1作者: ls612大约 1 个月前原帖
When it was first released in September the consensus was that it was one of the least stable MacOS versions in recent memory and so I have stuck on MacOS 15 so far. I'm curious if this has been improved enough with 26.2 to be worth updating to? Or should I hold off for 26.4 when the major Siri overhaul comes?
1作者: ZeroAurora大约 1 个月前原帖
Once again I&#x27;m seeing an agent development platform in GitHub Trending. Dify, RAGFlow, Flowise, Sim, Coze, countless to name, yet implementing the exact same goal. And there are still more emerging.<p>It seems like anything AI can be a trending, despite their repetitive works.
2作者: rubslopes大约 1 个月前原帖
I&#x27;m a developer who wrote everything by hand for years, but now I mostly use Claude Code and only dive into the code when necessary. I can still review everything and get my hands dirty when needed. The AI augments skills I already built.<p>My junior developers don&#x27;t have that foundation yet,but there&#x27;s pressure: management already expects shorter deadlines because &quot;AI makes development faster.&quot; So I need to push my team to be more productive, but if they rely heavily on AI tools, will they ever develop the debugging intuition that I built by struggling through problems manually? It feels like I&#x27;m asking them to learn to drive while also handing them a self-driving car.<p>On the other hand, LLMs really do make the team more productive, I use it a lot, and I also feel that I need to help them the adapt to modern tools.<p>Team leaders: how are you handling this tradeoff?
15作者: sberens大约 1 个月前原帖
There&#x27;s been a lot of discussion around the california budget and some proposed tax policies, so I asked claude code to research the budget and turn it into an interactive dashboard.<p>Using async subagents claude was able to research ~a dozen budget line items at once across multiple years, adding lots of helpful context and graphs to someone like me who was starting with little familiarity.<p>It still struggles with frontend changes, but for research this probably 20-40x&#x27;s my throughput.<p>Let me know any additional data or visualizations that would be interesting to add!
1作者: barelysapient大约 1 个月前原帖
Over the holiday I wrote an interdimensional &#x2F; multiverse news aggregator site inspired by Rick and Morty. Actually, I tried it in 2024 but the models weren&#x27;t there yet. This year I feel like it works.<p>The site works by scraping internet news and fits articles against user suggested universes. All of that is handled with Ollama, nemotron-3-nano on a Nvidia 4090. The images are nano-banana; although I&#x27;m thinking to explore using stable diffusion. The site is written in Golang and Templ.<p>Happy to answer questions and share about how it works.<p><a href="https:&#x2F;&#x2F;greenportal.news" rel="nofollow">https:&#x2F;&#x2F;greenportal.news</a>