The AI Window is Now

I had about twelve hours to kill in Vancouver yesterday. I went to my favorite tea place there (Vancha), played some pinball, and then went to a movie for the first time in five years or so. I saw the new F1 movie. I didn’t know anything about F1, and I don’t want to give any spoilers, but the big thing I took away from the movie was that even though the overall car and driver matter, there are only certain moments in the race when it’s really possible to jump ahead and move up from the back of the pack to the middle, or from the middle to the front.

My next sentence was originally, “now is one of those times for business”, but the opportunity is actually much broader than that. Now is one of those times for… almost anything.

Right now you can multiply your individual effort by at least 10X in many different fields, and very few people actually realize this or are doing it. You can get ahead right now, but that window will close over the next year as everyone one else catches on (or hires/subcontracts people who have caught on).

A few months ago when cohesive good quality video generation first made it to my feeds (and I assume I was late to it since I don’t really care about video generation), I called a friend in a panic. In high school he made a pretty awesome documentary about us camping out for Star Wars, he went to USC to study film, got an internship at a major TV show, etc. Then… nothing, even though he always wanted to make movies.

Hey, I said, you can make your own movie now! He’s technical, so he could easily figure out how to make an entire movie, or at least a short. Invest a little bit of money and a lot of time, and he could have made the first AI movie! The bar for making a movie now is ostensibly pretty high… but the bar for making an AI movie is probably pretty low. I’d make a movie, release it everywhere online for free, hype up the fact that it’s the first fully AI movie… and now you are a famous filmmaker. What doors does that open?

Almost exactly one month ago I tried Claude Code for the first time. It’s software that can interact with your computer and your codebase to complete tasks for you all on its own. I typed in something like, “Make a visualization of a cruise ship’s path over the year”. I figured it would ask me some questions and then make a table that showed what port the ship was in each day. Instead it found the credentials to my database, went in and looked at my data, and built a site where you could choose one of the ships and it displayed its path on a map with a color-coded trail and detailed information below. My jaw literally dropped and I just kept saying, “Oh my god…”. Even being technically savvy and already using AI for some coding, I had no idea it was this good.

Since then I have made the same amount of progress that a dedicated Tynan would have made in 1-2 years, and probably more progress than I’ve actually made in the past 8 years. 

One month ago, CruiseSheet was still running on bootstrap and jquery. I had been wanting to move to tailwind and native javascript, but it was a daunting task. If you don’t understand what that means, the gist is recoding every single part of the site that the user sees. The few times I started ended up being so frustrating that I tabled the project. With Claude I completed it in a week and also redesigned almost every single page in the process. Oh, and while I was at it I optimized every aspect of the site for both SEO and speed. 

Not only would that one job take me months… I just didn’t want to do it. It’s a slog, won’t immediately improve the business, and frankly I don’t have to do it so I didn’t. That’s one of the magic parts of AI— it does the parts that you really don’t want to do, so work feels like only doing the parts that you enjoy. I’ve worked 12-16 hours most days over the past month but my memory of that is mostly watching YouTube videos, browsing the web, hanging out with my friend on a cruise, and checking in on Claude every 5-10 minutes. I’ve also spent hours just thinking about what I want to build, rather than worrying about how difficult it would be.

Work is also much more satisfying because every single day CruiseSheet can really become whatever I want it to be. I’ve always wanted to have a port guide that I would actually use… three days later I had one. I wanted to make a cool visualizer to see where all of the cruise ships were at any given time. Done. I can’t even remember all of the things I’ve done in the past month because I’m moving so fast. Maybe more than anything— I’m proud of the site again. I’ve always been proud of making a site that shows the best deals, but the look and feel and performance of the site had become dated and atrophied over the years. Not any more!

I’m not too focused on competition or even on making money, but… has any other cruise site made this much progress in the past month? I doubt it. And I’m going to keep working on it every day because there’s really nothing I’d rather do. By the time other companies figure out that they could be doing this too… I think I will have lapped them. As I write this post Claude is working on generating deck plans for every single deck on every single cruise ship. 

I use coding as an example because it’s what I’m doing, but there are opportunities in really every field. You can generate images, video, text, and automations using n8n. You can have ChatGPT or Claude come up with strategies to use those tools. You could literally go from having an idea to making your first sale in a day.

My guess is that people reading this will fall mostly into two groups: people who are already doing it who will think, “Yeah, obviously… I’ve already been doing this for months” and people who aren’t doing it who will think, “Yeah, but its probably not actually that good”. I was in the latter camp a month ago. If you’re there, you should at least spend a day and see what’s out there. It’s wild.

Not only are these tools powerful, but they’re actually usable by normal people now. Up until a month ago every line of CruiseSheet was hand coded by me. In the past month I have maybe coded 50 lines total out of thousands of lines that have been made. Some technical understanding has been helpful, but you really don’t even need to know the code anymore. There are other tools like Lovable and Replit that require zero knowledge.

There are very few moments in a lifetime with opportunity this extraordinary. The last one was Bitcoin/crypto (which I also happened to write about very early https://tynan.com/bitcoin/), and the one before that was probably the internet boom, which also changed my life. Even if you don’t think AI can help you… you should look into it because I bet you have no idea what it’s capable of.

###

Photo is Shoshone falls in Idaho. My wife and I did a road trip from Vegas to Seattle a couple weeks ago. First time in an RV since I sold mine!

I made a bunch of blog-like posts on my X account with the idea that I might post blog posts there instead of on the blog. My first post got a ton of traction and I was encouraged… subsequent ones got less. We’ll see. Follow me there just in case.


Posted

in

by

Tags:

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *