GitHub Fusion与Gemini 4 Carbon表现引关注
Today was a really interesting day. Two things stood out:
Today was a really interesting day. Two things stood out:
今天是非常有趣的一天。有两件事格外引人注目：
1) GitHub Fusion. I think people still dont grasp how good this is. GitHub Copilot Fusion automatically selects and coordinates AI models for coding tasks, with Fusion Max scoring just behind Claude Code running Opus 5.5 Max on Artificial Analysis’s Coding Agent Index at roughly 50% lower cost per task and, according to GitHub, taking 40% less time. This is really cool.
1) GitHub Fusion。我认为人们还没有充分认识到它的优势。GitHub Copilot Fusion 会自动选择并协调用于编码任务的 AI 模型，Fusion Max 在 Artificial Analysis 的 Coding Agent Index 上得分仅次于使用 Opus 5.5 Max 运行的 Claude Code，且每项任务的成本低约 50%，并且根据 GitHub 的说法，耗时减少了 40%。这真的很酷。
2) Google's Gemini 4 Carbon has taken first place in the overall ranking. This is a surprise. Lowest hallucination rate, highest intelligence, basically the best model. The exciting part: Initially, it was a race between OpenAI and Anthropic. Then SpaceXSI re-entered the fray with Grok. Then Meta came along with Muse Wunderful, and now Google is suddenly back as well. What started as a two-horse race has become a five-horse race. Everything is wide open again, and the moat seems to be shrinking. Extremely exciting and surprising!
2) Google 的 Gemini 4 Carbon 在总排行榜中位居第一。这是一个惊喜。幻觉率最低、智能水平最高，基本上是最好的模型。令人兴奋的部分在于：起初是 OpenAI 和 Anthropic 之间的竞争。随后 SpaceX SI 带着 Grok 重返战场。接着 Meta 推出了 Muse Wunderful，现在 Google 也突然回归了。原本的两强争霸变成了五强混战。一切再次变得开放，护城河似乎正在缩小。极其令人兴奋且出乎意料！
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力