Click any tag below to further narrow down your results
Links
Thom Wolf argues that software is entering a transition period where releases blur into constant updates, closed code becomes effectively readable by LLMs, and users can generate their own features with a prompt. The hard problem he raises is how to customize software without losing the upstream updates that keep it working.
- Software now evolves continuously rather than in release cycles months apart, since one company reportedly merges over 30,000 code changes a day, and dependencies ship just as fast.
- Closed source is eroding: LLMs can already reverse-engineer Electron apps and binaries with debug symbols, and a description of a UI is enough to rebuild its internals, which weakens software IP.
- Bespoke features collide with continuous updates. Forking gives you your search box but cuts you off from upstream security fixes, so Wolf expects agents to rebase personal features onto each new release.
The author argues that coding agents, asked to build software that itself calls LLMs, reliably reach for brittle non-LLM approaches like regex or symbolic solvers. He offers three possible explanations and leans toward the idea that current agents are simply poor at software design.
- In one vibe-coded memory-contradiction tool, the agent built a complex symbolic logic system that couldn't work, even though it later admitted the underlying problem was a 130-year-old unsolved one; a single LLM call would have been the right tool.
- The author's own Autodeck agents kept categorizing flashcards with regex and keyword search instead of making another LLM call.
- Three candidate explanations are weighed: training code predates LLMs (ruled out, since models don't default to jQuery-era patterns either), RL environments rarely expose LLM API access, and agents are just bad at software design, which the author now finds most convincing.
@amQnese on X
The poster, writing in Indonesian, urges anyone using coding agents to run Cloudflare's security-audit-skill on every app they build, before problems surface later. The post is a short plea rather than a detailed review, pointing to the GitHub repo as the action to take.
- The author argues that coding-agent output should get a security audit as a routine step, not an afterthought.
- The recommendation applies to any app built with a coding agent, regardless of size or purpose.
- The main ask is simple: run the Cloudflare skill now to avoid regretting unaudited code later.
Developers and everyday users are using AI coding agents like Claude Code to remove ads, tracking, and bloatware from smart TVs, and replace the default home screens with cleaner ones. It's a growing DIY movement driven by frustration with ad-supported interfaces and viewing-data collection.
- Projects like "dumbtv" and "cleanyourtv" let users point an AI agent at a GitHub repo or a prompt to disable sponsored content, speed up the launcher, and turn off automatic content recognition (ACR), which fingerprints what's on screen.
- The fixes are not risk-free: some users found Claude refused simple "hack my TV" requests, others had agents disable important packages, one reportedly bricked an Android TV, and another got stuck partway through a Fire TV hack.
- Non-AI alternatives already exist for Android-based TVs, including ATV Launcher, Projectivy Launcher, and FLauncher, for people who don't want to hand control to an agent.
@PrajwalTomar_ on X
A post from the @ignytlabs agency lays out a five-part pre-launch security checklist for vibe-coded apps, expanding on a viral Reddit version. It argues that skipped security, like open Supabase databases and exposed API keys, is already costing builders money, data, and legal exposure.
- Most vibe-coded apps fail basic checks: missing Supabase Row Level Security, validation only on the client, and error messages that expose table and column names.
- Secret keys (service role, Stripe secret, OpenAI) must stay server-side, and any key found in the browser should be treated as already compromised and regenerated.
- Unrated API endpoints can run up bills fast; the author cites a Supabase bill rising from $20 to $200 in a day, so rate limits, hard API caps, and CAPTCHA on public forms are recommended.
@philvenables on X
A Kaggle whitepaper argues that software development is moving from casual vibe coding toward disciplined agentic engineering, with context engineering as the key to reliable output. It recasts developers as orchestrators of software factories and lays out a path for adopting AI tools without giving up production-grade standards.
- The paper contrasts fast, casual vibe coding with a verified, disciplined agentic engineering framework.
- Context engineering is presented as the main lever for reliability and consistency in AI-generated code.
- Developers become orchestrators of software factories rather than hand-writers of every line.
A survey of 167 software engineers reveals that while many feel they are keeping pace with AI coding tools, a significant number also express concerns about job security and productivity. The concept of "vibe-coding," popularized by Andrej Karpathy, highlights the changing landscape of software development, where AI assistance is both a boon and a potential hindrance. Engineers report mixed experiences, with some finding increased productivity while others struggle with over-reliance on AI-generated code.
- Of 167 software engineers surveyed, most feel they're keeping up with AI coding tools, but a notable share worry about job security and productivity.
- "Vibe-coding" (Karpathy's term) is reshaping workflows, with engineers split between productivity gains and frustration with over-reliance on AI-generated code.
The author explores the concept of "vibe coding," where non-coders use AI tools like Bolt to create applications with minimal technical knowledge. While the process seems accessible and fun, the resulting applications often lack depth and robustness, raising questions about their real-world effectiveness. The journey illustrates both the potential and limitations of such tools in empowering non-technical users.
- A non-coder used AI tool Bolt to build an app through "vibe coding" (describing what you want in plain language rather than writing code)
- The apps produced this way tend to look polished on the surface but lack real depth or robustness underneath
- The experiment highlights a gap between how accessible these AI tools feel and how functional/production-ready the output actually is