MikeTrendsTrends right now
AI: reposAPI

search

code reviewers

Trends

  1. 1
    Amateur says he found a possible new planet using Claude Code●I think I found a planet nobody knew existed. I used Claude Code to find itYhnSportBaseball12023 min ago

    A user reports discovering a previously unknown planet candidate by using Claude Code, Anthropic's AI coding assistant, to process astronomical data. The claim, shared on Reddit's Claude AI community, is drawing attention because it suggests AI tools can help amateurs do scientific discovery work normally reserved for professional astronomers. The finding has not been independently verified or published in a peer-reviewed venue.

  2. 2
    Music Video 'LGTM' Celebrates Claude Opus 5.5●LGTM (Looks Good to Me) – Claude Opus 5.5 Music VideoYhnLifeFood448 min ago

    A music video titled 'LGTM (Looks Good to Me)' focused on Claude Opus 5.5, Anthropic's AI model, is drawing attention online. The song riffs on the software-review phrase 'looks good to me,' blending developer culture with commentary on AI code review. Tech audiences are sharing it as a lighthearted take on AI-assisted programming workflows.

  3. 3
    Greg Kroah-Hartman on security in the age of LLMs●Greg Kroah-Hartman – Security in the LLM Age [video]YhnTechnologyAI3442 h ago

    Greg Kroah-Hartman, the longtime maintainer of the Linux kernel's stable branch, is featured discussing what large language models mean for software and kernel security. The talk examines how AI-generated code affects maintenance, review practices, and vulnerability risks in widely used open-source infrastructure, and it is drawing attention among developers weighing the benefits and dangers of AI-assisted programming.

  4. 4
    Developer argues AI code reviews should run independently●Why I want AI coding checks to run independently, with fallible model reviews, human-owned criteria and room for differeMmastodonTechnologySoftware34 h ago

    A software developer is arguing that AI-assisted coding checks should operate as an independent step rather than being bundled into a single AI workflow. The argument calls for human-owned review criteria, an acceptance that model reviews are fallible, and room for different development workflows. A related point making the rounds is that configuration files like CLAUDE.md, used to steer AI coding tools, are not a magic solution and do not replace genuine review practices.

Repos