Dan Luu 写系统问题时，习惯先测量、再对照、最后才下结论。把「Steve Ballmer was an underrated CEO」放到智能体、评测和线上系统里，真正要问的是：默认做法会不会系统性失败。下面用中文整理成可执行的工程笔记，去掉原站导航和无关链接。

Criticisms of Ballmer

There's a common narrative that Microsoft was moribund under Steve Ballmer and then later saved by the miraculous leadership of Satya Nadella. This is the dominant narrative in every online discussion about the topic I've seen and it's a commonly expressed belief "in real life" as well. While I don't have anything negative to say about Nadella's leadership in this post, this narrative underrates Ballmer's role in Microsoft's success. Not only did Microsoft

In addition to overseeing deep investments in areas that people would later credit Nadella for, Ballmer set Nadella up for success by clearing out political barriers for any successor. Much like Gary Bernhardt's talk, which was panned because he made the problem statement and solution so obvious that people didn't realize they'd learned something non-trivial , Ballmer set up Microsoft for future success so effectively that it's easy to criticize him for be

Ballmer's wins

For people who weren't around before the turn of the century, in the 90s, Microsoft used to be considered the biggest, baddest, company in town. But it wasn't long before people's opinions on Microsoft changed — by 2007, many people thought of Microsoft as the next IBM and Paul Graham wrote Microsoft is Dead, in which he noted that Microsoft being considered effective was ancient history:

A few days ago I suddenly realized Microsoft was dead. I was talking to a young startup founder about how Google was different from Yahoo. I said that Yahoo had been warped from the start by their fear of Microsoft . That was why they'd positioned themselves as a "media company" instead of a technology company. Then I looked at his face and realized he didn't understand. It was as if I'd told him how much girls liked Barry Manilow in the mid 80s. Barry who

Microsoft's misses under Ballmer

Microsoft? He didn't say anything, but I could tell he didn't quite believe anyone would be frightened of them.

These kinds of comments often came with comments that Microsoft's revenue was destined to fall, such as these comments by Graham:

The big picture

Actors and musicians occasionally make comebacks, but technology companies almost never do. Technology companies are projectiles. And because of that you can call them dead long before any problems show up on the balance sheet. Relevance may lead revenues by five or even ten years.

Graham names Google and the web as primary causes of Microsoft's death, which we'll discuss later. Although Graham doesn't name Ballmer or note his influence in Microsoft is Dead, Ballmer has been a favorite punching bag of techies for decades. Ballmer came up on the business side of things and later became EVP of Sales and Support; techies love belittling non-technical folks in tech 1 . A common criticism, then and now, is that Ballmer didn't understand t

Appendix: Microsoft's relevance under Ballmer

Dump Ballmer. Fire 40% of the workforce starting with the loser online services (they are never going to get any better). Reinvest the billions in start-up opportunities within the puget sound that can be accretive to MSFT and acquisition targets … Reset Windows – Desktop and Tablet. Get serious about business cloud (like Salesforce …)

To the extent that Ballmer defended himself, it was by pointing out that the market appeared to be undervaluing Microsoft . Ballmer noted that Microsoft's market cap at the time was extremely low relative to its fundamentals/financials relative to Amazon, Google, Apple, Oracle, IBM, and Salesforce. This seems to have been a fair assessment by Ballmer as Microsoft has outperformed all of those companies since then .

Appendix: my losing bet against Microsoft

When Microsoft's market cap took off after Nadella became CEO, it was only natural the narrative would be that Ballmer was killing Microsoft and that the company was struggling until Nadella turned it around. You can pick other discussions if you want, but just for example, if we look at the most recent time Microsoft is Dead hit #1 on HN, a quick ctrl+F has Ballmer's name showing up 24 times. Ballmer has some defenders, but the standard narrative that Bal

And if you look on less well informed forums, like Twitter or Reddit, you see the same attacks, but Ballmer has fewer defenders. On Twitter, when I search for "Ballmer", the first four results are unambiguously making fun of Ballmer. The fifth hit could go either way, but from the comments, seems to generally be taken as making of Ballmer, and as I far as I scrolled down, all but one of the remaining videos was making fun of Ballmer (the one that wasn't wa

值得单独记下的观察

• The enterprise sales arm, built and matured under Ballmer, was and is critical to the success of Azure and Office
• Like Azure, I don't personally like these products, but if Microsoft were to split up into major business units, the enterprise software suite is the business unit that could possibly rival Azure in market cap
• 2007: Microsoft releases LINQ, still fairly nice by in-use-by-practitioners standards today
• I'm not a fan of patents, but I would guess that the reason autocomplete/inference works fairly well in Excel and basically doesn't work at all in Google Sheets is that MS has a patent on this based on this work
• Heath Borders notes that Erich Gamma, hired in 2011, was highly influential here

落地时建议先做的 5 件事

1. 用自己的真实负载测，而不是只用公开榜或厂商数字。
2. 把评测设计成能抓到失败模式：平均分好看但尾部崩溃，仍然算失败。
3. 智能体默认不会好好用测试；要写进流程，而不是写在口头规范里。
4. 性能和正确性都要有基线，改模型或改语言前后必须能对比。
5. 结论写成可回滚的决策：哪一版配置、哪一版评测集、谁签字。

和智能体产品怎么接

龙虾PRO做 OpenClaw 落地时，同样吃「先测量再扩面」这条纪律：技能、数字员工和网关都要有可复现评测，而不是只看一次演示通过。

本文侧重全链路风控方法论。落地时请用自身业务单据做回放验证，不要把示例阈值直接当生产策略。 相关：风控体检 · 方案资源