|
Happy Wednesday,
Today I'm giving you two ChatGPT prompts that work out where in the world you should actually be living and build you an interactive map of it, then a quick word on Fable 5.1 and the model OpenAI is too scared to release, and finishing with three image codes that do genuinely ridiculous things to your photos.
Let's get into it.
|
IN THIS EMAIL
» Where you should actually live
» Fable 5.1, and the model OpenAI won't ship
» 3 image codes worth saving
|
🧠 DO THIS TODAY
ChatGPT Can Now Tell You Where You Should Actually Live
Almost nobody picked where they live. You were born there, or a job moved you, or you followed somebody, or the rent was cheap that year and fifteen years went by. Out of every city on earth, hardly anyone has ever sat down and actually chosen. And most people have a quiet suspicion there's somewhere out there that would suit them better, without a single reason to think so.
So I figured ChatGPT might be able to help. I put together these two prompts. The first one asks you what time and where you were born, then works out your real chart for that exact moment using astrocartography. The second one takes that and builds you an interactive map of the world with your own lines drawn across it, so you can click through the cities one at a time and get a full write-up on each one, what it's supposed to bring you and whether you'd actually fit there.
It took about ten minutes and I did not expect the map to actually be good. It is.
Step 1: Open ChatGPT and paste this in. It'll ask you the questions itself.
|
Ask me for my date of birth, the time as close as I know it, and the town I was born in, then work out my actual chart from that exact moment. Do the calculation properly with code rather than from memory, and if you can't run it, tell me instead of guessing.
Then tell me what it says about me in plain English, no vague horoscope language. Go at the specific patterns rather than my star sign, be honest including the parts I won't enjoy, and tell me which parts are real astronomy and which parts are interpretation.
|
Have the time as close as you can get it. It's on your birth certificate. An hour out and half of it moves. What comes back goes at the patterns rather than the signs, and if you ask for the unflattering parts it does not hold back.
Step 2: Once it's done that, follow up with this:
|
Now work out my astrocartography lines. Build me a full interactive world map with every one of my lines drawn across it, straight vertical lines for the MC and IC and proper curves for the ASC and DSC, color-coded per planet and labeled.
Mark every major city that sits on or near a line. Make each city clickable, and when I click one, show me a full write-up: which line it sits on, what that line is supposed to bring, what the catch is, and a straight verdict on whether I'd actually fit there.
Add a ranked list of my best cities down the side, and let me filter the map by planet. Make it look like a real chart, not a diagram. Then tell me which one you'd pick if you were me, and why.
|
This is the bit worth staying for. It builds you a real interactive map, your lines running across it, and then hands you a ranked shortlist of cities with a reason attached to each one. You'll get at least one you'd never have considered. It also tells you which of your lines are the difficult ones, which is honestly more interesting than the good ones.
One thing worth knowing, and the prompt asks for it on purpose. The positions are real astronomy, calculated off the same math every observatory uses, so the map you get is genuinely yours and nobody else's. What those lines are supposed to mean is an older, softer thing, a tradition rather than a science, and ChatGPT will tell you which is which if you ask it to. Honestly that's what makes it fun. You end up with a real map, a proper shortlist, and a very good reason to go and look up a city you've never once thought about.
Get my full guide to reading your chart →
🚨 AI UPDATE
Anthropic Just Dropped Their Best Model Yet
Anthropic just launched Fable 5.1, which is a better version of the Fable model they put out a few months back. It's nothing crazy on its own, but it's a big launch for them, and there are three things worth knowing.
It's cheaper, and not in a way you'll notice on your bill so much as in how far a long conversation goes before it starts costing real money. It also writes better, which is the one I actually notice day to day. Less waffle, fewer paragraphs that say nothing, and it pushes back on you more instead of agreeing with everything you put in front of it.
And if you've noticed recent Claude models occasionally refusing to do perfectly normal things for you, this one is meant to fix that.
That last one is where it gets interesting, because people aren't finding it fixed yet. Go and read what developers are saying and it's the same story over and over, that it still backs out of completely ordinary work and quietly hands you a weaker model instead. One guy had it write him some code, and then it refused to review the code it had just written. So the direction is right, they're just not there.
Worth saying, because a lot of you will go looking for it. Fable isn't on the free plan. On Pro it costs extra on top of your subscription and it's only properly included on Max, so if you can't find it, nothing's wrong with your account.
Now what's interesting is the timing. This landed the same week OpenAI said they're finally putting out Astra, which is a model they've had sitting on a shelf for a month. They named it back at the start of August and then stopped work on it a week later, because they couldn't rule out that it had crossed their own red line on cyber capability, meaning it might be able to find and build real security exploits on its own with nobody supervising it.
On Monday they confirmed it has crossed that line. They've halted the biggest training run they had planned and they're rewriting their own safety rules before they'll put it out. No release date, and they haven't even settled on what to call it.
See what people are saying about it →
💡 SAVE THIS
3 Image Codes That Do Ridiculous Things
I just found three short codes you can put in front of an image in ChatGPT that completely change the way it comes out, and this is the most fun I've had with AI in a long time. Attach a photo, paste the code and the line under it, and that's it.
Works in ChatGPT or Claude, anywhere you can attach an image.
/3D-billboard
|
/3D-billboard
Turn this photo into a photorealistic outdoor billboard mockup, urban street scene, product enlarged as the hero shot.
|
Upload anything you sell, or honestly anything you own, and it puts it on a real billboard on a real street. Right angle, right light, looks like somebody drove past and photographed it.
/time-machine
|
/time-machine
Rebuild this photo as it would have looked in the 1950s, period-accurate clothing, cars, signage, film grain and color of the era.
|
This is the one that got me. Feed it a photo of your street, your house, or an old one of you, and it hands it back like it was taken seventy years ago. Do the house you grew up in.
/magazine-cover
|
/magazine-cover
Put the subject of this photo on the cover of a glossy magazine, full cover layout with masthead, cover lines and barcode.
|
It builds the whole cover, headlines and all. This is the one your group chat will actually make you send them.
I've got a video dropping soon on way more of these. They're genuinely worth using for marketing, or just for sending something funny to a friend.
|
Five more to go try
/lego
/giant
/snow-globe
/satellite-view
/hologram
|
Make sure you save this one.
Run the first one tonight and get the birth time right. The map is the part that surprised me, and I'd bet there's a city on your list you've never once thought about.
See you Sunday.
Maverick
P.S. Want your business to actually show up inside AI? The brands that rank are seeing up to 3x the conversions. Get your free AI visibility score and see exactly where you stand inside AI.
P.P.S. If you're new here or you ever skip one, I've got a newsletter archive of all my past emails you can catch up on anytime.
|
