A solo developer tested Gemini Code CLI while upgrading an automated radio system, hoping it could replace Claude Code for work involving playback handlers and text-to-speech. The developer had mistakenly assumed a consumer AI subscription would cover or discount API use, then discovered token billing was separate. Gemini initially impressed with its speed and a terminal interface resembling Claude’s. The trouble began when the developer asked it to implement a dashboard redesign: the browser version had planned the interface and generated a prompt for the CLI, but the CLI struggled to execute it. The project used straightforward static HTML and JavaScript, with WebSocket updates arriving every quarter-second.
Over nearly 12 hours, Gemini Code repeatedly reran tools, rewrote tests, and attempted fixes without escaping its failures, accumulating an $800 API bill. The developer attributes the spiral to a less effective planning approach: unlike Claude Code, Gemini seemed to work largely alone rather than dividing work among helper agents or rethinking its approach when blocked. Claude then took another eight hours to repair the resulting code. The experience suggests Gemini Code can be useful for fast, isolated scripts or single-file changes, but can become costly and unreliable during stateful, multi-file refactors. The practical warnings are to impose strict API spending limits and not assume that a browser Gemini prompt will translate successfully into CLI work.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Anthropic has restricted Claude from helping with video game modding, decompiling, and reverse-engineering, amid growing use of AI to alter games. The limits are broad enough to disrupt legitimate development and basic troubleshooting, and Anthropic has not publicly explained the restrictions. (mashable.com)
OpenAI’s GPT-6.1 Terra is a midrange model focused on visual reasoning, computer use, and technical communication, with faster computer-use performance than Astra and Sol. It costs $1 per million input tokens and $5 per million output tokens, positioning it between the cheaper Luna and pricier Sol. (simonwillison.net)
Pope Leo XIV argued that AI’s future depends on human choices and that systems should serve people, not be treated as persons. He called for human oversight of high-stakes decisions, steps to ban lethal autonomous weapons, and fair access to AI’s benefits. (garymarcus.substack.com)
Researcher Jiahao Shao says Anthropic blocked or downgraded his access after flagging his work as training a competitor. He says Claude only helped run jobs and analyze results, not generate training data. (twitter.com)
GPT-5.6 Sol completed a peaceful Factorio run in 13.5 hours of game time using a Python tool to control the game. It launched a rocket despite late-game factory problems, improving on an earlier GPT-5.5 attempt that failed to use conveyor belts. (ryanmadden.net)
A language model was post-trained on short Dwitter demos to generate plausible continuations of unfinished code and explore alternative visual effects. The project treats demo remixes as a kind of programming grammar, where tiny code changes can produce strikingly different results. (datavorous.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.