Building MVP with ChatGPT 🤖
Cem Özçelik
39 replies
I was curious about how I can get assistance from ChatGPT during the process of creating an MVP.
I tested its capabilities from the idea phase to the testing phase.
Result: I think, ChatGPT can indeed be used as an assistant to quickly develop an MVP.
Sample prompts here: https://resources.usemagnetiq.com/building-your-first-mvp-with-chatgpt-a-roadmap-for-success/?utm_source=producthunt&utm_medium=mvp-chatgpt
Question: Do you genuinely believe that ChatGPT can be effective in this subject?
Replies
Swapratim Roy@swapratim_roy
That's an interesting topic indeed
Share
Wonderslide
I believe that large models will become indispensable assistants. But we will still need to choose the course for the ship, and LLM can sit on the oars.
@renat_abyasov I agree!
thanks for the resource 🙌
FunContact
GPT can be an awesome EA, but it does not have any original thoughts really, I would not use it for ideation/ idea validation. Only for tasks that do not require real creativity.
LogicLoop AI SQL Copilot
Absolutely! Btw if any part of your MVP involves writing SQL with just text, feel free to use LogicLoop https://www.producthunt.com/post... (we're launching today, so please check us out too!)
@jesika_haria Thanks Jesika and good luck today 🙌
Sweet discussion, and thanks for sharing the prompts!
@blake_whittington You're welcome Blake
Launching soon!
I have used chatgpt assistance in most technical use cases but your prompts are impressive and helpful.
@dineshan_sithamparanathan Thank you🙏
this is an interesting hypothesis. Must try it out.
3Common
i would argue that “MVP” should be something that you can build with your current skills, and not something that necessarily requires technical ability. many great digital product MVPs are built using basic tools like Facebook, Mailchimp, Instagram, Notion, etc.
Hunted Space
Launching soon!
Yess, ChatGPT is amazing. It can help you with idea validation, decide on most critical features, help you with UXI and many other things, but you cant rely on it completely as it still is AI. All of it's answer should be reviewed and validated by humans 🙂
@sandradjajic I agree, but for now 🙃 I can't imagine 5 years from now
I used GPT-4, paid $23, but was able to build Extension for Twitter in 10 days. It was not possible without the GPT's code assistance because we already spent a month trying to build an extension earlier. It is worth X10 of money. I started building by 11 May and now it is approved, live, and has users. Check this
https://chrome.google.com/websto...
@janak_patel56 I'm using GPT plus too. It's quite helpful tool with idea generation and code reviewing
@janak_patel56 impressive
@vladislav_pavlenko It is time-saver if you know how to exploit
@janak_patel56 what a story! LFG
Wishing you best of luck @cem_ozcelik
@shaur_ul_asar Thank you Shaur
Embroidery patches are decorative designs or logos created by stitching thread onto a fabric base.
https://austintrim.co/custom-emb...
100%. I built a prototype with ChatGPT - not a public MVP, but still, it blew my mind. From no coding experience at all, I had a working prototype using Python, HTML, Airtable, Open Banking - all mainly through learning and iteration from GPT.
Have since handed over to an engineering team, but the prototype was invaluable.
Tamly: Automate B2B Sales Outreach
This is very smart because you can build and validate everything with ChatGPT. Lean startup is getting much more leaner ...
Definitely great for idea exploration and also more. Isolated code it produces like a breeze! What are your thoughts?
@milappert In my opinion when it comes to isolated code, it is important to ensure that the code is well structured, modular and follows best 2 player games practices. Separate code can be easier to maintain, debug, and test, which can ultimately improve the overall quality of the software or application being developed.
AppManager by CompanyDNA AI
For sure, ChatGPT is great for every step of creating an MVP. It can also generate amazing product improvement ideas if you provide enough data sources.