Alper Tayfur

This makes a ton of sense. Pencil is amazing for speed, but Figma is still where things actually get finalized and shared. Right now most people I know either redo screens manually in Figma or accept a messy import and clean it up later — both are pretty painful. An export that keeps layers and Auto Layout intact sounds like exactly the missing bridge. Curious how robust it is on more complex...

I’ve had both ends of the spectrum 😅 Best moment: using AI to scaffold a real feature in hours that would’ve taken days — tests, edge cases, even decent docs. Felt like cheating in the best way. Worst moment: trusting a refactor suggestion a bit too much, running it, and realizing half the repo logic subtly changed. Nothing “broke” immediately, which made it even scarier. Big lesson for me:...

I’m somewhere in the middle. I trust automation for the boring, repetitive stuff — categorizing transactions, pulling balances, recurring reports. That saves time and reduces mistakes. But I still want manual control for anything that affects decisions or money moving around. I like to review, sanity-check, and approve. Automation should assist, not decide. So: automate collection, keep humans...

My expectations are pretty grounded but optimistic. Where I think Opus 4.6 really shines is anything that needs long context + consistency: big codebases, multi-step refactors, agent-style planning where earlier decisions actually matter later. The 1M token context (even in beta) is the most interesting part for me, more than raw IQ gains. First places I’d try it: understanding and modifying...

This really resonates. Productivity tools almost never “fail” — they just slowly drift away from how you actually want to work. I’ve felt the same with ClickUp (powerful but overwhelming) and Notion (flexible but gated in annoying ways). There’s always that tradeoff between simplicity and capability, and most tools eventually tip too far to one side. I think that constant experimenting you...

This is exactly the kind of wake-up call people have been warning about. Agent ecosystems are starting to look like early browser extensions or npm — powerful, fast-moving, and full of trust assumptions. Once “install a skill” becomes normal behavior, malware just follows the distribution channel. The takeaway for me isn’t “don’t use agents,” it’s: treat skills like untrusted code assume...

Yeah, I draw pretty hard lines too. Anything irreversible or deeply personal stays human for me. That includes: full access to finances (I’ll allow read-only or capped actions at most) health, biometric, or identity data private communications where trust or intent really matters decisions with legal or long-term consequences AI agents are great for prep, analysis, drafts, and coordination —...

@nikolas_dimitroulakis You’re asking the right questions at the right time. From patterns I’ve seen across successful open-source dev tools, a few hard-earned lessons keep repeating: 1. Be opinionated from day one The biggest early mistake is trying to please everyone. Healthy projects stay clear about: what problems they will solve what’s explicitly out of scope Open source doesn’t mean...

Hello @gleedo This is a really solid reality check, not pessimism. My take: vibe coding is great for getting to truth faster, not for skipping fundamentals. You’re right that the build is no longer the hard part. Shipping an app is almost table stakes now. The hard parts didn’t disappear — they just moved: understanding users deeply deciding what not to build handling edge cases, reliability,...

@alonhmdt This is such a wholesome use of “vibe coding” 😄 Also, building it as a gift instead of a startup pitch makes it way cooler. I really like that you leaned into the linear, no-algorithm surprise aspect. That feeling of just turning something on and letting it run is exactly what streaming lost along the way. The YouTube IFrame workaround for a continuous, ad-free stream is a great find...

@vik_sh This lines up with what I’ve been seeing too. Mobile is great for discovery — scrolling listings, saving jobs, quick interest — but when it comes to actually applying, people still seem to switch to desktop. Resumes, cover letters, forms, attachments… it’s just more comfortable and less error-prone on a bigger screen. So “mobile-first” might be true for browsing, but not for completion...

Thank you for this post @busmark_w_nika . This is a solid list already 👍 One thing I’d add: be consistent in how you show up. If someone clicks your profile after seeing a comment, they should immediately understand what you’re into and why your opinion matters. Random activity across unrelated topics can dilute that. Also, quality > quantity in comments. A few thoughtful replies people upvote...

Possible, but not likely in the way people imagine. Money and compute are accelerating insanely fast — that’s the part that is undeniable. We’ll almost certainly get models that feel more general, more autonomous, and more capable across domains in the next 5 years. But the science side is the bottleneck. We still don’t really understand: reasoning vs pattern matching long-term learning without...

Hello Matt, Yes, there are basically two “defacto” approaches now, depending on whether you want visual eyes (screenshots) or structured inspection (DOM/DevTools). 1) Best for “bad padding/alignment”: Playwright MCP or Chrome DevTools MCP Instead of you screen-capping, the model can: open your page inspect DOM/layout/accessibility tree read computed styles / bounding boxes take screenshots when...

I mostly see a difference between experimentation and avoidance. Early on, trying multiple things quickly is healthy. But when “Founder” repeats 6–10 times with no visible outcomes, it signals a lack of follow-through and comfort with the hard scaling phase. For me, it reduces credibility around execution depth. On LinkedIn, listing every short-lived venture usually hurts more than it helps....