@tjrobertson52

Has anyone else noticed that we've had AI that's about as good as humans at a lot of things for a while now? Almost no reputable companies are actually automating much with AI. And why is that?

Well, I think it's because we're learning that about as good as a human isn't good enough if it's not at least as good of a human. It just feels like AI. It feels cheap, feels impersonal.

As AI just becomes more capable and people become more aware of what AI can do, our expectations go up. We have higher expectations for our software and higher expectations for humans. What seemed magical two years ago is just standard today.

So your vibe coded SAS product that can do an SEO audit of my website almost as well as a human can or can evaluate my tax return almost as well as a real accountant has effectively seen its value drop to zero.

We've reached a point where the only AI tools or processes that have any value are those that can do the job better than a human. And those do exist. We're building them at my agency.

And I know a lot of other companies are doing the same thing, but it's a lot harder.

6, people are realising that there's a lot more applications where AI can now do a better job than humans. And as Claude continues to ship more than one new feature every day, and models keep improving it, that's just gonna keep expanding.

So even though it is really hard to build these tools, you need someone with real domain expertise. You need to iterate on it for many, many hours. It's well worth it, and companies are noticing that, which means there's going to be a ton of competition.

And our standards for what an AI tool needs to do to have any value are going to continue to go up very quickly.

So we're past the days where you can just vibe code a cool project over the weekend and impress everyone. But if you have real domain expertise and you're willing to put in the time, 2026 seems to be the inflection point.