A Founder's Perspective on the Technology That Changes Everything

Dean Mazlish Founder & CEO, Bridge AI Consulting March 2026

The Bus Ride That Changed Everything
The Hype and the Skeptics
What AI Actually Is (Without the Hype)
How AI Changes Business
The Job Question (Let's Be Honest)
Schools, Kids, and AI Literacy
The Future of Human Creativity
The Risks Nobody Talks About
The AGI Conversation Is Already Outdated
How I Actually Use AI Every Day
Where We Go From Here

Part 1: The Bus Ride That Changed Everything

December 3rd, 2022. I was on a bus, coming home from a basketball game we'd just lost. The mood was sour. My teammates had their heads back, earbuds in, doing the thing you do when your team just got beat and you don't really want to talk about it.

But somewhere along the dark stretch of highway, I opened my phone, tapped on a link I saw everywhere on my twitter feed, and typed a question into a chatbot that had launched three days earlier.

That chatbot was ChatGPT. I was among the first million users. And within about ninety seconds of using it, I started tearing up.

Not sad crying. The other kind. The kind where something hits you so hard and so fast that your body doesn't know what to do except leak. It was excitement and awe and the unmistakable feeling of shit, everything is about to change. My ah-ha moment.

I couldn't keep it to myself. I grabbed my friend sitting next to me and shoved my phone in his face, because when something hits you like that, you need someone else to see it too. And he did. For the next five and a half hours, we didn't sleep and we didn't mope. We didn't talk about the game. We asked this thing to write rap songs about obscure topics. We had it draft research papers. Write us stories. I even fed it parts of my computer science project that was due the following week, and it just... did it. Not perfectly, but well enough to make my jaw drop.

One of my first ChatGPT conversations — a Harry Potter text adventure on that late-night bus ride

From that moment, I was hooked. I spent the rest of my senior year at Grinnell College testing the limits, playing with every feature, and showing it to every single person I cared about. Not because I wanted to be the "AI guy." Because I genuinely felt like I'd stumbled onto something that everybody needed to see, and I wanted the people in my life to share in the awe.

That feeling has never gone away. And it led me, eventually, to start Bridge AI Consulting, where I now spend my days helping businesses understand and implement the technology that made me cry on a bus more than three years ago.

But I'm getting ahead of myself. Let me back up.

Part 2: The Hype and the Skeptics

Fast forward two years from that bus ride. I'd been promoted to be a leader on AI initiatives for the Integration Engineering division at Epic Systems, the largest healthcare software company in the world. Our division had about 500 people, and a big part of my job was essentially to figure out how AI could make all of them better at theirs.

Now, I want to be honest about something: I didn't get that role because I was the most technically brilliant person on the team. I got it because I was the most fucking obsessed. I was the person who wouldn't shut up about AI, who was constantly testing new tools, who kept finding little ways to shave hours off repetitive tasks. And in those early days, that mattered more than knowing the math behind transformer architectures.

A big part of my job was sitting down with people one-on-one and talking to them about why and how to use AI in their work. And this is where I learned the most important lesson of my career so far.

Most people are not on the AI bus the way I am.

I don't mean they haven't tried it. I mean they're standing at the bus stop with their arms crossed, suspicious of the driver, and not entirely sure the bus goes where they want to go.

Up until that point, I'd been living in what I'll call AI Hype World. If you spend any time on X (formerly Twitter) or certain corners of LinkedIn, you know this world well. It's a place where every day brings a breathless announcement about some new model that's 10x better than the last one, where founders post screenshots of revenue dashboards they built in an afternoon, and where the general vibe is that if you're not already using AI for everything, you're basically a caveman. (Side note – I love The Rundown AI Newsletter as a way to stay up-to-date.)

Inside the AI hype world, everyone is running through a field of lilies into the sunset, holding virtual hands, singing about the future.

Outside the hype world? It's a very different picture.

Some people were worried about the environment (training these models uses enormous amounts of energy). Some were worried about their jobs. Some just didn't want to support big tech companies getting bigger. And some, honestly, were just afraid of change in a way that's completely human and completely understandable.

It wasn't that everyone was against it. It was more like a bell curve. On one end, you had people like me, the true believers who couldn't shut up about it. On the other end, the firm skeptics who wanted nothing to do with it. But the biggest group, the fat middle of the curve, were people who were interested but cautious. Excited but unsure. Not on the bus, not walking the other direction, just kind of standing at the stop wondering if it was safe to board. And among the cautious middle and the skeptics, the most frustrating pattern I saw was this:

Someone who was already on the fence would give AI a shot. They'd ask it something it wasn't really built for, or throw it a task that even a human expert would struggle with on the first try, and when it didn't nail it, that was it. Case closed. 'Yeah, I tried AI. It sucks.' One bad experience and they'd written off the most powerful tool of their lifetime.

This drove me crazy, and it still does. Because you wouldn't pick up a guitar for the first time, strum a faulty chord, and say "Music sucks." You wouldn't open Excel, stare at an incorrect formula, and conclude that spreadsheets are useless. Tools take time to learn. AI is the most powerful tool most people will ever have access to, and too many people are judging it on a single bad first date.

That experience at Epic taught me something that became the foundation of Bridge AI Consulting: prioritizing people always comes first when understanding how AI can help them. There isn't a one-size-fits-all way to use AI, and if you don't meet people where they are, you'll never get them to where AI can take them.

Everyone's aha moment comes at a different time. My dad's came just a few weeks ago when Claude's Opus 4.6 built him a perfect spreadsheet in seconds. Mine was rap songs on a bus at 11pm. A client's was watching AI generated videos. The trigger is different for everyone, but that feeling of 'oh shit, this is real' is universal. Part of my job at Bridge AI is helping people find their aha moment faster. That realization led me to think hard about how to actually explain this stuff. Because if I was going to meet people where they were, I needed to be able to describe what AI is and what it isn't in a way that didn't make their eyes glaze over. So here's my attempt.

Part 3: What AI Actually Is (Without the Hype)

Let's get the technical stuff out of the way, because I think one of the biggest reasons people misunderstand AI is that they've never had it explained to them in plain language.

AI as a broad term has been around for decades. Your spam filter is AI. Netflix recommendations are AI. Siri is AI. However, when most people say 'AI' right now, they're talking about generative AI. Generative AI is the branch of artificial intelligence that creates things: text, images, video, code, music. It's what powers ChatGPT, Claude, Gemini, and the rest of the tools you keep hearing about.

At its core, here's what generative AI is: advanced pattern recognition.

That's it. I know that sounds reductive, and yes, the engineering behind it is complex. But the fundamental concept is not. These systems have been trained on enormous amounts of human-created content, text from the internet, books, code, images, and they've learned the patterns in that content so well that they can generate new content that follows those same patterns.

When you ask ChatGPT a question, it's not "thinking" the way you or I think. It's predicting, with remarkable accuracy, what the most helpful response should look like based on patterns it has learned from billions of examples. It's like having a conversation with someone who has read every book ever written and has a supernaturally good memory, but who has never actually experienced anything. They can tell you what a sunset looks like because they've read ten thousand descriptions of sunsets. But they've never watched one.

This distinction matters because it explains both why AI is incredibly useful and why it sometimes gets things spectacularly wrong.

The Magic Wand Problem

Here's the single biggest mistake I saw people make with AI in those early days, and honestly I still see it every single day: they treat it like a magic wand instead of a tool. Wave it around, get pissed when nothing happens, and blame the wand. (Think Ron Weasley with Wingardium Leviosa)

A tool, you have to learn how to use. You have to understand what it's good at, what it's bad at, and what it was never designed to do. A hammer is amazing for nails and terrible for screws, and nobody blames the hammer for that.

A hammer is amazing for nails and terrible for screws — and nobody blames the hammer for that.

Now, I'll push back on my own analogy. Nobody ever marketed a hammer as a tool that can do literally anything. But that is exactly how AI has been presented. The hype machine, the breathless posts, the commercials, they've all sold AI as this omnipotent thing. So when someone tries it and it can't do the one thing they needed? I get the frustration. The marketing and hype lied to you. The tool didn't. AI is incredibly powerful and getting more powerful every day, but it is not yet a do-anything tool. You have to know what it's good at, how to talk to it, and where its limits are. That's a big part of what we do at Bridge AI: cut through the hype and help you understand what AI can actually do for your business right now.

Even those early large language models were spectacularly good at synthesizing information, drafting content, brainstorming ideas, writing and debugging code, summarizing large documents, translating between languages, and explaining complex topics in simple ways. But they were mediocre to bad at precise mathematical calculations (they process language, not numbers, so math is like asking a writer to do plumbing), real-time information retrieval, consistently following long and complex multi-step instructions, and anything requiring a genuine understanding of physical reality. (Today's models are a different story. We'll get there.)

The people who got the most out of AI back then weren't the most technically sophisticated. They were the ones who took the time to learn where the tool shined and where it didn't, and who developed a feel for how to talk to it. It's kind of like being good at approaching a stranger at a bar. You've got to have confidence. You've got to know how to talk smooth. You've got to read the room. And that only comes with practice.

Those people are a step ahead today. Because even now, when AI has become incredibly powerful and it inches closer to a magic wand with every update, you still need to know how to set it up correctly, how to prompt it well, and how to understand its capabilities. The limitations are rapidly shrinking, but they're real. And people who know how to use AI get dramatically better results than people who don't, not because the tool is different, but because the operator is. That's exactly what we do at Bridge AI Consulting: we speed up that learning curve for you so you're not fumbling through it alone.

On Hallucinations

You've probably heard that AI "hallucinates," meaning it sometimes makes things up and presents them as fact. This was a big problem in the early days. It's a much smaller problem now because models have gotten significantly better. But it hasn't gone away entirely, and I don't think it ever fully will.

Here's how I think about it: treat AI output the way you should treat anything you read on the internet. Most of the time, it's accurate. Sometimes, especially for well-established facts, you can be 99% confident. But for anything important, anything you'd bet money or your reputation on, you verify. You don't blindly trust a Wikipedia article, and you shouldn't blindly trust an AI either.

The people who know how to use AI get significantly fewer hallucinations than those who don't. Not because they have access to a better version of the tool, but because they understand what it's good at and what to double-check. Most of that just comes with practice. The responsibility is on you to understand the limitations and use the tool accordingly. That's not a flaw of AI. That's just what responsible use of any powerful tool looks like.

The Closed vs. Open Source Debate

There's an ongoing debate in the AI world about whether the most powerful models should be open-source (freely available to anyone, like Meta's Llama) or closed-source (controlled by companies like OpenAI and Anthropic).

I'll be direct: I lean toward closed-source for the most powerful models. And I know that's not the popular opinion in many circles.

My reasoning is simple. I don't fully trust big corporations to always act in humanity's best interest. But I trust them more than I trust the entire population. Open-sourcing the most capable AI models means giving every person on earth, including malicious actors, access to technology that could do real harm. There's no perfect solution here, but given the genuine existential questions AI raises, I'd rather bet on a few companies treating the technology carefully than guarantee that bad actors have unrestricted access. Sure, I don't trust the companies either and the potential for negatives is still huge (think weaponized misinformation at scale, autonomous systems making decisions without any human oversight, or surveillance infrastructure that makes today's look quaint) but it is our only real option here. As we get closer to the Magic Wand I want truly malicious actors nowhere near it.

If I had to pick a company I trust most on safety, it's Anthropic. Their CEO Dario Amodei has been the most thoughtful and transparent voice on AI safety, and their recent decisions, like declining military applications, suggest they're walking the walk. That's part of why Claude is my daily driver (They also just kick ass). Between growing my own technical understanding, the people skills, and the obsessive tinkering, eventually I came to the same question: how do you actually help a real business use this stuff? That's what led me to start Bridge AI.

Part 4: How AI Changes Business

Mark Cuban said it best in a recent interview: "There are millions of companies that have one, five, 10, 50, 100, 500 people that aren't going to have AI budgets, that aren't going to have AI experts." He's right. Every company needs an AI team right now. But most can't afford $1,000/hour consultants, and most don't need a full-time AI hire. They need someone who can walk in, understand their business, and show them what's possible. That's exactly why Bridge AI Consulting exists. We're the AI team for businesses that need one but can't justify building one in-house.

I talk to small business owners every single day. And the most common thing I hear, in various forms, is some version of: "I know AI is important, but I don't know where to start."

My answer is always the same: start by using it yourself.

Open Claude (or your preferred model). Pick a task you do every week that's tedious, time-consuming, or just annoying. An email you have to write, a report you have to compile, a spreadsheet you have to update. And just... have a conversation with the AI about it. Don't expect it to solve everything on the first try (it still might). But explore. Play with it. Get a feel for what it can do.

At Bridge AI Consulting, when we work with a new client, we don't walk in and say "Here's your AI strategy." We start by understanding their workflows, their pain points, their team. Because the right AI solution for a five-person marketing agency is completely different from the right solution for a twenty-person accounting firm, and both are different from what works for a solo consultant.

Sometimes the solution is as simple as showing someone how to write a better prompt. Sometimes it means building a custom AI agent, a piece of software that can take actions on its own based on rules you set, that automates an entire workflow. And sometimes, honestly, the answer is that AI isn't the right tool for that particular problem, and we help build a traditional solution instead.

The ROI Question

One thing I'm passionate about is not building AI tools just because they're cool. Every solution should be tied to a return on investment.

The framework is straightforward: estimate the time a task currently takes, multiply by the average cost of the person doing it, and compare that to the cost of the AI solution (including development, ongoing API costs, and maintenance). If the math doesn't work, don't build it. If the math does work, build it fast because your competitors will.

Don't create AI tools because they're trendy (if you are using AI to create your Studio Ghibli-style profile photo and calling it a business strategy, you are doing it wrong). Create them because they save money, improve quality, or let you do things that were previously impossible.

The Small Business Opportunity

Here's something I find genuinely exciting: I think AI could bring back the age of the small business.

For the last few decades, scale has been king. Big companies could afford better software, bigger teams, more data, more everything. Small businesses were often stuck competing with one hand tied behind their back.

AI changes that equation. A two-person company with the right AI tools can now produce marketing content, analyze financial data, handle customer support, and manage projects at a level that used to require a team of twenty. AI reduces costs for businesses of all sizes, but proportionally, it helps small businesses more because it can be the difference between profit and loss. A Fortune 500 company saves some money. A small business becomes viable.

That's not hype. I've seen it. At Bridge AI, we've helped businesses go from struggling with manual processes to running lean, AI-augmented operations that freed up their founders to actually focus on growth instead of drowning in busywork. From marketing to business plans to finances to product development, there isn't an area where AI doesn't help when it's implemented thoughtfully. But I'd be doing you a disservice if I only told the optimistic side. Because the same technology that's leveling the playing field for small businesses is also changing the rules for your average employee.

Part 5: The Job Question (Let's Be Honest)

This is the part of the AI conversation where people get uncomfortable, so let me just say it plainly: AI is going to eliminate some jobs. That is not a maybe. It's already happening and pretending otherwise is sticking your head in the sand.

Junior software engineering roles are the canary in the coal mine. Right now, one experienced developer with good AI skills can do the work that used to require a small team of junior engineers. You don't need five people writing boilerplate code when one person can prompt an AI to generate, test, and debug it in a fraction of the time. You just need one person who knows how to architect solutions and manage AI output, essentially a director rather than a line worker.

Following software engineering, the white-collar workforce more broadly is going to come under pressure. Data entry, basic analysis, first-draft writing, routine customer service, standard legal document review, entry-level financial modeling... these are all tasks that AI either already does well or will do well within the next few months. If your job consists primarily of processing information in predictable ways, the ground underneath you is shifting.

This isn't just theoretical. In early February 2026, Anthropic released new Claude Cowork plugins for legal, finance, and sales, and the stock market lost roughly $285 billion in a single day. Thomson Reuters and LegalZoom dropped over 15%. Intuit, PayPal, Salesforce, all got hammered. Cybersecurity stocks tanked when Claude added vulnerability scanning. Sector after sector, the market looked at what AI could now do and panicked. That's not irrational fear. That's investors doing math and realizing the economics of entire industries just changed overnight. Anthropic's own CEO Dario Amodei has warned that AI could displace half of all entry-level white-collar jobs in the next one to five years. I don't say that to scare you. I say it because the people who see this coming and adapt are going to be fine. The people who pretend it isn't happening are the ones who should be worried.

Now, here's what I want to say to anyone reading this who is feeling a knot in their stomach: learn how to use the tools.

Humans have always adapted to new technology. We adapted to the printing press, the assembly line, the personal computer, the internet. Each of these disruptions killed some jobs and created others that didn't exist before. AI will do the same. And if the jobs don't come back this time? Well, you probably weren't going to be the person solving mass unemployment anyway. But if you are that person, and you are reading this right now, please do a good job (Looking at you Zach).

But the transition won't be painless, and I'd be lying if I said otherwise. In the short term, I'm genuinely concerned about job displacement and a widening wealth gap. The benefits of AI are going to accrue faster to people and companies who adopt it, which means the early divide could be stark.

In the long run, I'm optimistic that AI increases the mean quality of life for humanity. What that looks like and how long it takes, I'm not sure. But I believe it.

In the meantime, my advice is specific and actionable: stop being a person who "does things they're told" and start being the person who is good at explaining a task, who truly understands a task, and who can write it down clearly enough for an AI to execute it. If you can translate goals into instructions and learn how to use the latest tools, you will thrive in the AI era. That is the single most important career skill of the next decade. In short... don't be a sheep (sorry Mom – nothing against real sheep).

Part 6: Schools, Kids, and AI Literacy

If there is one topic where I could write an entire separate paper, it's AI literacy in education. When I say AI literacy, I mean understanding what AI can and can't do, knowing how to use it as a thinking partner rather than a replacement for thinking, and being able to evaluate whether what it gives you is actually good. It's the difference between someone who uses a calculator and still understands math, and someone who can't add without one. I'll try to be concise, but understand that I feel strongly about this.

Banning AI in schools is one of the worst ideas I've heard in education. It's like banning calculators in the 1980s. You can do it, and some schools did, and then those students entered a world where everyone used calculators and they were behind. Thankfully, the majority of schools I have talked to are at least taking an open approach to AI.

I don't think giving students free rein with AI tools and changing nothing else about education is the answer either. The answer is teaching AI literacy, and it needs to be thoughtful and intentional.

Here's an example of what I think smart AI integration looks like. A teacher assigns an essay on, say, 18th century France. Students are told they can use any AI tools they want. There are no restrictions. But then, after they submit their essays, the teacher uses AI to create a personalized exam for each student based on what they wrote. And the students take that exam in person, with no tools.

See what this does? It incentivizes students to actually learn the material, because they know they'll be tested on it in a way that's tailored to their specific essay. It teaches them how to use AI as a tool for research and writing, which is a skill they'll need. And it shows teachers how to use AI to personalize and improve learning, creating individual assessments instead of one-size-fits-all tests.

The bigger picture here is about critical thinking. AI has the potential to enhance critical thinking skills or to destroy them. It depends entirely on how it's taught. If kids learn to use AI as a thinking partner, a tool that helps them explore ideas, bring them to life and test arguments, it's incredible. If they learn to use it as a copy-paste machine that does their thinking for them, it's a disaster.

Copy-paste AI: using it without thinking is the real risk.

Countries that invest heavily in AI literacy over the next ten to twenty years will produce citizens who think more clearly, work more effectively, and contribute more to their economies. Countries that don't will produce a generation that's dependent on tools they don't understand. The stakes are extremely high.

As for college: I still think a degree is valuable, mostly for the relationships you build and for learning how to learn. But that value depends enormously on how colleges adapt. A university that's still teaching the way it taught in 2015 is going to produce graduates who are dangerously unprepared for the world they're entering (but then again don't they already do that? Who leaves college understanding taxes... or debt).

If I had kids today, I'd want them to learn in conjunction with AI from an early age. I'd want them to develop two skills above all others: creativity and the ability to explain their ideas clearly (hmm maybe the English major make a comeback for the ages). In a world where AI can execute, the people who can imagine and communicate will be the ones who lead. And that brings me to something I think about a lot: what happens to the people whose entire job is imagining and communicating? The artists, the writers, the creators?

Part 7: The Future of Human Creativity

Here's a prediction that might surprise you coming from someone who runs an AI company: I think art and artists are going to become more important than ever.

Let me explain.

AI can generate a painting in seconds. It can compose music, write poetry, produce video. The technical skill of creation, the ability to actually make the thing, is being rapidly democratized. Which means the craft of creation is becoming less scarce. Anyone with a laptop can now produce something that looks professional (or even write a full book with Claude Cowork).

However, I think we're heading toward a world where knowing that something was human-created becomes a selling point. Not for everything. I don't particularly care if the marketing video for my hotel was made with AI tools (it was, and it's great). I don't care if the software I use was written by AI.

But for art that's meant to be art? For books, paintings, music, film? I think human authenticity is going to matter more, not less. And I think verification tools that can distinguish human-created from AI-created work will become increasingly important. I want to physically stand there and watch my favorite author write every single word (c'mon George it has been 15 years).

The concept I keep coming back to is the "artist architect." With AI tools putting high-quality production capabilities in the hands of more people, what will separate the extraordinary from the mediocre isn't technical skill. It's creative vision. Having big money may not matter if everyone can use tools to make high-quality content. What will matter is creativity and the ability to tell an amazing story.

True creatives and special talents should win out. And I'd love to run experiments to test this. We are almost fully at the point where people can't tell AI-generated art from human art in a blind test, but I would wager that the majority of people will care deeply about the distinction once they know. If that's true, it means authenticity becomes its own form of art.

Should AI-generated content be labeled? For art, absolutely. For business content like marketing materials? That's less clear to me. Art for art and video for marketing are different things with different expectations, and I think the norms around labeling will evolve differently for each. The bottom line is this: AI makes it easier than ever to create. But creating something that actually matters, something that moves people, that still requires a human with something to say. The tools are democratized. The vision isn't.

Part 8: The Risks Nobody Talks About

I'm an AI optimist. I want to be clear about that. But being an optimist doesn't mean being naive, and there are risks that I think most people don't understand well enough (people inside and outside the AI hype world fit this category).

The Everyday Risk: Copy-Paste AI

The risk that keeps me up at night isn't some sci-fi scenario where robots take over. It's something much more mundane and much more likely: people slowly losing their ability to think critically because they've outsourced their thinking to AI.

I call it "copy-paste AI." It's when someone asks an AI a question, takes whatever comes back, and uses it without thinking. Without checking. Without engaging their own brain to evaluate whether the output is good, accurate, or appropriate.

This is already happening, and it's not just lazy students. It's professionals sending AI-generated emails they didn't proofread. It's business owners making decisions based on AI analysis they didn't verify. It's people gradually atrophying the very mental muscles that make them valuable.

AI has the ability to be genuinely negative for you if you don't understand the risks and don't know how to use it properly. Over-reliance leads to dependency. Dependency leads to declining skill. And declining skill in a world that's changing this fast is a dangerous place to be. I will say again... no I will shout AI LITERACY, AI LITERACY, AI LITERACY.

Deepfakes and Misinformation

The ability to generate realistic fake content, whether it's a video of a politician saying something they never said or a phone call in someone's voice, is genuinely scary. It's not hypothetical. It's happening right now.

Like any powerful technology, AI comes with the potential for both tremendous benefit and tremendous harm. The answer isn't to stop developing the technology. The answer is to stay vigilant, to build detection tools, and to create and enforce rules and regulations around AI use. Will it be perfect? No. There will always be malicious actors, just as there are malicious actors with every technology we've ever invented. But we have to work hard to be just and hold accountable those who would use AI to deceive and harm.

The Existential Question

People sometimes ask me if I think AI could be truly dangerous on an existential level. Here's my honest answer.

The scenarios that worry serious researchers aren't Terminator. They're AI systems optimizing for goals that don't align with human wellbeing. That can take many different shapes and yes, it does mean there is potential for very bad things to happen. But the same is true with nuclear weapons, and after a horrible start, we've managed that risk imperfectly but functionally (without ending the world) for eighty years. AI safety research is extremely important, and I think the people and organizations doing that work (Anthropic chief among them) are spending their time on one of the most important problems in human history.

My honest guess is that in the long run, the net benefit of AI will be positive for humanity. AI will eventually become self-improving and will advance in ways that are hard to predict. But I don't think that means the end of the world or vast negatives beyond our control.

And if I'm wrong about that, if things do go sideways at a civilizational level, then my prediction probably doesn't matter anyway, and there's likely nothing realistic any of us could have done to stop it. So, we may as well be optimistic and work hard to make that optimism justified.

That said, optimism without action is just wishful thinking. We should be optimistic and we should try our hardest to ensure that optimism comes true. We may not be able to prevent every worst-case scenario, but we have enormous control over how much unnecessary negative we have to endure along the way. A lot of the existential worry I just described tends to center on one concept: AGI. So let's talk about it.

Part 9: The AGI Conversation Is Already Outdated

Everyone wants to know: when is AGI coming? Artificial general intelligence, the point where AI can do anything a human can do.

To understand the debate, it helps to know the spectrum. Narrow AI is what we've had for years: spam filters, recommendation algorithms, voice assistants. It does one thing well. General intelligence, AGI, would mean AI that can do anything a human can across every domain. And beyond that, some researchers talk about superintelligence, AI that surpasses human capability entirely.

Here's my probably-unpopular take: there will never be an "AGI Day."

The definition of AGI is too fuzzy, the goalposts keep moving, and the reality is that we're already in the process. AI is already better than most humans at many cognitive tasks. It's still worse than humans at other things. That gap will keep narrowing, and it will narrow exponentially but to us it will still feel gradual.

I think even 1 year from now, AI will be dramatically more capable than it is today, in ways that would shock us if we could see them now. But most people won't feel like their lives have changed drastically day to day, because the change is incremental. It's like how the internet transformed everything about modern life, but there was no single day when you woke up and said "Oh, the internet has officially changed everything." It just... happened. Gradually, and then all at once.

If you asked the average person today whether their life has changed dramatically because of AI, most would probably say no. Ask them again in 2027. They'll probably still say no, even though by then AI will be writing significant portions of the world's code, handling most routine customer service, tutoring children, and managing complex business operations. The change will have been enormous (and it is already underway). It just won't have felt that way while it was happening.

AI will continue to get smarter. It will continue to do more. Humans will continue to be better at certain things. And the debate over whether we've hit "true AGI" will probably rage on forever, even as AI systems routinely do things that would have been considered science fiction a decade ago.

Part 10: How I Actually Use AI Every Day

I want to pull back the curtain on my own AI use, because I think seeing how someone actually integrates these tools into their daily work is more useful than abstract advice.

My current favorite tools are the suite that Anthropic provides: Claude Chat for thinking through problems, Claude Code for building software, and Claude Cowork for everything else. Together, they let me take ideas from concept to working prototype in hours. Not days, not weeks. Hours. For example, I had Claude Code build the Bridge AI Consulting website, CRM and application tracking system and Claude Cowork organized my folders beautifully before migrating over to Sharepoint.

But here's the critical caveat: I am the architect and director. AI doesn't run my business. I run my business, with AI as my most capable tool (Although if you are interested check out Felix the AI agent running his own business). I need to know what I'm doing, what I'm asking for, and whether what I'm getting back is good. AI amplifies my abilities. It doesn't replace my judgment.

At Bridge AI Consulting, we help businesses across the full spectrum. From marketing to business plans to finances to product development, there isn't an area where AI doesn't help when implemented correctly. Sometimes the solution is as simple as writing a good prompt. Sometimes you need to build a separate AI tool, a traditional software tool, or a full AI agent. All of it is available in mere moments to companies that know how to use it or know someone who does.

Here are two real examples. We worked with a basketball scouting startup (Basketball Scout AI) that wanted AI at the center of everything they do. That meant we were embedded across the entire business. On the internal side, we worked with their sales lead to build AI-powered workflows for outreach and lead tracking. They now utilize a workflow including AI note-taking and Claude that dramatically increases efficiency and quality of follow-ups, action items, and lead tracking after every meeting. That alone saved him hours a week. On the product side, we helped them build out the actual AI scouting platform, working hands-on with Claude, Gemini, and OpenAI APIs to design the backend that powers their core product. We didn't advise from the outside. We became their AI team, helping them save time internally while building the thing they're selling to customers.

On a completely different end of the spectrum, we worked with a walking cane company (Walking Canes). No AI scouting platform, no API integrations. They needed help with marketing and growth. We created over 100 AI-generated product videos for their canes, dramatically increasing their content output without cutting quality or blowing their budget. When they wanted to launch a new offshoot website for their sword cane line, we built that too, using tools like Claude Code to get it working exactly the way their team needed. Every solution was different because every problem was different, but the approach was always the same: figure out what actually helps, and then go do it. That's the range. A tech-forward startup and a traditional product company, both getting exactly what they needed because we started with their business, not with the technology.

Three years ago, the things I do daily would have cost tens of thousands of dollars or required a team of specialists. Today, one person with the right tools and the right knowledge can accomplish what used to take a department. That's not an exaggeration. That's my Tuesday.

Part 11: Where We Go From Here

I started this paper on a bus. So let me end it there.

On December 3rd, 2022, I had a feeling that everything was going to change. Three years later, I can tell you: it has. Not in the dramatic, Hollywood, robots-walking-the-streets way. In a quieter, deeper, more fundamental way. The tools we use to think, to create, to build, to communicate, to work, they are all different now. And they're going to keep changing.

I am optimistic about AI. I'm optimistic that it will help more people than it hurts, that it will create more than it destroys, that it will, on balance, make the world better. But I hold that optimism alongside a genuine respect for the risks, the turbulence, and the hard work it will take to get the good outcomes and avoid the bad ones.

If you're a business owner or future business owner reading this, here's what I want you to hear: now is your opportunity. Being a small business is hard. It takes long hours, random problem-solving, and usually a lot of inefficient processes that exist because you never had the time or money to fix them. In the age of AI, you don't have to accept "okay." You can increase your productivity and quality in ways that would have been unimaginable three years ago.

But you have to learn how to use it right. Not just "use AI" in the vague sense. Learn which tools are right for your specific business. Understand what they can and can't do. Build workflows that actually save time and improve quality. Don't just chase the hype.

That's what we do at Bridge AI Consulting. We don't sell AI for the sake of AI. We sit down with you, understand your business, and figure out where AI genuinely makes a difference (or where you need a deterministic solution instead). Sometimes the answer is a simple prompt template that saves you five hours a week. Sometimes it's a custom-built system that transforms how you operate. Either way, we start with you and your business, not with the technology.

AI is here. It's real. It's powerful. It can help you, and it can hurt you if you don't know what you're doing.

Educate yourself. Be optimistic. Be careful. Be excited.

From a Bus Ride to a Business: A Founder's Thoughts on AI

Table of Contents