So you’re curious about open source AI (and a little intimidated)?

By Heiko HeiligApril 16, 2025
cloudriftopen-source-aiai-toolsai-for-designers

So you’re curious about open source AI (and a little intimidated)? Here’s why you should dive in anyway

You don’t need to be an engineer to dive into AI — you just need to be curious (and willing to get a little lost).

Not too long ago, I thought I was pretty tech-savvy, especially when it came to trendy new AI design tools. I was using those Figma plugins that auto-generate style guides pretty early on. I’d tested UX-Pilot, Galileo, and even Uizard (“You-ai-zard”? “Yooiezard”?) before they were an alleged must-have in everyone’s stack. Some of those tools were game-changers, others were meh, but just knowing they existed made me feel like I had my finger on the pulse.

screenshot of the Figma Interface, containing an image of a GPU and an option to remove the background of that image.
Back in the good old days, I’d go into Photoshop, fiddle with contrast, selection feathering, and channels to get the perfect mask and remove a background. Now Figma just does it for you. You kids don’t even realize how crazy this is.
screenshot of the Figma Interface, containing an image of a GPU that has its background removed
I took entire SEMINARS on how to remove backgrounds!! SEMINARS!!!

And then… I started a new role as a product designer at a cloud and AI startup — and I got introduced to the world of open-source AI. And holy crap.

Imagine a world where subscriptions don’t exist. Imagine a place where you use a tool and it DOESN’T tell you to buy the pro license after three uses, or ask you to pay just to use a different font. Where everything is free and, well… open.

It felt like someone had yanked the rug out from under me and I’d fallen into some parallel universe I didn’t know existed.

Suddenly, I was staring into the Terminal most of the time, navigating forums where people casually dropped terms like “LoRA,” “negative prompting,” and “inpainting” into conversations — and everyone just nodded along like this was completely normal human speech.

The tools weren’t things you simply downloaded or logged into anymore — they were GitHub repos you had to clone, somehow get running in Terminal, all while praying you wouldn’t brick your laptop. (Because lord knows my ol’ reliable MSI laptop from 2018 used to be top tier, but now it sounds like a helicopter during lift-off most of the time.)
“Using AI” wasn’t just about crafting clever prompts and waiting for magic to happen. It meant building actual workflows, testing models, breaking stuff, fixing stuff, and joining communities that evolved faster than new AI trends on LinkedIn.

And here’s the weird part: I didn’t hate it. It actually sparked some joy — because it also meant freedom, control, and a much higher degree of agency and transparency when working with AI.

From “What the Heck Is Happening” to “I Can’t Stop Tinkering”

Let’s be real — at first, it was overwhelming. I walked in thinking I had a decent grasp on AI just because I’d read some articles, used a few closed-source AI design tools, and watched people turn themselves into action figures on LinkedIn.

Then came the reality check.

I quickly realized that AI is so much more than ChatGPT, Claude, Gemini, Midjourney, or Grok.
When you dive into open-source AI tools and forums, you realize there’s an entire ecosystem out there — still figuring itself out. It’s wildly janky sometimes, it breaks in weird ways, but it’s incredibly exciting to explore.
Because it gives you a glimpse into the future. Because it lets you see what people are building and what these tools are capable of.

And contrary to what I initially thought, it’s not just for engineers or ML researchers with PhDs.
It’s also for regular people like you and me — designers, creatives, makers, and the just-plain-curious who want to build cool stuff and learn by doing.

You don’t need to become an overnight expert.
You just need to be interested enough to be okay with looking like a complete newbie sometimes.

But Wait — Isn’t AI That Scary Thing Coming for My Job?

Yeah, I hear you. I’ve done the late-night LinkedIn doomscrolling too.

I get it. There are legitimate conversations happening about automation, creative ownership, ethics, and what work will look like in five years. But there’s also a ton of noise and fear-mongering.

I’m not going to go into the weeds here, but when we only consume headlines without ever touching the tools, it’s easy to spiral into anxiety about something we haven’t even experienced firsthand.

What I’ve discovered is that actually using AI — seeing its uses and, more importantly, its limitations — is way less terrifying than imagining what it might do.

The moment I started playing with image models, experimenting with prompts, and building simple workflows, I stopped seeing AI as this looming threat and started seeing it as a weird, powerful tool I could use to create things I couldn’t make before.

It changes a lot.
Especially if you’re a designer.
Especially right now.

Two Rabbit Holes Worth Falling Down

Screenshot of the homeage of civitai.com, an AI community focused on AI generated Images and Videos

CivitAI: The Wild West of AI Art Models

If you have even a slight interest in visuals — whether you’re into design, illustration, concept art, or whatever — CivitAI might become your new obsession. It’s a massive, open platform where people share AI image models and all their bizarre remixes.
You can download and play with everything from photorealistic portraits to… well, vaporwave-anime-vampire hybrids (Is that a real thing I found or did I just make it up? Only one way to find out — and the answer might surprise you).

Screenshot depicting the huggingface platform, an AI community that inlcudes a forum and provides access to different ai models and pre-build apps called spaces.

The best part? The community is surprisingly welcoming (From what I can tell — I’m still new). People share their workflows, give actual helpful feedback, and aren’t afraid to post their hilarious failures. You don’t need to understand the complex math to appreciate what’s happening — you just need enough curiosity to click around and try stuff.

If that sounds interesting to you, here are two great articles that go into a little more depth:

Hugging Face: The GitHub of Machine Learning (But Way Less Intimidating)

If CivitAI is like an art rave, Hugging Face is more like a library that somehow merged with a makerspace. It’s packed with models for text, vision, audio — plus datasets, demos, and these little apps called “Spaces” where you can try things directly in your browser without installing anything — there’s a good chance you’ll be waiting a while for a result, but it’s cool and it’s free.

It’s definitely nerdy, incredibly powerful, but also surprisingly approachable. You’ll absorb so much just by poking around and playing with the demos.

So… What’s a UX Designer Supposed to Do With All This?

If, like me, you’re in UX or product design, you might be thinking:
“Cool story bro, but where exactly do I fit into this AI chaos?”

Short answer? Probably right in the middle of it all.

We’re the ones who ask the uncomfortable questions. Who obsess over those edge cases everyone else ignores. Who make sure these powerful tools aren’t just functional, but actually usable and desirable by real humans.
AI doesn’t just need better algorithms — it desperately needs better interfaces, more intuitive flows, clearer language, and actual accessibility. That’s literally our whole job.

AI generated portrait of a realistic looking cat wearing an astronaut’s suit looking into the camera
Seriously, getting this out of an open-source AI tool you set up yourself — without throwing money at ChatGPT — feels liberating.

Even just browsing CivitAI or Hugging Face becomes a fascinating case study in design:
What’s working well? What’s confusing as hell? What would you change if you were in charge?
This entire ecosystem is full of real-world UX problems begging to be solved.

AI Isn’t Just Online: There Are Actual Humans Involved

And just so you know — this isn’t all happening behind a screen.

Depending on where you live, there are in-person AI meetups, weekend hackathons, creative tech workshops, and community spaces where people are building stuff together.
And it’s not just stereotypical tech bros in hoodies (shame on you for thinking that!) — you’ll find artists, designers, writers, researchers — and plenty of regular folks who are just curious about the technology.

You Don’t Have to Know What You’re Doing (Seriously)

Here’s the thing nobody tells you when you’re getting into AI:
It’s totally fine to mess around without having a master plan.

You don’t need to have an awesome project mapped out from start to finish. You just start with a simple feeling:

“This seems cool, and I want to see what it can do.”

Some of my early creations looked absolutely terrible. Some broke halfway through.
I once tried to generate epic fantasy portraits and ended up with something that looked like a biblically accurate pile of crap. (Or like something I definitely wouldn’t want my wife to see.)

AI generated Portrait of a medieval minotaur-looking character, tastefully censored. The AI messed up and depicts the character with a capeand  boots, but without pants or underwear.
Open-source AI tools are often unpredictable and a little janky. Be prepared to face the occasional NSFW content.

But every time, I learned something new. Picked up another weird term. Found another tool. Googled yet another cryptic error message. Had small wins that kept me going.

The more I embraced the messiness, the more I realized: this chaotic energy?
That’s where the real magic of AI happens.

So Where Do You Actually Start?

Start small. Watch a YouTube tutorial that doesn’t make your eyes glaze over.
Create a Hugging Face account. Scroll through CivitAI for 20 minutes.
Tinker with something. Break it. Ask questions that might seem dumb.
Trust me, you’re in good company.

Or maybe just wait for my next post, where I’ll walk you through using ComfyUI (an image generation tool with a visual interface that won’t make you cry), how to rent a GPU (without frying your laptop or emptying your bank account), and how it all fits together to help you build your own AI workflows.

OR if you’re feeling curious now, you could check out some other articles on Medium that dive deeper into the topics I touched on here:

On Open Source AI in general:

Could open source provide a better, more accessible AI?

On Huggingface:

7 models on HuggingFace you probably didn’t knew existed

On CivitAI:

INTRODUCING CIVITAI

TL;DR:

  • You absolutely don’t need to be a developer to get into open source AI.
  • There are tools and communities (both online and IRL!) waiting for you.
  • Starting messy is infinitely better than waiting until you “get it.”
  • UX and product designers have a huge role to play in making AI usable.
  • You can take a course, but the real fun starts when you actually try things.
  • This is honestly the most exciting time to play with these new tools — especially if you’re creative.

So you’re curious about open source AI (and a little intimidated)? was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.