It feels ironic…
No-code was built so everyone could create without coding.
Now with AI, everyone is coding.
Prompt → Generate → Deploy.
The barrier isn’t writing code anymore — it’s knowing what to build.
What do you think?
Is no-code evolving… or being replaced?
Ouch 
Wasn’t there some talk when no-code came out that there wouldn’t be a need for developers who code anymore?
Guess it’s just a different day with different panics 
That chart reflects detected domains running Bubble, not traffic or depth of adoption. It trends upward as long as more sites are created than taken down.
With AI generation, domain count inflation is easier than ever.
I spun up 30+ apps in a few weeks just testing Bubble’s AI output and comparing it to Claude 4.6. They show up as domains yet they represent zero real apps.
The flattening in mid-2025 happens right before Bubble began rolling out heavier AI app generation features, with the larger AI Agent push later in the year. The bounce roughly lines up with that timing.
So footprint growth alone doesn’t tell us much about long-term platform viability.
Moreover, for the first time we’re seeing public migration guides, “how to leave Bubble” posts, and goodbye threads. That simply wasn’t happening a few months ago.
As I said on earlier on idea posts, btc and ai is a bubble, with U.S gov deficit is declining YOY, BTC is popped already (60% reduction from ATH keeps going), now as we saw on gpt 5 release, and now with claude release getting backlash, I don’t think AI is getting too far.
Now, you can be agreeable person or highly disagreeable like me, it doesn’t matter, a fact is a fact.
I’m not sure I’m following your point. Are you saying AI is just hype, similar to crypto?
If so, I’m curious what data you’re basing that on.
No, I have tried to build a few apps with AI, and although I was able to get them working, it was still very difficult. Much harder than anything I would do in Bubble. It broke more code than it fixed.
What I am saying is that money is currently dried up (if private credit creation stays flat), AI is not similar to crypto, crypto is total bs. AI is a good complementary product, thats it. But it will fall short of expectations, like AGI stuff total bs. And we haven’t seen regulations yet.
Hence, AI will be the comfort zone of highly agreeable or lazy people thats it, think of it like astrology.
Yes. We know AI is hype, that’s why the world’s best hype man, Elon Musk went all in. Newest way to get investors to inflate stock price for future earnings that will never materialize.
I mean, fully autonomous self driving cars are only 2 years away since 2016…now it’s extreme abundance such that we will all have robots watching our kids (and for what? If I have extreme abundance, I have nothing to do but watch my kids)
AGI will never come about through LLMs. LLMs will not train themselves, since data is already in that LLMs have basically ran out of new data to train on and new models trained on AI generated data produce worst models.
Models have basically peaked. ROI for training new models non existent. LLMs showed in record time the flaws of Moores law, things do not forever get 2X better every two years. Processing chips will soon be facing same realities (10-20 years).
Plus quantum computing will in 4 years time take most VC money and investment away from AI and hack all crypto walkets.
Crypto won’t bounce back, especially when regulators see citizenry losing savings through crypto devaluations, they will slap regulations onto it. Most institutional investors hold only around 1% of their assets in crypto so it’s not really fully adopted and entrenched. Plus the vail has been lifted, it’s not digital gold, a store of value nor a new way to conduct transactions to replace credit cards nor is it decentralized.
We will not have a need for data centers in space…
I can only speak from my experience - worked with bubble 6 years or so, ran a bubble agency, shipped complex apps, saas, you name it.
Few weeks ago, I cancelled all of my paid bubble apps because I went all in on AI-assisted code. For about 6 months I was experimenting, trying to find that battle-tested stack and I did (next.js, postgresql, vercel, AI sdk, inngest). It’s a no brainer for me - the speed of iteration, absolutely no limitations, owning every single piece of the product and having full control is paramount.
I respect how far the bubble have come and the hustle - Emmanuel and the team are working hard to stay relevant in the fast paced AI tech world. I wish them nothing but success in their niche.
But for me, I will prob never touch another bubble app ever again. I’m rebuilding my bubble apps from scratch - using codex, claude, glm in the good ol vscode. Just today i deployed one of the apps i rebuilt with cc and codex. Claude Code did all the coding and Codex handled Stripe - created products, webhooks via stripe cli, tested it all so I did not even touch stripe dashboard but provided AI with the stripe key and it handled everything.
You can check it out for yourself:
The next natural step for me would be to ship a native mobile app with AI, but thats for another post!
Cheers!
Nik
Oh no…AI killed Bubble…again. At this rate I should be buying stock in Bubble. Pretty resilient for something that gets killed every few weeks.
I do not think it will replace no-code but I do think it will drastically reduce the market and its relevance. If you think otherwise then you probably believe that consumers are rational actors.
Code syntax is becoming irrelevant. It will get hidden behind the bonnet just like “1s and 0s” currently are and abstracted away. Nobody will think in code, nobody will need to look at it. The novelty of seeing it and “feeling like a proper programmer” will soon fade away.
Building a website will be a combination of English prompting & Visual, for everyone.
The development platforms that end up on top are those that will integrate both of those aspects together in the most user friendly unified “project management experience”.
Developers tolerate fragmented stacks but most people don’t really want to think about “tech stacks” (“postgresql, vercel, AI sdk, inngest blar blar blar”)… They don’t want their data and pages spread all over the place. They want a stable unified environment that allows them to get from idea to result (a fast secure and gorgeous website) in as frictionless & enjoyable a way as possible, where everything is under one roof and totally dependable, where they can invest time & become experts of, alongside a thriving community.
No-code isn’t dead, it’s the future, and if Bubble gets it’s act together, fixing it’s flaws, restrictions & pain points, it’s going to be in a very healthy position because it’s already sat exactly where all the other companies are eventually going to end up anyway.
Well, you’ll have to do it in the NASDAQ private market (yes, that’s a thing) since Bubble isn’t public
The whole AI talk is a lot more complex than most understand. It’s more technical, philosophical, ideological, humanistic, and any other almost nonsensical word.
A person above built an app using AI. Great, now they’re in a 24/7 race with every other vibe coder. I mean, if someone can easily build an app using AI, I or any Vibe coder could easily replicate it in a day.
Also, theoretically speaking, with AI, why do we need most apps if we can just ask our own AI to do what they do?
There are a couple of ways in 2026 to ensure you can’t be copied or purely used by AI. But that’s another whole Netflix mini-series
I heard this too from a friend, what is this ?. I don’t think people grasp really how big the world really is.
Yes, this is essentially a continuation of an already existing trend. Assembly programmers have always looked down at languages like Python because they don’t even consider that programming. That was the original “no-code.” But in the same way, you will always need a few guys who understand code at the lowest levels of abstraction.
I agree - code syntax is indeed becoming irrelevant. I also agree with the creator of the Openclaw - 80% of SaaS apps will disappear. There will be API endpoints for AI agents that will connect to any service the user wants by simply speaking into the phone.
I want to know what goes into my app, what kind of stack I am using…just like what goes into my food. When I don’t care, I use autocoder that codes away on my VPS non-stop, for like 24 hours, in a loop - completing tasks, doing regression testing and fixing bugs it finds.
A person above says they can easily replicate an app in a day using AI. Sure thing - this is exactly what I am doing. Any project, any idea i ever had - I can ship it in a day, away from my computer, sleeping, cooking, walking my dog or playing with my kiddo. Then I pull out my phone, preview, and send it a voice message telling it where to iterate.
I closed multiple clients just like that - have a call, transcribe, AI, clickable prototype next day while I sleep, deploy, person is blown away, tells me please take my money.
I just leave the screenshot here.
AI has been an interesting subject for me ever since it started a few years ago.
If you take the current AI and its ability to write code…that is very amateur in the big picture of things. If you think about the whole code thing…it’s created by humans and used by current AI. This is AI using very human technology.
Neural Software is where we’re headed. It doesn’t use any code.
Neural Software is just AI doing its thing without any code. Yes, it takes a bit to wrap your mind around that.
So, now, we can replicate apps very easily. In the near future, we won’t need most of the apps we have because our AI will just use Neural Software without any code to do what we want. Code will eventually become obsolete
