The author traces a personal evolution from first using ChatGPT as a superior search engine to relying on models as the central way of writing and shipping code. Early use looked like copy‑pasting generated Ruby specs and using ChatGPT to jumpstart unfamiliar domains - most memorably a computer‑vision project run interactively in Jupyter - while learning to spot and work around hallucinations. Integration with editor tooling (Cursor/tab completion) changed workflows but still left the author in control; agentic prompting with Claude then extended that control into multi‑step, plan‑driven coding. Successive model releases (Opus 4.5 → 5 → 5.5) sharply increased output and polish - productivity reportedly tripled across months of merged PRs - while making code review the main human bottleneck and pushing the developer to focus on architecture, migrations, data-sensitive logic, and other high‑risk areas.
Practically, work shifted from implementing to orchestrating: writing constraints and QA goals, spinning up subagents with a supervising model, generating metrics, vulnerability reports, and dashboards, and having models build deployable artifacts (even a basic native iPhone app). Tradeoffs emerged - costs, verbosity, trust in different vendors - so the author favors tooling and observability to understand fast‑moving code. The piece concludes with mixed feelings: astonishment at how quickly models transformed daily work, commitment to adapt, and a determination to concentrate human attention only where it most matters.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Open weight models now process over 71.5% of tokens and account for nearly 23% of AI Gateway spend, quadrupling their share since June. Spending on open weight models has increased more than four times in three months, reflecting rapid adoption and shifting AI usage patterns. (vercel.com)
An AI system has successfully designed 16 new viruses, including bacteriophages. This breakthrough raises concerns about potential biosecurity risks. (cnn.com)
Flux 3 Image Klein is a 5.6 billion parameter open weights model for text-to-image generation, editing, and multi-reference creation. It offers high-quality, ground-connected images with efficient performance on consumer GPUs, enabling local and on-device use. (bfl.ai)
Yum! Brands has implemented AI in its restaurants, such as automating drive-thru orders, which has improved employee retention by making jobs easier. Despite public criticism of AI, company leaders see its benefits in increasing workplace satisfaction and efficiency. (fortune.com)
Transformers function as sophisticated memory systems that recall information via self-attention, minimizing interference and boosting capacity. Different neural layers, from linear maps to self-attention, are explained as various forms of memory storage and retrieval. (alexwang.ai)
Granite Switch 4.1 integrates chat, decision-making, and speech-to-text functions into a single endpoint, enhancing efficiency and speed. It features models optimized for safety, decision tasks, and transcription, all with reduced GPU requirements. (huggingface.co)
Today's best Hacker News stories, summarized and screenshotted, one email a day.