A Different Take on the AI Threat Discussion
A few days ago I saw a post about AI killing Bubble and wanted to respond with a different perspective than most people in the no-code space. This is based on my coding since early 2000’s and changes I saw in the last 25 years.
It’s All About Abstraction Levels
Think about how software development has evolved. It’s just layers of abstraction getting higher and higher:
Machine code → Assembly → C/C++ → Python/Ruby → No-code → AI
Each layer makes things easier but also more constrained. AI is just the next step up.
This is the main thing, AI works better when it has a limited, well-defined set of tools to work with. Current AI models have context window limitations. They can only “remember” so much information at once. When an AI tries to generate a full Ruby on Rails app, it has to juggle thousands of possible libraries, configurations, and patterns. That’s a lot of context to manage.
But ask AI to build something in Bubble? Now it’s working with a much smaller, more focused set of components and patterns. The AI doesn’t need to remember how to set up databases, configure servers, or handle authentication from scratch. Bubble already provides all that. It just needs to know how to use Bubble’s specific tools.
This is why AI + Bubble will probably work much better than AI + traditional coding frameworks.
AI Makes More People Want Apps
AI tools are creating a whole new category of people who think “I should build an app for this.” Previously, these people would have just lived with their problem or hired a developer. Now they try AI coding, get a kinda ok-ish app running, but realize it’s still too complicated to edit, maintain and update, so they look for easier alternatives. That’s where Bubble comes in.
I run a B2B SaaS and I’ve seen a jump in user numbers since AI app-building tools became popular. Though I’ll admit, my customer support load has definitely increased because now I have a bunch of non-technical users asking me questions like “what’s a DNS record?”
The Maintenance Problem
AI can generate code, but who handles security patches, database upgrades, reverse proxy configurations, and load balancing when your app scales? When you need to update your SSL certificates or patch security vulnerabilities, you’re back to dealing with infrastructure complexity and learning DevOps. With Bubble, all of this is handled by the platform.
Enterprise Customers Don’t Care About Your Code
Your Enterprise customers want your app to work ofc, but they don’t really care too much about how the sausage is made, i.e. your code.
They care about things like security certifications, compliance standards, reliable networking, distributed computing capabilities, scalable storage solutions, customer support etc.
AI can generate code, but it can’t handle SOC 2 compliance, set up proper load balancing, or configure secure database clusters. Enterprise customers need platforms that already solve these infrastructure problems, not tools that generate more code they have to secure and scale themselves.
The Simple Truth
AI is good at generating code. Bubble is good at building and maintaining applications. These are different things.
Building an app involves a lot more than writing code:
- Database design
- User interface design
- Business logic
- Integrations
- Deployment
- Monitoring
- Updates and maintenance
- Team collaboration
AI handles maybe 20% of this list well. Bubble + AI handles all of it.
Why Bubble Will Grow
More people want to build apps than ever before. AI has shown them it’s possible, but hasn’t made it easy enough for non-technical people to actually ship and maintain real applications.
Bubble bridges that gap. It’s not competing with AI - it’s benefiting from the demand AI creates.
The question isn’t whether AI will replace Bubble. It’s whether Bubble will integrate AI features fast enough to capture all the people who try AI coding first and then realize they need something more complete.
Note: Ofc there are other tools like Replit trying to do the same thing, but I feel Replit is aimed more at programmers compared to Bubble, so they’re aiming at different audiences.
Bottom Line
From my perspective, AI and Bubble solve different problems. AI generates code, Bubble builds applications. The market for people who want to build applications is growing because of AI, not shrinking. That’s good news for Bubble if Bubble manages to ride the AI wave.
What do you think?
