I'm not afraid of AI. I know that is apparently a controversial thing to say now, but I am not.

I keep hearing people talk like the second AI gets smarter than us, the human race is over. That jump does not make sense to me. Smarter than us is not the same thing as wanting to kill us. Even the hypothetical where we become the dog and something smarter is in charge — I am not volunteering to be the dog, but that is still not extinction. You cannot describe one outcome and then sell people a completely different one because the second headline is scarier.

A lot of this feels less like analysis and more like a personality type I already know. You see a post about a new venue downtown, something that might bring people in and actually improve the place, and the first comment is about parking. That person can never see the whole thing. They have found one possible inconvenience, and now the whole idea is bad. AI has become the new venue. The doomers are in the comments asking about parking.

The surveillance panic gets me for the same reason. People act like a license-plate camera is the exact moment privacy disappeared. Meanwhile, a phone sits in the middle of their life, their money moves through systems, their identity exists in systems, and they rely on those systems every day. Even the person cosplaying as completely off the grid still needs groceries, hospitals, and an identity when the real world catches up. You are part of the system. I am part of it. AI did not start that deal.

What I really do not understand is the mood around technological advancement. I do not remember feeling this much outright hostility when smartphones appeared. Maybe I am remembering it wrong. I do not know. But it seems like a lot of people have decided that new automatically means threatening, and I think they need to grow up about it. Use the thing. Figure out where it improves your life. Find the parts that are genuinely dangerous and put boundaries around them. But stop acting like refusing to touch it is some kind of moral achievement.

I also believe AI is going to be built whether people like it or not. That is a belief, not a fact I can prove about the future. But I do not believe development stops everywhere because one group decides to sit it out. If I am right, then I want us involved. I want the data centers here. I want people here working on the models, deciding how they are trained, testing what they do, and having some say in what gets built. Burying your head does not give you control. It gives control to whoever kept building.

Then there is the ant argument. People say an AI will be so much smarter that we will be ants to it. Fine. Why does that automatically end with extermination? We created these systems. We are teaching them and trying to steer them. I do not accept that greater intelligence automatically produces something more barbaric. In my own head, I associate intelligence with being less barbaric. Maybe that instinct is wrong. I can say that plainly. What I will not do is pretend the opposite instinct is a measured certainty just because it is frightening.

The other group I want to shake a little is the one saying AI is a bubble and does not really do anything. I keep hearing people describe agentic AI as though it is still a chatbot sitting in a browser waiting for a clever question. That is not how I use it. I use it to work on files, code, workflows, plans, and actual things I want completed. It has made ideas I carried around for years feel buildable. That is my experience. It is also why I think the labs have done a poor job showing normal people what these tools can actually become when you stop treating them like a search box.

If you have not seen that side of it, do something small. Give an action-capable tool a folder that needs to be organized and ask it to help. Watch what happens. Do not start with a grand theory about the future of humanity. Start with one annoying job on your own computer. That is where the ideas begin. You stop asking whether AI is impressive and start asking what else you could hand it.

There is responsibility in that. You can ask a system to agree with everything you say. You can make yourself a very capable stooge. Or you can tell it to check you, challenge the plan, and push back before it runs with a bad idea. I prefer the second one. These systems are imperfect, like coworkers are imperfect. The fact that they can be wrong is not an argument for making them useless. It is an argument for setting them up responsibly and paying attention.

So yes, I am optimistic. I could be wrong about where this ends. Anybody pretending they cannot be wrong is the person I trust least in this conversation.

What I know from my own work is simpler: the technology is productive, it has expanded what I can build, and hiding from it would make me less capable. I believe it gets built with or without the people complaining about it. I would rather have my hands on it, configure it responsibly, and help other people see what is possible.

And if your first response is still about the parking, I do not know what to tell you.

Full episode transcript

I'm not afraid of AI, and the doomers haven't earned your fear. There are people out there right now talking like extinction is basically guaranteed — like the moment something gets smarter than us, that's it, lights out for humanity. And you just can't say that. Superintelligent does not automatically mean human extinction. It's just not the truth. Even if the outcome is that something smarter keeps us around the way we keep dogs. I'm not saying being the dog would be great—but that isn't extinction. Those are two completely different outcomes, and the doomers keep selling one as the other. And I know exactly who these people are. These are the same people who see a Facebook post about a new venue going up downtown—something that might actually revitalize the place—and their whole contribution is, ‘But where are people going to park?’ Total inability to see the bigger picture on anything, ever. They don't want an answer. They need something to complain about. That's not analysis, that's a personality type, and right now that personality type has found AI. Same crowd on the surveillance stuff — grow up, we already live with a lot of surveillance. You really think the license-plate cameras I keep seeing people melt down over are the thing that finally takes us there? We carry phones that sit in the middle of our lives. You've already agreed — at least implicitly — that a lot of your life is in the open and that you're beholden to systems you rely on. Even the people cosplaying as living off the land still go to the grocery store, still show up at the hospital, still have a Social Security number. You're part of it. AI didn't start that deal. And what I actually don't get is the mood. I don't remember it being like this when smartphones came out. There's something going on with people that turns them against technological advancement in a way that feels new to me. I don't know what to blame for it, and honestly it doesn't really matter. People need to grow up, embrace it, use it in ways that improve their lives, and stop whining. A lot of things could go wrong. We're all going to die someday. What we can do is our best with what's in front of us. Because I believe this is going to get built. That's what I talked about in yesterday's video — I don't believe it stops just because we decide not to build it here. It happens somewhere else. So the real question isn't AI or no AI. It's: do you want to be the ones in control of these things — engaged in how they're trained and what goes into them — or do you want to bury your head in the sand and hope for something that I don't think is going to happen? That's why I want the data centers built here. I want us in the room. And personally? I tend to believe this is all going to be great. People love the ant line — they'll be so smart that we'll be like ants to them. No. You can't just assume that something we created is going to totally disregard us and toss us aside. Is there a chance? Maybe. But we're the ones teaching these systems and trying to steer them. So why is destroying us automatically more likely than working with us, or helping us? Nobody ever finishes that argument. They just say ants and drop the mic. Here's my read, and I'll label it plainly: this is my belief, not a law of physics. I associate greater intelligence with being less barbaric, not more. Maybe that instinct is wrong. But I don't see why a superintelligent AI automatically turns into a murderer, especially toward the species that created it. Even if you want to argue it would see us the way we see ants, that still doesn't get you to extinction. I could be wrong about this. But the doomers could be wrong too — and they're the ones claiming certainty. Now, the people who say this stuff isn't productive — that it's a big bubble, it doesn't really do anything. I keep hearing people judge agentic AI while describing a chatbot you just talk to. No. This stuff can do stuff. In my own use, it takes action on files, code, workflows, and real technological work. It's unbelievable — truly unbelievable — and I don't think enough people know about it yet. And honestly, I don't think the labs have done a good enough job bridging that gap and showing people what's actually possible. That's what I'm here for. That's the mission. Maybe being harsh in a video isn't the best way to do it — but maybe it wakes somebody up. I'll give you my own experience: this has made my wildest ideas and dreams actually buildable. If you've ever had a business idea or something you wanted to create, open an action-capable tool and just get started. Don't limit yourself. Say what you want to do. Ask it what it can do. Tell it you want to work with it, and don't be afraid to use it. Want a concrete first job? Give it a folder on your computer that needs to be reorganized, and then watch what it can do. That's a simple, easy thing — but it flips a switch in your head: okay, what can we do from here? What else is possible? That's when the ideas start lighting up. In my work it has been unbelievably productive, and you can get more complicated from there. It standardizes, it organizes, it collaborates on your ideas — and it checks you, if you set it up to. That last part matters. I hear people talk about unhealthy relationships with AI as though the machine alone is the whole story. I don't buy that. If you have control over a thing, you have a responsibility not to push it into something unhealthy — and you do have control over how you ask it to work with you. You can ask it to agree with everything you say and be your stooge. Or you can tell it to check you and push back on bad ideas. It's not perfect — it's an imperfect system, like a coworker. But the setup is yours, and so is the responsibility. So here's where I land. I believe it's going to get built — here or somewhere else — and we're going to have to learn to live with it. I tend to believe living with it is going to be wonderful. Not because I'm naive about risk — a lot of things could go wrong. But burying your head in the sand is not a plan, and standing in the comment section complaining about parking while the venue goes up has never built anything. Get your hands on this stuff. Steer it. Set it up to push back on you. And stop whining.

Sources

https://theaiinsider.tech/2026/07/06/ai-safety-expert-roman-yampolskiy-believes-ai-has-a-99-9-chance-of-wiping-out-humanity/