The point everyone is making is that you haven’t backed your statement. You can’t ask someone to back up an argument to a statement you made first that you did not back up.
It’ll really help your argument if you can share the stats with sources.
this is quickly devolving from discussing no-code in the age of AI to just ranting about Bubble. I suggest to you all to start a new discussion topic for this
Now that VC funding of these vibe-coding companies drying up, they are slowly turning to general purpose AI’s, next step is them to simply pop. I never taught they were any competition to Bubble, but they were enough to Bubble to allocate some resources to AI, anyways.
With private credit stagnant, gov spending is stallin YOY, and on top of that energy prices keep rising, fall of AI will be accelerated.
I hope Bubble switches course.
Again, @code-escapee give evidence next time, this is not social media
Saw some statistics the other day on the whole AI builder thing out there…
it’s not good for all the gold rush AI builders.
Thousands and thousands try them…less than 2 percent stay and pay the monthly fee long term.
Less than 2% ever actually go ‘live’…and of those, less than 1% stay ‘live’ for over a month or two.
I tried Bubble’s AI again a couple of weeks ago. It was ok for the first paint…Then I ran into some issues when I tried to add or change things. I gave up on it.
Myself, I wouldn’t care if Bubble only came out with an AI that you could ask it questions about building something, and it would just give you an answer…not sure it’s important to actually build it.
Anyway, don’t believe all the hype about the AI builders; there was a lot of money being invested with a lot of hopes and dreams that aren’t working in the long run.
I would rather Bubble work on things that real-life app developers who have a bona fide idea really care about
Bubble just needs to do a deal with George buildprint , integrate that, and then put every ounce of resource spared towards polishing the core product.
Definitely this. Then just leverage OPM and let community handle AI stuff. I’ve been saying this for nearly 2 years. Bubble should just focus on everything the community can not do.
You still impress me being one of the few who understand it properly.
Nobody in the thread said what the numbers are, but as mentioned, Emmanuel said in an AMA it’s proving a healthy uptick. And just think, if Bootstrapped for 7 years before first $6M funding, then second and most recent $100M in 2021 has Bubble still at top of the industry they created 5 years on, I’m sure they understand their internal metrics enough to change course if it were prudent. AI didn’t kill and won’t kill Bubble.
Yo let me make it even better before death by Jira
Not sure why this cracked me up
The name’s George
George Buildprint
Quality street reputation right there
AI is great for getting something off the ground quickly, but turning that into something stable and scalable still takes a lot of work. It’s also much harder to understand how to fix something that’s not working if you vibe coded with AI, since you don’t understand how everything fit together without building it yourself. In my experience, making one small change can take a ton of time because the AI doesn’t always follow instructions perfectly.
One thing i’m really happy about is how little we have to worry about code syntax/debugging (which AI has been helpful for), but I still think no-code will be preferred over AI.
I watched a video last night explaining the increase in hiring of senior developers just because of the large need to review AI code. Apparently it makes mistakes at a rate of 1.7x of a human developer and one stat implied 9/10 times it makes an error on a task.
Lots of technical debt piling up in codebases generated by AI
Been following this thread for the comments — great insights 
I’m not sure AI will completely kill no-code, but it’s already changing the game a lot.
I used to struggle moving from Bubble to other no-code tools. There was always friction. But switching from Bubble to building with AI (VS Code and Claude Code) felt completely different, the speed is on another level.
It’s not just incremental improvement, it’s a shift in how you build. I’m shipping faster, with more flexibility, and way less limitation compared to traditional no-code platforms.
It’s been such a big change that I even created a new brand for my service around this approach.
Jizreel
tangerineai.dev
I built this on Bubble, spent hours creating a plugin just to draw pasture polygons on the map.
The project is complex. We need to store daily snapshots for every animal group or lot, and if a user edits one retroactively, we have to update all the upcoming ones in a cascade. This was tricky on Bubble, and I was starting to worry about scaling.
So I decided to switch to coding with AI and rebuilt the app. This is what I got in 30 days, working in spare hours (and added more features than on the Bubble app):
The funny part is that I went to the supermarket while Opus was building those map features for me. When I got back home, it was ready and working.
Building with AI and spec frameworks is just much faster. We spend a bit more time planning, testing, and reviewing, while in Bubble we spend way more time actually building.
That’s been my experience, not just a theory.
Jizreel
tangerineai.dev
Vibe / prompt coding has been a weird one for me… It is speeding a lot of things up, but also slows a lot of things down. It always seems fast & fun to begin with - then as we get into the weeds, starts to feel less & less enjoyable…and leaves me feeling not quite so close to & “at one with” my app anymore.
One of the biggest issues I noticed is that it doesn’t allow me to get into the same kind of “flow state”. Same with making music or doing art … There’s a trance that occurrs when lost in creative flow , which feels like it’s very good for us at a mental or even soul level. I just don’t get the same level of flow state with prompt-coding and it can actually leave me feeling very drained & fragmented at the end of the day.
Recently I did a full session building a fairly complex page in Bubble without touching LLMs at all…and it honestly felt great. More focused, more enjoyable - and I think even faster to get the page completed in the long run.
From brain health perspective it was a nice mix of puzzle solving & creativity + flow state which somehow felt healthy … whereas vibe-only might eventually make us all mentally impaired in some way. Time will tell!
The other benefit of course was that I ended up knowing exactly how every single field, state, and piece of logic worked & generally gaining “skill”.
Not knocking vibe coding at all - that’s the way the world is going … Just worth highlighting some tangible benefits of NOT relying on it all the time 
I agree completely. I hate the fragmented nature of using LLMs and really enjoy getting into a flow state for building out features
This. I love building in Bubble because of this. Bubble makes building apps fun and it hits the dopamine receptors when I hit preview and things work. When things go wrong I enjoy being to apply fixes and see them get fixed in a few minutes.
This is not true with code. Wrong syntax and typos are not fun to find and fix. Troubleshooting AI code is worse.
I’ve been in marketing and video production for over 10 years. I enjoy the creative process and working with Illustrator, Premiere Pro and After Effects. Bubble gives me the same high.