Sr. Content Developer at Microsoft, working remotely in PA, TechBash conference organizer, former Microsoft MVP, Husband, Dad and Geek.
145544 stories
·
32 followers

Stitch Design Variants: A Picture Really Is Worth a Thousand Words?

1 Share

My head has been in the world of code generation for the last few years. Now that I’m also thinking about design generation, I’m seeing a lot of similarities but also some fascinating differences.

For example, it is cheap to ask AI for things. I will ask it to start from scratch and do a lot of extra work that I would never ask of my team mates.

What isn’t cheap though is the human side of reviewing all of the code that AI creates. Reading through them, parsing the differences, holding them in working memory—it’s real cognitive load. It takes time, and time is expensive.

With designs though, the calculus is a bit different. It is still cheap for AI to spin up variations, but human review? Our visual brain is ridiculously efficient. You can glance at six screens side-by-side and instantly know which one sparks “ugh” and which one sparks “oh yes!”. That “emergence” moment—that’s the magic.

Which brings me to a new Stitch feature: Design Variants. It does exactly this. When looking at one of your app UI screens, simply click on “Create variants” and you get a spread of designs.

Here you see the entropy, the unexpected angles, the one idea you wouldn’t have thought of yourself. You select, delete, iterate. Fast. Natural. Fun.

This is just the first version. We’ll be adding a lot more ways to let the AI flood the canvas with possibilities while keeping the cost of selection close to zero.

A picture is worth a thousand words. Design Variants is here to give you … ten thousand?

NOTE: The example image above is a shared project … another feature the team shipped this week! I would love to see any of your Stitch projects!

Read the whole story
alvinashcraft
5 hours ago
reply
Pennsylvania, USA
Share this story
Delete

My Experience Asking GenAI to Design My Blog

1 Share

What was my experience using GenAI tools to design my blog? Well, you're looking at it! As I mentioned last week, my new design came from one of my experiments using GenAI to help me design a new theme, but I wanted to share a bit more about the experience when I had time, and that time is now.

About two or so months ago, I had the idea of testing out GenAI to create themes for small web apps. While my blog is actually huge (near seven thousand pages), design wise it's basically:

  • A home page
  • A post page
  • A "everything else" page

That's just three basic pages, all sharing a main layout with slight differences in what's the main content of the page. I thought it would be fun to give multiple different GenAI tools a chance to spec out a blog theme to see what would happen.

The Process

For my tests, I made use of the same prompt every time:

I want to create a new theme for my blog. My blog is developer-centric. 
My theme should consist of one main HTML file, one CSS file, and minimal 
if any JavaScript. (You can create an empty JS file and include it 
though if you wish.)

Absolutely NO framework must be used. Absolutely no React, or Vue, or 
anything, and no build process. I literally just want static HTML and CSS. 

I want a dark green centric theme as green is my favorite color. 

The theme should be responsive and look on mobile devices.

The main page you create should include a title for the blog and 10 
blog entries listed in reverse chronological order. Each blog post 
should include a date.

The top level menu should include a links to: Home, About, Speaking 
(I'm a frequent speaker at conferences), Subscribe. You can also 
include a link to a search page, or perhaps a small search field. 

Finally, using this template, create a page for an individual blog post. 
That template should include a date and a list of categories associated 
with the blog post. 

I tried to be specific in terms of what I wanted as well as what I didn't want.

For the tools I used, I wanted something that integrated into my editor and could write to the file system. I didn't always do this as one solution I'll share below made use of the CLI.

Once the initial results were created, I'd open them up and see if I wanted any tweaks. To be clear, not in regards to make the design 100% complete, just initial reactions and things I thought should change immediately.

Finally, for each of these (except one), I made a video, and that's primarily what I'll be sharing below. Ok, let's get to it!

Copilot in Visual Studio Code

My first attempt made use of Copilot in Visual Studio Code. This is baked into VSC now with no need to install anything extract, and you get a decent free tier with a GitHub sign-in I believe. Here's how it went down.

Play Video

You can find the resulting code here: https://github.com/cfjedimaster/gen_my_new_blog_theme/tree/main/copilot_in_vsc

And you can see it in action here: https://cfjedimaster.github.io/gen_my_new_blog_theme/copilot_in_vsc/index.html

All in all not bad, but not my favorite shade of green.

Gemini in Visual Studio Code

Next was Google Gemini in Visual Studio Code. Nearly all of my experience with GenAI has been via Google Gemini, so it has a special place in my heart. This one had a few issues while recording though.

Play Video

You can find the code for this version here: https://github.com/cfjedimaster/gen_my_new_blog_theme/tree/main/gemini_in_vsc

And demo it here: https://cfjedimaster.github.io/gen_my_new_blog_theme/gemini_cli/index.html

Cursor

Next was Cursor, an editor built on the open source portion of Visual Studio Code. You can see how it went down here:

Play Video

The code may be found here: https://github.com/cfjedimaster/gen_my_new_blog_theme/tree/main/cursor

You'll note it did not listen to my instruction to create a blank JavaScript file, but to be honest, this is pretty much what I do in a lot of apps when starting anyway:

// DevBlog JavaScript
// This file is intentionally minimal as requested
// Future enhancements can be added here

console.log('DevBlog loaded successfully!');

// Placeholder for future functionality
document.addEventListener('DOMContentLoaded', function() {
    // Future JavaScript functionality can be added here
});

You can demo this one here: https://cfjedimaster.github.io/gen_my_new_blog_theme/cursor/index.html

This was almost the design I picked.

Gemini via CLI

Ok, so this is cheating a bit, but it was suggested I give the Gemini CLI a try and I figured, why not. As I said, I'm a huge fan of Gemini but I had not tried out the CLI. Here's how this one went:

Play Video

You can peruse the source here: https://github.com/cfjedimaster/gen_my_new_blog_theme/tree/main/gemini_cli

And see it in action here: https://cfjedimaster.github.io/gen_my_new_blog_theme/gemini_cli/index.html

Not surprisingly, it's pretty similar to the Gemini in VSC one.

Claude

For my fifth attempt, I used Claude, again in Visual Studio Code:

Play Video

The generated code can be found here: https://github.com/cfjedimaster/gen_my_new_blog_theme/tree/main/claude

The resulting blog can be seen here: https://cfjedimaster.github.io/gen_my_new_blog_theme/claude/index.html

And the Winner Is...

None of the above! Ok, technically it's Claude, but I recently got access to Kiro, a new AI IDE from Amazon, which uses Claude, and when I tried out my prompt there, I was incredibly happy with the results. I didn't record a video for it, but I felt like I hand landed on my final design and just went for it. What's cool is - after I decided to implement it on my blog here, I went back to Kiro multiple times to ask for tweaks I needed for various parts of my site. As one example, I asked for a good blockquote UI.

And you can see the result here.

I also asked for a good design for a form field and button, which you can see in the newsletter signup below. All in all it just really clicked with me, even though I didn't even try the spec-driven development that is the core feature of the product.

All in all, I'm pretty darn pleased with this experiment. I've already used GenAI a few times for small design challenges and I think this one turned out great. Let me know what you think - leave me a comment below!

Image by Victoria from Pixabay

Read the whole story
alvinashcraft
6 hours ago
reply
Pennsylvania, USA
Share this story
Delete

Ranking cities by AI startup funding: Bay Area dominates, Seattle is No. 4

1 Share
(GeekWire File Photo / Kurt Schlosser)

As investors continue to pour money into AI startups in the U.S., Seattle ranks fourth in the nation in AI funding, according to new data from financial tech company Carta.

In a ranking of 10 metro areas, across seven sectors that also includes SaaS, hardware, biotech, health tech, fintech and consumer, Seattle’s startup ecosystem accounts for 5.1% of AI-related funding.

The number is a small piece of the AI startup pie that’s mostly consumed by California’s Bay Area, which gobbles up 51% of funding. New York (11%) and Boston (5.5%) are also ahead of Seattle.

Seattle ranks sixth overall with 3.9% or $3.46 billion of the $104 billion invested into U.S. startups on Carta between the third quarter of 2024 and the second quarter of this year.

In the categories aside from AI, Seattle is No. 3 in SaaS (5%), No. 7 in hardware (2%), No. 5 in biotech (3.8%), No. 9 in health tech (3.1%), and not in the top 10 in fintech or consumer.

Seattle is trying to keep up in the race to lure and/or retain AI companies and investor dollars. Despite the hype around the region as a major AI hub, there are no Seattle-area companies listed among the top 100 AI startup funding deals so far this year, according to PitchBook.

But given the density of Seattle-area engineering talent that’s crucial to building companies in the burgeoning era of artificial intelligence, entrepreneurs, investors, and startup leaders remain optimistic that the city’s best AI days are ahead.

The new ranking data comes from Carta’s platform, which is used by startups to manage their cap tables.

San Francisco-based Carta, which employs about 135 people in the Seattle region, recently moved into new office space in the Columbia Center downtown, occupying 23,500 square feet.

Related:

Read the whole story
alvinashcraft
6 hours ago
reply
Pennsylvania, USA
Share this story
Delete

Certificate Automation with Todd Gardner

1 Share

The days of the one-year SSL certificate are coming to an end - are you ready? Richard chats with Todd Gardner about the upcoming requirement from the Certification Authority Browser Forum to limit SSL certs to 200 days starting March 2026 - and they keep getting shorter until by 2029, certificates will last no longer than 47 days! If you haven't already automated certificate renewal, it's time to start moving in that direction. Todd talks about CertKit as a new approach to automating certificate renewal across various services, making it much easier. Still in the experimental phase, you can be part of the process and make your certificate life easier now!

Links

Recorded August 19, 2025





Download audio: https://cdn.simplecast.com/audio/c2165e35-09c6-4ae8-b29e-2d26dad5aece/episodes/561b49e3-acf8-499c-959c-f0c98f067dc4/audio/08be8708-6a1e-4e8b-9978-5370e786d117/default_tc.mp3?aid=rss_feed&feed=cRTTfxcT
Read the whole story
alvinashcraft
6 hours ago
reply
Pennsylvania, USA
Share this story
Delete

Fedora Linux 43 Beta Released

1 Share
BrianFagioli shares a report from NERDS.xyz: The Fedora Project has announced Fedora Linux 43 Beta, giving users and developers the opportunity to test the distribution ahead of its final release. This beta introduces improvements across installation, system tools, and programming languages while continuing Fedora's pattern of cleaning out older components. The beta can be downloaded in Workstation, KDE Plasma, Server, IoT, and Cloud editions. Spins and Labs are also available, though Mate and i3 are not provided in some builds. Existing systems can be upgraded with DNF system-upgrade. Fedora CoreOS will follow one week later through its "next" stream. The beta brings enhancements to its Anaconda WebUI, moves to Python 3.14, and supports Wayland-only GNOME, among many other changes. A full list of improvements and system enhancements can be found here. The official release should be available in late October or early November.

Read more of this story at Slashdot.

Read the whole story
alvinashcraft
7 hours ago
reply
Pennsylvania, USA
Share this story
Delete

The right wing is creating a society of snitches

1 Share

The post was up on Natalie's private Facebook page for 20 minutes - a crude joke about Charlie Kirk's death. When her husband suggested the comment was a bit much, she deleted it. The next day, Natalie received the first email telling her to kill herself. It didn't even take 24 hours to go viral.

In the 20 minutes the post was live, someone had taken a screenshot of it. She doesn't know how the screenshot got to the right-wing influencer who posted it. Natalie, who owns her own small business, was on a client call as the death threats started to roll in. She told herself to focus on the call, but she was shaking so hard that she was having …

Read the full story at The Verge.

Read the whole story
alvinashcraft
7 hours ago
reply
Pennsylvania, USA
Share this story
Delete
Next Page of Stories