky. It takes away the practice. When the system can automate drafts, code, research plans, people stop doing it normally and then become supervisors, which sounds like promotion or nightmare, depending on what you like about your job.

But here’s the thing. The system runs fine most of the time. And because it runs fine most of the time, you’re least prepared for when it doesn’t. And that’s not an AI problem, or even an automation problem really.

Your teams using AI to crank out client recommendations, right, week after week, you get polished decks. They’re clean, you can turn them out quickly. But one day the model misses something, security compliance, actual business reality, and now you have to catch it. But you haven’t been building that muscle because the system made it feel unnecessary.

Are you gonna catch it or is your process gonna let it through? Again, the differentiator isn’t just who can generate, everyone can generate. It’s who keeps judgment sharp, verification habits, escalation paths, decision logs, and failure drills for the weird cases.

Lightly edited for clarity.

We’ve all built a habit that when a computer says something, we think it must be right. I made an appointment. Well, it’s not in the system. I sent an email. Well, clearly you didn’t, because I didn’t get it. This program produced this report, so it’s right.

And the more advanced the system looks, the more likely we are to just follow along. We see it all the time. People will follow their GPS even when it’s clearly wrong. And in business it’s the same deal. When an AI recommendation comes in, we’re more likely to go with it, especially because it’s coherent. But that doesn’t mean it’s correct.

We’ve all heard stories of people following their GPS, onto a non existent road or even into a lake, and it’s easy to laugh at. But the same thing can happen in business, like major financial decisions or deciding who to hire. And having AI doesn’t set you apart. Everyone has AI. What sets you apart is the habit of stepping back and asking, wait, does this actually make sense.

Lightly edited for clarity.

I’ve driven from my home to my office countless times, I’m sure most of you have as well, and we’ve all become accustomed to using GPS to get us there, efficiently. But when you turn your GPS off, what happens? You still know how to get to work, right, but you start second guessing specifics like, this is the road I should turn on, right?

It’s called cognitive offloading. The more we rely on external tools, whether it’s GPS or search engines or AI assistance, the less we build our own mental models. We remember how to get to the answer, but not the answer itself. And that’s not necessarily bad. Tools have always done this. Calculators did it for arithmetic, search engines did it for facts.

But AI is different in one very important way. It doesn’t just help you find answers, it helps you produce them. So if you rely on it for analysis, reasoning, or writing, something subtle can happen. Over time, you get faster at producing work, while your mental model gets weaker. You still recognize a good answer when you see one, but starting from a blank page starts to get harder.

And the goal is not to avoid using tools. It’s not to avoid using AI. That’s ridiculous. The goal is to make sure you’re still building your mental model while you use it. Efficiency is great, but when the situation changes, and it always changes, understanding is what gets you to your destination.

Lightly edited for clarity.

So AI is cranking out a lot of the work now, and what you’re hearing is you need to provide judgment and taste, right?

Well, judgment is a lot trickier than people think. You’re not born with it, and it’s not some superpower you get just automatically by being human. Judgment works really well when patterns repeat, and there’s a well-defined set of rules, like chess, for example. Experienced players can look at a board and immediately understand what’s going on in the game, and good players can quickly determine who’s likely to win and how.

But most business decisions don’t work like that. The feedback is slow, the signals are messy, and a lot of context isn’t available, right? Conversations in the hallway, relationships between people, organizational politics, real-world constraints.

And when you’re making decisions like restructuring a team, or launching a new product, or entering a new market, you might not know for a year or more whether it was brilliant or a total mistake, or as is most often the case, somewhere in the middle.

So judgment’s not something you automatically move up into once AI does the work. It’s something you build. You test ideas, you get feedback, you challenge your assumptions, you pay attention to what actually happens. In other words, judgment isn’t something you have just because you’re human. It’s something you train.

Lightly edited for clarity.

I was talking to my brother the other day about how he uses AI, and one of the things that he does is ask it to build the fastest, most efficient plan possible, for whatever task he’s trying to accomplish, reduce the manual steps, cut the friction, get to the result quicker, which is really smart and makes perfect sense.

But as a counterpoint, AI is already making the work faster, 10faster50faster, sometimes10times faster. And so before you go looking for even more speed, maybe slow down for a second, and make sure you actually understand the problem. Look at your options. Build a mental model, train your judgment. See if there are different ways to approach this problem. Make sure you’ve got your car pointed in the right direction, before you hit the nitrous. Cause once AI’s in the loop, speed amplifies everything. If you’re headed the right way, that’s amazing. But if you’re headed the wrong way, and regardless, you still need structure and auditing and feedback loops, a way to catch mistakes as early as possible. Because speed doesn’t just amplify progress. It amplifies errors too.

Lightly edited for clarity.

The real risk of AI is that we stop thinking. And this happens with automation all the time. When a system works most of the time, people naturally stop practicing the underlying skill.

Pilots are a good example. Modern aircraft fly on autopilot most of the time. It’s incredibly safe. It’s incredibly efficient. But because of that, pilots train for the moments when something breaks, bad weather, system failures, situations where they suddenly have to take manual control. Because when those moments happen, they can’t be rusty.

And AI is creating similar dynamics in knowledge work. If the model writes the draft, and it summarizes the research, and it generates the analysis, it’s easy to slide into a role where you’re mostly just approving what the system produces. And that works right up until the moment something important is wrong.

So the goal isn’t to replace people. It’s to design systems that keep them thinking, review processes and decision logs and second opinions. And I like to do regular manual mode, where you work through problems yourself. And not because AI isn’t useful, but because the skill you stop using is the skill you lose. And do you really wanna lose the skill of thinking.

Lightly edited for clarity.

AI agents don’t just generate, they execute. So think about how you use something like ChatGPT today. You ask it to write something, summarize something, explain something. It gives you an answer. Hopefully you iterate, but then you’re done.

An agent doesn’t stop. It’s the difference between someone who writes you an email explaining what you need to do and someone who writes you an email explaining about the work they’ve already done.

So here’s what that looks like in practice. Instead of asking AI to summarize a Jira ticket, you give an agent access to Jira and GitHub and say, pull everything that changed in the last 24 hours and generate a status update. The agent AI will query Jira, it’ll read the GitHub commit history, it’ll correlate those changes, and produce a report.

And that’s not one prompt. That’s a sequence of actions across real systems. And once you have that foundation, where agents can connect to your systems, then you can start to expand on them and leverage some advanced capabilities of AI.

So not just pull everything that changed in the last 24 hours and give me a report, but review the requirements for ticket AI X5, 78. Review the code that was created in branch bug fix AI X5, 78. Analyze what requirements were met, what requirements possibly weren’t met, and specific areas that should be focused on by manual QA.

The agents sit inside your workflows. They’ll take actions on your behalf, and when you implement them properly, they can dramatically extend the capabilities and efficiencies of your team.

Lightly edited for clarity.

If you want to understand where agents fit, look for the boring parts of the job. The best place to find agent use cases is not in high-level planning; it’s in coordination work. It’s the tedious, cross-system, nobody-wants-to-do-it layer that sits between decisions and results.

Take marketing teams. A typical weekly report is not really analysis. It’s pulling numbers from analytics tools, combining it with campaign data, and then formatting it into slides or dashboards.

Same in software development. Status reporting usually means checking tickets, and reviewing recent commits, and piecing together what was actually shipped. Things like bug fixing and ADA issues, SEO issues, browser specific things, they’re all straightforward, well defined tasks, but they’re tedious and time consuming for developers and cloud.

Agents are ideal for these cases, not just writing code, but also generating summaries and PR descriptions and workflow outputs.

And it’s not just in tech. In a law firm, agents can pull case history, flag relevant precedents, and draft a brief outline before a human attorney ever opens it.

And that pattern is the same everywhere. The work that connects systems is where agents win, not replacing the decision, just handling the legwork that surrounds it.

Lightly edited for clarity.

Agents don’t replace people, they change what people are responsible for. If agents actually replace people, companies would already be running them into end to end. But that’s not what I see happening.

So let’s talk about how they’re actually being used. Take coding workflows, for example. An agent can take an issue, generate code, and open a pull request, but it doesn’t merge it into production. A human reviews it. Why? Because the hard part’s not writing the code, it’s knowing whether it’s the right code.

Does it break something else? Does it fit the architecture? Does it handle edge cases? Does it actually solve what the customer need is? And agents don’t necessarily have that full context.

So what happens is execution gets faster, but the responsibility shifts. You’ll spend less time doing the work, but more time evaluating whether that work is correct.

And that part is something most people underestimate. It’s not the job that disappears, it’s what you’re responsible for that changes.

Lightly edited for clarity.

If your process isn’t well defined, an agent’s going to make it worse, not better. The agent use cases that work all look the same. There are very clear inputs, there’s clear outputs, and the sequence of steps that someone has actually thought through.

So for example, take new weeds from the CRM, enrich them with this external data, and prepare a summary that I can use on my sales call. There’s a defined start and a defined end, and it works.

Now if you compare that to respond to customer complaints and resolve them appropriately, that might sound reasonable, but resolved and appropriately are carrying a lot of undefined weight there. They’re not instructions, they’re judgments.

What counts as resolved? Who decides what’s appropriate in what context, with what authority and for which customers? And we’ve already seen multiple cases of what happens when you skip that step.

But I recently read a Reddit thread where a small UK company put an AI chatbot on their website to handle customer questions, and a customer spent time steering the conversation until the chatbot started offering discounts. And it ended up generating an 80 discount on an 8,000 euro order.

And it’s not because the system was malicious, it’s because it was never given a clear boundary for what it was allowed to do. That’s the issue.

And this is where organizations get into trouble, reaching for agents on problems that look structured but aren’t. Agents don’t just bring order to a messy process, they run it faster.

If you want an agent to work well, the real work is upstream, defining the process clearly enough that a person could follow it without asking questions.

Lightly edited for clarity.

Most agent failures aren’t really AI problems, they’re system design problems. When agents fail, it’s usually because something around the model got something wrong, not the model itself, because the system wasn’t built carefully enough.

As a real version of this, let’s say we’re gonna build an agent to generate a weekly status report. So it will go to Jira for ticket status, and it will go to GitHub for code changes, and it will go to Teams for chats and updates.

And this seems straightforward, but in Jira, if you’ve got some tickets that are marked as done while others are marked as closed, or you’ve got tickets that just haven’t been updated properly, or in GitHub your commit messages don’t map cleanly to tickets in Jira, and in Teams, half the actual decisions are happening in threads that never get captured or monitored.

So in that case, what does our agent do? It’s still going to pull everything together, and it’s going to give you a very clean, well-written, completely misleading report because LLMs match tokens; they don’t make judgments.

The model did exactly what the system allowed it to do. The system wasn’t designed well enough.

And then in addition to that, there’s the permissions problem. If that same agent has write access, it’s not just reporting. It could update tickets, or trigger workflows, or send communications based on bad assumptions.

So you give it too much access, and it can do the wrong thing, and too little access, and it misses what matters.

And that’s why security teams are increasingly focusing on agent-specific risks—not just bad outputs, but bad actions. The model’s not usually where it breaks down; it is the system design.

Lightly edited for clarity.

When an agent makes a mistake, who’s responsible? That question should make you think about how you design agentic systems.

Agents do eliminate work, but they also create new work, and they move where that work happens. When you automate a workflow, you remove manual steps, but you create new ones.

If an agent writes daily reports across 10 different projects, someone now has to review 10 reports. You scale that up, and you built a review bottleneck that didn’t exist before.

Same problem with data. An agent’s only as good as what it connects to, garbage in, garbage out, but faster and at a higher volume.

And then there’s the accountability question when something goes wrong. So a report is off, a change gets made incorrectly. Was it the model, the data, the workflow, the permissions? And most organizations genuinely don’t know how to answer that yet.

But the teams that get this right, that design the system carefully, and put human judgment in the right places, they’re not just saving time, they’re building systems that are reliable enough to trust, and fast enough to increase their competitive advantage.

Lightly edited for clarity.

Demis Hassabis, who’s the CEO of Google DeepMind, recently stated that we are in the foothills of the Singularity. Jensen Huang, who’s the CEO of Nvidia, has stated that he thinks they’ve already achieved AGI. And AGI, if you’re not aware, is artificial general intelligence. People love to talk about AGI. It’s simple to understand. It’s the benchmark which we have: an AI that’s just as good as a person is at doing whatever.

But I think the reason that most people really like to talk about AGI is because it’s kind of scary. It’s once we have AGI, then what? Right, if I have an AI that’s as good as a person at doing a job well, then why would you hire the person for 100 times the price when you could have AI do it? And so it’s terrifying, the idea that we’ll have AGI that is as good as us at doing effectively anything. So that’s scary.

I just don’t think it matters. Like, what we have with the current technology with AI is already so powerful and transformative that whether or not we have AGI is almost beside the point.

I haven’t made videos for a month because I’ve been busy, and most of what I’m doing at work is business process consulting, business process automation, and modernization. So we are talking to companies about how to reorganize their business processes within their organization to effectively take advantage of AI. It’s what do we need to do to change the way we do work to take advantage of this technology, to be able to do more work faster and cheaper?

And this is the change that most large organizations are going through now. These are the things that they are working on. The scope and scale is massive. It’s effectively every business process in every organization can be affected by AI.

And so whether or not we get to AGI, whether or not we already have AGI, if you’re waiting for AGI to figure out the then what moment, you’re too late. We’re already at the then what moment.