As a product manager, I tried all these tools (lovable, bolt,v0 ,replit) but then reverted back to AI assisted Design and development because of these reasons:
1. I have peripheral coding knowledge, understanding of system design so I had some pre-work done for me initially but someone with almost no coding knowledge, the dependence on these tools to create full fledged apps/features is almost impossible.
2. While these tools build but they may not take into account things like user centricity,context, demographics, persona etc. to build a satisfying user experience for the end user. So a lot of work definitely needs to be done there.
3. Sometimes coding with these tools,as mentioned, becomes messy so you need to put on your project manager hat to neatly modularize your project.
4. Testing and validation is something that you would need to continuously do.Sometimes, while fixing one error, these tools also mess up with things that were previously correct. A strict prompting and version management also need to be followed.
Overall, in my opinion, it will be about selecting a tool and learning the best techniques to use it for a given use case.
My current stack: ChatGPT(free tier), Claude (Free tier), v0 , VS Code with CoPilot
Definitely worth trying lovable.dev, I had the exact issues with bolt, lovable is better (but not perfect, so many odd UX quirks due to the chat only interface, still great)
On the Upgrade point - I keep upgrading each time I hit the limit, much better!
On the eggs - Here is the app I created, took about 12 hours over a weekend, definitely felt proud https://www.latestplan.com/ (for a family member complaining about schedule management)
I have a feeling that because you're a UX/UI designer, you felt like the app should have taken your input before going ahead and designing the app on its own.
As a PM, I only care about prototypes, so honestly, it won't matter to me whether the corners are curved. I want the app to understand my instructions correctly and create the needed feature in the app. So, the egg might be user-dependent.
I'd like to see smarter automation where the app first clarifies the details to obtain a full context before creating the app draft.
I agree with trying Loveable is essential here. Two things about loveable:
1. They never ask for going through any connection steps to supabase until you really want to do it. Instead I get to create my app straight away -
I love it!
2. They actually have a better Free-plan than the Basic-plan. The Free-plan gives me 5 prompts a day - which leads to me building a habit of using Loveable EVERY day to make the most of my 5 prompts. The Basic-plan instead gives Monthly prompts - which turns me to a procrastinator by abandoning my daily prompt-everyday-behavior.
I’m a designer and one thing I’ve done a lot of in my career is to get the teams I’ve worked on to define their terms.
Product design, UX, UI, craft: definitions of these terms are not agreed upon in tech. Even big companies who have aligned on leveling still haven’t aligned on titles: Microsoft uses UX and Product design in ways that seem interchangeable and sometimes just refer to these titles as Designer [focus area].
It looks like this is the first time you’re using Bolt, so I’d argue there should be some first time UX/inline onboarding that helps avoid the app making assumptions, leading it to do things you don’t want it to.
Even better: by not assuming what you want—at least the first time—this solves for the problem I raise as well as the problem you raise: there are likely to be a wide variety of users of this product and their command of technical terminology is going to be uneven. Design for that by having the AI do some discovery about the user along the way to build some context and preferences so when you say, “help me enhance my ui/ui,” you get what you expect, not mansplained.
As a product manager, I tried all these tools (lovable, bolt,v0 ,replit) but then reverted back to AI assisted Design and development because of these reasons:
1. I have peripheral coding knowledge, understanding of system design so I had some pre-work done for me initially but someone with almost no coding knowledge, the dependence on these tools to create full fledged apps/features is almost impossible.
2. While these tools build but they may not take into account things like user centricity,context, demographics, persona etc. to build a satisfying user experience for the end user. So a lot of work definitely needs to be done there.
3. Sometimes coding with these tools,as mentioned, becomes messy so you need to put on your project manager hat to neatly modularize your project.
4. Testing and validation is something that you would need to continuously do.Sometimes, while fixing one error, these tools also mess up with things that were previously correct. A strict prompting and version management also need to be followed.
Overall, in my opinion, it will be about selecting a tool and learning the best techniques to use it for a given use case.
My current stack: ChatGPT(free tier), Claude (Free tier), v0 , VS Code with CoPilot
Definitely worth trying lovable.dev, I had the exact issues with bolt, lovable is better (but not perfect, so many odd UX quirks due to the chat only interface, still great)
On the Upgrade point - I keep upgrading each time I hit the limit, much better!
On the eggs - Here is the app I created, took about 12 hours over a weekend, definitely felt proud https://www.latestplan.com/ (for a family member complaining about schedule management)
I have a feeling that because you're a UX/UI designer, you felt like the app should have taken your input before going ahead and designing the app on its own.
As a PM, I only care about prototypes, so honestly, it won't matter to me whether the corners are curved. I want the app to understand my instructions correctly and create the needed feature in the app. So, the egg might be user-dependent.
I'd like to see smarter automation where the app first clarifies the details to obtain a full context before creating the app draft.
Hello beautiful
I agree with trying Loveable is essential here. Two things about loveable:
1. They never ask for going through any connection steps to supabase until you really want to do it. Instead I get to create my app straight away -
I love it!
2. They actually have a better Free-plan than the Basic-plan. The Free-plan gives me 5 prompts a day - which leads to me building a habit of using Loveable EVERY day to make the most of my 5 prompts. The Basic-plan instead gives Monthly prompts - which turns me to a procrastinator by abandoning my daily prompt-everyday-behavior.
So yes - they also need a good UX-designer!
I’m a designer and one thing I’ve done a lot of in my career is to get the teams I’ve worked on to define their terms.
Product design, UX, UI, craft: definitions of these terms are not agreed upon in tech. Even big companies who have aligned on leveling still haven’t aligned on titles: Microsoft uses UX and Product design in ways that seem interchangeable and sometimes just refer to these titles as Designer [focus area].
It looks like this is the first time you’re using Bolt, so I’d argue there should be some first time UX/inline onboarding that helps avoid the app making assumptions, leading it to do things you don’t want it to.
Even better: by not assuming what you want—at least the first time—this solves for the problem I raise as well as the problem you raise: there are likely to be a wide variety of users of this product and their command of technical terminology is going to be uneven. Design for that by having the AI do some discovery about the user along the way to build some context and preferences so when you say, “help me enhance my ui/ui,” you get what you expect, not mansplained.