I work at an early stage startup and I'd estimate 70-80% of our codebase is vibe coded (510k lines). To be clear, it's not 1 shot "build this feature." More like, "implement get_slim_documents for Jira in the exact same way we did it for the Confluence connector." Comfort with AI coding tools is actually something we gauge during interviews/work trials. Looking at our peer companies, it's exactly the same. My hypothesis/assertion is that companies founded ~2022+ are fundamentally intertwined with "vibe coding." In 5 years, programming will connote vibe coding more than it will connote non-AI assisted work. Am I crazy? Pigeon-holed in the SF startup world? Naive? Would love to hear more thoughts/diverse perspectives on this.
I don't know if this works for everyone or only for me. First of all, I think when AI coding tools first became popular, the biggest excitement was that anyone could build software without years of programming experience. But after experimenting with these tools, I think the hardest part has shifted. Writing prompts is becoming easier. The real challenges are:
Knowing what to build
Turning an idea into a product users actually need
Debugging unexpected problems
Making the app reliable enough for real users
AI can help you create a prototype quickly, but building a real product still requires product thinking, iteration, and understanding users. I m curious about people who have used AI coding tools to launch something:
What was the biggest bottleneck after your first version was built?
I m one of the devs who are writing less code now and doing more Prompting , damn I hate all of these fancy new buzzwords. Anyway, I m bored and don t want to write long prompts all the time. Anyone know a good tool (should work with Cursor)? If nobody knows one, I will need to do one :(
While building out Aiden's keyboard input handling, we hit something strange: modifier-key shortcuts (Cmd+V, Cmd+A, Shift+X) would intermittently fail on iPhone, while plain letters typed fine every time.
Took a proper controlled experiment to isolate it, keyboard-only setups worked 32/32, adding a mouse brought the failure back, removing it again fixed it, 22/22 clean. Turned out iOS was routing the command to the wrong place (SpringBoard, the home screen process, instead of whatever app was actually open) whenever AssistiveTouch was on and it saw both a keyboard and a mouse connected. We confirmed it's not specific to our hardware, it reproduces on a totally unrelated gaming keyboard and a Bluetooth keyboard too.
Our fix switches Aiden to a keyboard-only USB profile right before any modifier-key action, then restores mouse capability after. Workaround, not a real fix (that's on Apple's side), but it works reliably now.
Anthropic just shipped Opus 4.7 today and i had to write about it somewhere because the jump is weird.
I ran the same backlog task on 4.6 and 4.7 back to back. same repo, same prompt, same tools. 4.6 looped on a bug for 25 minutes and was not going to solve it. 4.7 closed it in eleven, and the part that freaked me out is that it paused in the middle to sanity-check an assumption i had not asked it to check. literally wrote "before i write this migration, let me verify the actual shape of the response object, because my assumption here might be wrong" and then went and verified it. unprompted.
That self-verification behavior is the thing. Vercel is reporting it does proofs on systems code before starting work. Hex says it flags missing data instead of making up plausible-but-wrong fallbacks. Genspark measured loop rates on hard queries and 4.7 basically stopped looping. different teams, different harnesses, same pattern.
Hi, I m an app developer who s shipped many projects the old-school way (hand-coding) for over a decade. Recently AI tools have exploded - speeding up my production like crazy.
What s happening:
- Idea to MVP: Creators focus on ideas while AI writes most of the code.
