Reaching limits (maybe showing people how to save token usage by doing simple tasks instead of getting ai to do it. Copy paste a simple bit of code instead of gettingvitvtobrewrite the full thing for instance.)
2025-08-16 01:27:02
7
Dragos B :
Time :) and from a technical standpoint, I realized that the agents tend to get super optimistic and tick boxes saying Ready when it’s not
2025-08-16 05:05:18
6
Birgir :
Building in Lovable (sometimes Replit) how do I make sure my SaaS is secure, is there an external tools that can "audit" my security?
2025-08-16 07:20:41
5
CS :
after my 8 years of programing and 2 years using AI for work , my opinion is , if you really want to build something which can be called production ready project, today, you must at least know how code looks and to understand very good the technologies you will use , your stack . You need to have the terminology of a programmer to can ask ai to do what you want . And most important the planning and steps but for this you must understand your stack and deep understanding of the app mechanic, what how and when will happen. In the and ai is a tool a smart tool but like any tool you must learn to use it
2025-08-17 16:39:45
4
Happyswe2022 :
An step by step guide what you need to do for getting your app out on the market. When you have it finished?
2025-08-16 04:54:32
4
Movie Maven :
my relationship 🥺💔
2025-08-16 01:30:08
3
zookcruzr :
having an issue 'fixed' for it to result in something unrelated being removed
2025-08-18 06:46:09
3
vlix.ch :
That Lovable will turn off its Legacy Mode…
2025-08-17 12:29:58
2
Panner Selvam7468 :
Limit reach quickly cursor
2025-08-18 08:21:52
2
Eric • Vibing Philosopher :
Struggling with jumping around different tools and not building up a confidence and momentum..
2025-08-16 15:02:21
2
gunnarodland :
Package/dependencies chaos…
2025-08-16 06:58:29
2
eduardo.alexiso :
I reach the limit wit chaude code in less than one hour, today I check codex from OpenAI and also has the option to login with the subscription and the limit is 3000 for gpt5 thinking mode
2025-08-16 02:00:03
2
Swypr :
security
2025-08-16 02:46:55
2
ted finkleman :
Circular code breaks and fixes
2025-08-16 05:39:02
1
Security By Mitu :
Syntax errors, just why !!!
2025-08-25 22:41:45
1
Michal Varyš - Varyshop :
I assume.. debuging and performance 😃
2025-08-23 14:39:07
1
GXer :
time
2025-08-16 01:50:17
1
Grazzhopper :
Cost
2025-08-17 21:21:21
1
CS :
and I like your videos bro . :)
2025-08-17 16:41:38
1
StackOverflow@2025 :
Education AI tutor
2025-08-16 02:06:33
1
Believer_of_Christ_2023 :
Why are my comments being deleted
2025-08-17 16:38:43
1
ffffdv :
Rate limits, ai repeating same fix that ddnt work before, not giving the full code and using mockup data when asked not to
2025-08-17 03:49:49
1
🇨🇦🍁David824🍁🇨🇦 :
Redoing and using tokens when the prompt goes wrong.
2025-08-16 05:31:29
1
@ghuraba :
stack in API
2025-08-16 13:41:43
1
Ron :
if I just let the model build, it does whatever it wants and the result isn’t what I imagined (even with PRD and mocks). I realized I have to picture it clearly first, break it into small steps, and describe every detail so it looks exactly like I want. Now I make an MD file for each feature with rules for the AI to follow, and in that file I force it to ask questions before starting and again after finishing. That’s the only way I stay in control.
2025-08-16 07:27:55
1
To see more videos from user @marcinteodoru, please go to the Tikwm
homepage.