There’s a very real thing you’re missing in your ‘did AI kill this or that’
Humans want personal attraction. Humans are pack animals; they follow people, not imagination.
An AI-written song won’t do that.
Humans follow humans and want to feel a personal connection.
This is one of the main reasons AI won’t kill a lot of the things people think it will.
A whole philosophy behind all this. If you look into it, it will convince you that humans still will be mainly attracted to other humans and what they do, etc.
I could replicate a site and do 10,000 times better marketing for it than most ever could…and make 10,000 times more money doing it…just by understanding human behavior.
There are still principles of human behaviour, even with AI, that stand the test of time.
The thing is, every tech and creative forum in the world is having versions of this same conversation right now. Music forums, graphic design, architecture, filmmaking - you name it.
AI is holding a mirror up to us and forcing some deeper questions. What is this actually about? What’s the point? What are we really trying to achieve? And why are we always in such a rush to get there faster and more hands-off anyway (often at the cost of the very journey that makes life meaningful in the first place)?
In the music business, some are getting exited about being able to ‘prompt a track’, others think that is souless & pointless. Many are arguing… meanwhile some dude is currently sat in his basement on his old dusty Atari St knocking out absolute bangers not giving a crap about any of it.
tbh I’m not even sure I know what I’m trying to say anymore I guess: use and support whatever tools & community that brings the most joy into your life? If that’s Bubble, like it is for me, then support it. Jump into the community, be constructive, and help shape it in to the best version of itself that it can be.
I keep seeing these posts about moving from bubble to an AI built app, or how AI is killing the no-code industry, but the one thing I haven’t seen is any application fully built on AI from no-coders currently in production.
I don’t know why people aren’t showing off what they have done, but it makes these posts pretty meaningless to me.
And I’m not saying that it’s impossible, I’m sure it’s happened a handful of times but at least some screenshots would be nice.
Nice site and I wish you all the best with it. I think the original one is messier but has more personality… the new one looks like it was created by ai rather than a designer and is extremely neat & functional but also very bland and like 100 other ai sites I’ve seen recently.
One of the things I love about Bubble is the graphic editor - some people complain about it but as a graphic designer with years of experience on traditional design apps (InDesign etc) and webflow & pinegrow on web, I can say that it’s actually a very great visual editor & should get more respect. I love that I have a total control over every pixel and moving stuff around is a simple as a couple of mouse movements or key presses. Setting up conditionals on elements etc also very slick and easy - would hate to resort to ‘prompting’ all of that micro “attention to detail” stuff…
Do you think that the AI version can’t be made to look EXACTLY like the Bubble version if he wanted?
Agreed with you on Bubble’s visual editor, but that’s90% doable in AI apps now anyways. There’s Claude Code Workflow Studio and then just this week a super cool back and forth MCP with Figma… you edit the wireframe and the agent updates the code…
I am sure it can… and vice versa. The more design minded someone is, the better results they can get. Bubble gives us design tools to do it the manual way AND via prompting too (not great yet but only a matter of time) and mix and match. Prompt to create rough start point, then have precise visual control too with a nice editor if needed. Convert bits to reusables. Easily feed in dynamic data and conditionals on the text. Total control over the design. Lovely. So it’s another win for bubble imo.
you expressed preference for the bubble version. however he obv prefers the other version because if he wanted the bubble (more personality) version it would be a few minutes of work and he would have it in the code version.
and I’ve never heard anyone promote Bubble as the best design tool / front end UI or even in the top tier (and this was before AI). Bubble is great as a full stack app (if you are more of a CRUD application and dont need AI, looping, parallel processing etc then its the BEST ). The issue is it takes way too long to do admin and basic things when compared to any platform where AI is properly leveraged to increase output.
No, i actually think the Ai generated one is much more professional and better. My point was that the Ai one also feels a bit bland and ‘ai generic’. There are loads of sites looking very similar right now, because of ai, and that’s something ai prompt builders without design skills should be wary of.
I expect people aren’t shouting what they built with AI because they’re trying to sell their product/service? If they say it didn’t take long to build then it devalues your craft in the eye of the consumer. It would come across like saying you completed a computer game using a cheat code. Who wants to admit that?
IMHO, using AI to build is addictive and fascinating… but prompting simple front-end changes is tedious, especially when you have navigated this with Bubble before where it’s fast. Bubble is more fun to work with in this regard.
The “all apps look the same” could be said about Bubble. It’s been debated before on the forum. They all start the same and Bubble doesn’t allow all modern styling options (why they don’t add more when relaunching the new editor I don’t know…) like say, Figma…
Both of these things might become irrelevant debates very soon as it matures. Just this week, in Claude you can now replicate content in Figma > You can edit the content and feed it back to your app so it kinda solves this problem of design.
I was a bit skeptical 12 months ago but then Gemini 3 dropped, Sonnet 4 came out and it’s only going to get better, faster, less error prone, cheaper.
I love Bubble but it feels like it’s losing appeal - I want to see them innovate more aggressively. I haven’t seen anything in its AI builder that is useful (and I can’t use it for existing app…) or actually solves a problem. the most exciting thing about Bubble right now is George Collier’s Buildprint. But I wouldn’t be surprised if it’s also contingency, in the event there’s a demand for a service to migrate to another platform.
I feel like I could have built my last app from scratch in Bubble. New apps I’m on are all via Claude/Antigravity right now… it’s more fun and substantially quicker… and the skills you learn in Bubble about security, privacy rules, performance and best practices all apply to code you just have to know what you’re doing.
FWIW, I don’t think Bubble will go anywhere any time soon. There are too many apps on it. If it declines it will be slow. I hope that’s not the case!
I don’t understand how he’s shipping faster than an entire VC-backed company. Are they so mired in technical debt that it’s literally impossible to move beyond a glacial pace?
That’s unfair to Bubble and an apples to oranges comparison.
NQU is building it as a passion project with no expectations. At Bubble’s scale there’s a lot of moving parts to consider. Besides if someone is already working on it, and I am very sure Bubble are very aware of the project, and doing it well - why not embrace it instead of building a native competitor.
I don’t think any technical debt is issue. All it is is using the internal write APIs to make changes to the JSON via understanding of IDs necessary. The JSON structure is not that difficult to untangle. If you know how IDs for elements, workflows etc are created, and have list of all that exist then creating new ones that don’t disrupt is not complicated for an LLM.
With all of that, I don’t know why bubble is so slow at it, but I don’t think it’s their technical debt.
Not talking about the backend. It’s the frontend. Part of why they created the new property editor was because the previous one was very difficult to change. Lots of legacy code and technical debt.
Well, they aren’t sabotaging themselves on purpose and they’re not lazy, so by elimination that leaves technical debt.
I don’t think AI killed no code, it just changed the entry point. AI lowers the barrier to writing code, but you still need structure, logic, and product thinking. No code tools still win when it comes to speed, visual workflows, and non technical teams shipping reliably.
If anything, AI and no code are merging. AI helps generate logic, and no code helps organize and deploy it. The real shift isn’t coding versus no code, it’s who can turn ideas into working products fastest.
I guess: use and support whatever tools & community that brings the most joy into your life? If that’s Bubble, like it is for me, then support it. Jump into the community, be constructive, and help shape it in to the best version of itself that it can be.