Generative AI recipes

Introduced idea am calling generative recipes which are easy to follow instructions to give AI for vibe coding. Is simpler to give one than explain further. Got Grok through SuperGrok access to graciously make me one for here! Quite grateful.

-----

**Generative Recipe: Local Station Weather Blend**

A small public-service experiment in getting weather data closer to the ground.

Most weather apps give you a single number. That number usually comes from a model that has already decided which stations matter and how to smooth them. You rarely see the stations themselves.

This generative recipe does the opposite. It finds the three nearest public weather observation stations to your location, pulls their latest readings, and blends them with a simple inverse-distance weighting. You get both the blended result and the individual station data so you can judge it yourself.

It uses only free, public U.S. National Weather Service data. No API key. No commercial weather provider in the middle.

### The Recipe

**Goal**  

Build a lightweight web page that:

- Asks for (or detects) your location

- Finds the three nearest NWS observation stations with recent data

- Fetches their latest observations

- Computes a distance-weighted blend of temperature, humidity, and wind

- Shows the blend plus the raw station readings for transparency

**Data source**  

National Weather Service API (`https://api.weather.gov`). Public domain. Requires only a descriptive User-Agent header.

**Core steps an AI coding assistant should follow**

1. Get the user’s coordinates (browser geolocation with a manual lat/lon fallback).

2. Call `/points/{lat},{lon}` to obtain the list of nearby observation stations.

3. Take the closest stations that have a recent temperature reading (ideally under two hours old).

4. For each station, fetch `/stations/{id}/observations/latest`.

5. Calculate inverse-distance weights: weight = 1 / distance^p (p = 1 or 2 works well).

6. Produce a blended temperature, humidity, and wind speed. Show wind direction from the nearest station or a simple vector average.

7. Display:

   - The blended values clearly

   - Each station’s name, distance, observation age, and raw values

   - A refresh button

**Constraints for a clean public version**

- Prefer a single HTML file or very small set of files so anyone can save and run it locally.

- No backend required.

- Clear unit toggles (°C/°F, m/s or mph).

- Graceful handling when fewer than three good stations are available.

- Always send a proper User-Agent string.

**Optional but useful**

- Side-by-side comparison with a model-based value (for example from Open-Meteo) so readers can see the difference.

- Simple confidence note based on station distance and observation age.

### Why this is worth publishing

It is deliberately modest. It does not claim to be more accurate than every commercial product. It simply makes the underlying public observations visible and lets the user see how a blend is constructed. That transparency is the point.

It also serves as a concrete example of a **generative recipe**: a short, reusable set of instructions written so that an AI coding tool (or a human following carefully) can produce a working result with minimal extra decision-making. The recipe itself is the artifact.

Anyone interested can take the recipe above, paste it into an AI coding assistant, and generate their own version. The resulting page can be saved and used offline or hosted anywhere.

This is a small proof of concept for making public data more directly usable, and for treating clear instructions as a shareable product in their own right.

Globloc social media company idea

Found myself thinking a few hours ago about how many social media companies I use, and then started wondering what kind of social media company I might create, and then brainstormed for an hour or so and decided, why not do a post? So here are some quick thoughts on a framework for a social media company built on a global-local focus, which I call Globloc for short. So I'm just idly speculating and tossing something out to finish out the year, so criticism of it is boring. Just going to start typing now and worry about editing it later.

There is an oddity to social media to me in that it's not like when you pick up a phone as you can assume some social media company has to receive your data but Globloc would give you a key certificate, allowing you to connect to other Globloc clients on other people's computing devices. Then information, including files, would be shared by bittorrent.

Globloc would maintain a global list of users though, but wouldn't care what people shared privately, unless law enforcement came knocking of course. But even then Globloc wouldn't have any information about what was shared, only whether or not a user was valid on the system, oh and be able to offer de-encryption keys, but getting ahead of myself.

Globloc users would, could share with an expectation of privacy, like phone users already have, so nothing new in that sense, and shared data would be encrypted on their device's storage, using what I call DMESE as I may as well use my ideas, in a Possibly Public space. But you'd only have the full thing, like a picture, if requested, as otherwise you'd get a digest version in the stream.

So digest version would be by default. If you, say, clicked on something interesting to you, the request would go to the device of the originator which would serve it to you by bittorrent. Then if the originator had given permission you could save to your device, in case you wanted to see it later, but it would be encrypted. To leave the Globloc system, the item would have to be tagged as exportable, and you'd have to export, like to get a JPEG file if it were a picture. That would trigger a log back to the originator as an exceptional use. Favoriting the item or simply viewing it would not, as considered expected--as putting it out there in the digest you expect people to look at it.

Oh yeah, so clearing a lot of privacy concerns with these notions. People could share with a private network on the local aspect of the Globloc system, and even then some information wouldn't automatically be shared, while giving maximum flexibility, so like they wouldn't have to keep serving the same data over and over again if they wanted to let their friends store locally. But even then it's encrypted, and they get notified if the data is exported out.

Globally people could publish to the Globloc public stream, which is where it looks like other social media companies, except I'd give the option of follow without broadcast, where you can get public posts from anyone public without giving out who you are. Or follow broadcast, where you'd be seen as a follower. But there would be a count of you if you followed without broadcast, so that user might have a million followers, but only twenty thousand broadcast their identity.

The data meant to be broadcast public would be kept on Globloc servers. Everything else is being traded by users locally. So that's the global-local characteristic.

Even locally though the Globloc app would imprint requesting user information on data which I call IDDI in various ways, including ways on photographs meant to be invisible to users but detectable by machines.

Oh yeah, almost forgot--system would try to eliminate use of passwords, and at least between Globloc globally and the local client would use one of my favorite ideas. Hmmm...looks like I never named that idea. But that's where the key server would come into things, which is what would keep up with the user key certificates.

So Globloc could yank your keys, eliminating your ability to talk to other Globloc clients, if you did something really bad, like violating the rules or something.

And I'm starting to run out here...hmmm...seem to think there were some other things when I was brainstorming earlier, but maybe I lost them. That's the trouble if you don't write things down.

Oh yeah! Not thinking a lot about legal things. This idea is free and open source and I make no pretension of having even begun to handle legal issues. The "Globloc" mentioned above is a hypothetical notion for a possible company which someone may or may not implement in the future.

If you like these ideas, steal them!!! I don't care.

Actually, it wouldn't really be stealing as I'm giving them away.

But regardless I might conceivably build this company myself, if I could get some help. There is SO MUCH left that would have to be done, like designing the user interface, actually implementing concrete examples of things quickly mentioned here which could be major coding challenges. And people would have to go for it.

Why might they?

Well I like the idea of not just assuming some company is looking over everything I send, with the same expectation as with a phone call, which is how I grew up. And I like local encryption protecting the data from being illegally shared, as well as the methods to catch people if they share without permission.

And I like the idea of having most of the data shared locally so the company at the top level deals has far less data than social media companies today, which also could protect from, um, certain folks pushing national security down your throat as they routinely raid, hack or otherwise try to get inside your tech company so they can SPY all data...um, ranting now.

Maybe I should stop there. Wonder if I'll keep this thing up! Wonder if I'll post it. Yeah, I'll post it. Going to post now.

And coming back, found I did limited editing, as kind of intrigued at this thing being mostly a stream of ideas as more of just a brainstorm post without me worrying about making it pretty. Maybe will think later tomorrow, but that's another year. Time to push away from the keyboard. To next year.


James Harris

Ideas as attention engines

Coming up with ideas is fun, as I love ideas. I also like tossing an idea out here on the web. The potential then is that you can draw a good bit of attention, so I like to say: ideas can be attention engines.

But to me that opinion calls for a demonstration.

And one of my more recent ideas is SO simple and basic it seems like it would be fun to give myself permission to just run wild with speculations. So I gave away a completely free and open source idea which is to use shared images to help in validating guests, for some kind of event like a party. At its simplest, it's just that: someone shows up at the door brings out their smartphone, opens the app, and taps the screen and it shows the correct image, validating them. It is in essence an image invitation, or an image ticket.

I don't know if there are any apps out there doing that and I came up with it just idly thinking about wedding crashers and ways to make that harder. It's so basic I don't take it too seriously and with no intentions of developing it myself, it gives me what I like to call a throwaway idea where I can just have fun. And no, would not care if somehow this is a valuable business idea and someone made a billion dollars with it, and gave me none, as they would owe me none. That gives me complete freedom to run wild with some speculative analysis. Play at a business case.

Which is the point of this post. So point of disclaimers is: free and open source idea, I consider it a give-away, it's so basic I don't think it should be charged for anyway, and now I'm going to play with speculations.

Ok, so with images you have automatically out of the box that this approach has international potential. It's a very basic approach, so limited explanation, and the transaction value is highly specific which I like. So yes, anyone can share images to potential guests but an app can block that sharing so only people you list get the image, allow a professional feel, and monitor as guests arrive so you have a constantly updated list of guests onsite. And that's just what comes to me just kind of freethinking it.

With a potential customer, who has friends willing to have the app on their smartphones, a purchase decision, where let's say $1 US for an event of one hundred or fewer guests just to throw out numbers, would involve convenience, security, and trust. The value of the transaction is in giving a convenient tool that allows guests to have a "ticket" to the event with minimum fuss or hastle.

Pricing is about value to the customer. So, for instance at 500 guests you could have a higher price point not necessarily because the app would work that much harder, but because of the accepted benefit for the customer, so it's pricing on security, professionalism, ease of use and benefit to guests and host.

And pricing would vary by so many ways as it's about the value to the host! So shifting to Tokyo which is another premium level international city, price points could vary based on how valuable people in the city see this type of security for an event.

At the transaction level it would be a lot about what value a host sees in the service and pricing at that value which could be much higher than costs. But it's like singing, how much really is it for that person to belt out a song? But how much does the value shift? Quite a lot depending on who is singing and what, and where.

Oh yeah, images as I brainstorm here, could be very high profile at higher price points, like even designers or artists? Willing to allow use of their images for the promotional value? That could be a prestige point as well for an event.

At the highest level price points, you could link to even more security services, or a security firm might have such an app as part of its service, allowing it to control access to the event, and easily check guests. They just bring out their smartphone and you have an image to consider.

Visual images communicate well, and cross language barriers.

For instance here's one of my favorite photos of San Francisco.

It would also not hamper the visually impaired as though they might not see the image themselves, the point is for others checking the guests, so they could still use the app. App could have features designed to ease their use of it, like auditory aids.

It would be harder to bluff your way into an event, if it was clear that you should have the required app on your smartphone though someone might say they lost it, but then they'd be shifted over to a higher level of scrutiny to validate, and wouldn't make it past the first person at the door, who wouldn't have to think at all, just check an image.

In terms of potential market I think automatically of New York City as a premiere top level city, and the potential market is every event in the city during some particular time. And price points can vary based on host easily. At the bottom level price point you could have a basic image and the app priced to move, like at the $1 per event level mentioned. At higher levels you might have $10 per event for added security. And at premium levels you could use your imagination, including high level art images, and linkage with onsite physical security or even concierge services.

Fun exercise so far. These are the kinds of things I do. Sit around and muse about all kinds of things, and what I like is just putting something out there with an actual idea because it is fun to speculate with something that could actually be done by someone. I'm kind of wondering if something like this thing is already out there, and if not, why not?

Images are cool. They travel well. Using shared images as a security feature is easily explained. Hardest thing might be getting everyone to download an app to their smartphones, but that app would be free, and people have all kinds of apps now.

Gonna toss these speculations up, and maybe fiddle with this thing on an ongoing basis. I like posting things and editing which is a process that can take days or weeks or even longer--as I write it, so I can change it.

Will give me time to consider if I covered everything that I wish to ponder. Oh yeah, so the fun to me with this idea is that it IS given away so I can just run wild with speculations. I think the concept of using shared images is so basic that it would seem silly to patent something or whatever but that's just my personal feeling where others do not have to agree.

To me it's like a great basic potential community tool that can help facilitate greater community: just giving to the world this possible path to helping secure events, like say, weddings from things like wedding crashers.

And it's fun to put out some of my process. For me there's this giddy excitement in doing this kind of thing. Where I'm someone who just enjoys playing around with ideas--and throwing them out there. It can be a very messy process too! So if all the above just sounds wacky, I don't care. I'm having fun.

So how much attention can an idea like this draw? Well it can go wherever the web goes.

I try to monitor ideas as best I can, and one way is with search!

To see what results with your own search with this one you might try: guest validate app

Or: guest validate idea


James Harris

Guest validate app idea

Started wondering to myself about how you might use smartphones to validate guests, and came up with an idea. And it's a free and open source idea.

The way this app idea would work is with it on your smartphone, and on the smartphones of all your friends or guests or whatever. It could be a free app, which only charges you if you use it for a wedding or a party or whatever your social event is.

So with this, um, let's call it Guest Validator app, you could send invites to a list of people, where I guess they would be contacts on your phone. They'd get the invite telling them the time and place of the event.

When they arrived at the event, they'd push a button on the app, like a button on the screen, and it'd show a randomly created color coded "ticket", which would be unique to that event. And would not show until they arrived either by time, or location or both, where the Guest Validator would check the web for the time or GPS for location, to keep people from doing sketchy things like changing the time on their phone.

Of course it could be any kind of image, including ones the host had taken. Like here's an opportunity to toss in a photo of mine:

So this unique ticket would show, to the surprise of the guest who wouldn't even know what it looked like until the event, and the host could just look at their own phone, which would show the same thing, and let the person in, and from then on, anyone valid would have the same ticket, for like five minutes.

For more security it could change every five minutes, until lock out, when it would lock to the last image.

Or the color could slowly shift minute by minute.

And for additional security, the image would only show for 2 to 5 seconds with each check. Quite long enough for most people to see what it is, and compare without it just sitting there.

(You can keep adding security layers depending on your level of paranoia.)

And yeah, security is what you'd be paying for if anyone ever implements this idea, as otherwise you could just email a photo to all your guests for them to show when they arrive. But what if they share that photo with someone you didn't invite?

If the host or anyone at the event wishes to check a guest they'd just ask them to show their smartphone and hit the button, and the app would show the image from when that guest arrived, validating them. But by then, everyone would have the full list of pictures so anyone at the party could validate anyone else, by matching their smartphone's image against the valid list of images.

(Oh yeah, need to make sure it's actually their smartphone of course, if you get suspicious.)

The host would also be able to see which guests are checked off in the full list of invitees.

And that's it.

I like it because it's visual. Also it allows anyone at the event to check guests, so it allows you to in a sense crowd-source to some extent. Say, people know whomever is near the door would app validate people who are just arriving.

It'd be really hard to counterfeit a "ticket", as images change for a while, and also it has the backup that based on the image you know roughly when a guest arrived.

And that's just a quick throw-out-there idea for me. I don't like investing too much time in these things. Have them a lot. So not worth it to think it out too much.

Oh yeah, I have no idea if such a thing already exists. Did a real quick search and didn't see anything which doesn't mean it's not out there, but at least I do try.


James Harris