The most wanted feature on Bubble: AI Copilot for Bubble app creation

ok bubblers, let’s say it out loud. what is the most wanted feature in bubble.io? What will be the most groundbreaking tech in nocode realm and bubble has all the potential to revolutionize this industry?
Right, GPT. Generative large language model, or simply AI that based on bubblers’ prompts will create bubble applications.

I have to confess without ChatGPT I would not be able to progress that far with the development of my app.
Even though I have to write long messages in ChatGPT, clarify and reiterate when AI’s advice does not work, I still do much faster searching on forum and consulting bubbler community.

Now the question to Bubble.io decision makers, strategists, founders, how long will you hesitate to undertake a full spec integration of GPT (open source, why the hell you should pay for proprietary LLM).

Make us happy, tell us that you are on it!

They are working on it (read Josh’s monthly update for February), but I can assure you it’s not the most wanted feature by the majority of the folks (or maybe just the vocal minority) in the forum.

5 Likes

I think it could be useful at times, but it’s not the top of my list, I’m most pumped for native app compiling :fire::fire::slightly_smiling_face:

9 Likes

Native app compiling for sure! Can’t wait for this.

6 Likes
  • A native mobile app is the most wanted feature for the majority of the Bubble community.
  • User Interface/Experience Improvements with better responsiveness, and easier design workflows
  • Allowing to export of code to run some editable and self-deployed applications.
  • Improved Performance and Scalability
  • Advanced AI Integration and *Machine Learning Integration: Allowing integration of machine learning models for tasks like image recognition or sentiment analysis could open up new possibilities.
  • Offline Functionality Enabling apps to function even when offline could improve user experience and expand applicability.
  • Auto-documentation of the database and workflows
3 Likes

This is the only smart use of AI for Bubble right now

  • return data from backend workflow
  • EU hosting
  • better performance in both editor and run-mode
  • mobile native
  • loops
7 Likes

NATIVE APPS!!!

Not that hard to sort the idea board to find that out.

Seems like #2 is native PDF creation, which would be great. And #3 is better performance, which would be game changer to have ability for performance to be on par with traditional code.

AI is pretty far down on the list for good reason. Who really needs AI to do something for us that we can learn how to do ourselves? I mean really, who needs AI to just do things for us that we can do ourselves.

AI can not make our Bubble webapps into Native apps. AI can not make native PDF generation, and AI definitely is not making Bubble performance any better.

So, let’s say it out loud…drop the initiative on AI design and just make the Figma import function properly because Figma the design platform is already making (might have already released) their AI design. Bubble should have from the beginning just been thinking a bit smartly in the corporate board room when they decided to sink so many resources into building AI design capacity when they could have just piggy backed off of a platform that does what they do best (implying Figma does design best and Bubble has other strengths that make it shine).

All those resources wasted on building their own AI design could have been spent more wisely on making performance better so that more potential users of Bubble would take the head first dive into it and all the existing users could have benefited from the improvements so as to see staying on Bubble and scaling on Bubble a more feasible option and not just jump ship to custom code (I know some of the reasons client jump ship for custom code is over ownership issues, but a lot has to do with performance and scaling).

I could personally care less about BuzzWords.

7 Likes

I’m honestly not looking forward to AI on Bubble. It’s absolutely not what bubble needs.
As mentioned above by Bubble power users, there are so many other functionalities that Bubble could implement that would be much more appreciated than AI.
AI is incredibly powerful and I do believe it should be leveraged as much as possible, but i’m tired of seeing it get thrown around as an all-powerful, set-it-and-forget, can-do-anything-you-ask-it-to tool. It’s not.

simply? :sweat_smile:

Even if we set ashide how many resources would be needed to actually create an AI bubble builder, imagine being presented with a complex, custom app built by an AI and having to make sure that it has no vulnerabilities, no errors, doesn’t have bloat. It definitely would have all three. Imagine having to change/add functionalities.
I would honestly rather use Templates than an AI built app, and I don’t like templates.

I do use AI for coding, its good at outputting functions, giving tips, giving explanations, but even then it often needs to be corrected. And its incredible at generating language (much more so than code), and it’s a great feature to add to your Bubble apps for a huge range of applications.

But outright building full, complex, custom apps… I just don’t see it happening.

3 Likes

So true!

Bubble’s marketing team likes to think that Bubble is easy and good for non-technical people, and anyone can no-code when that’s just untrue. The AI feature will just bring a shit load of people here complaining that their app doesn’t work because they have no clue what’s going on. Using AI to build it means you skip the important stage of making/experimenting/learning from mistakes and can’t actually understand how your own app is build. I’m sure they’ll start with AI UI components rather than workflows/DB/logic, but like you said, that’s rubbish and should make the Figma link work.

3 Likes

We will then need a @mikeloc AI to attend to the horde.

3 Likes

Not top of my list for the moment. AI will make bad apps better and good apps worse. You need creative thinking time and ‘what-if’ time while you build to create fast, efficient, secure, non-ambiguous, non-bloat apps. How is AI going to wake up at 3am with a ‘eureka’ moment’?

We would have to to use AI to find out why an AI-developed app doesn’t work properly as it could create a vastly complex app that only it can understand… policing itself…hmmm

1 Like

Still, this could indeed turn out to be invaluable some time.

The combined power of AI and NoCode could potentially change the industry in a couple of years.

It would be almost negligent if Bubble doesn’t recognize this and fails to see the joint opportunity presented by NoCode and AI.

1 Like

I do agree kaiphilip. It needs to be there for us to experiment with and who knows, maybe I’ll be on a beach asking it to create a super complex app but my experiments with ChatGPT on other types of coding have resulting in some very lengthy troubleshooting as the logic simply wasn’t there. Exciting times though.