Ben Horowitz: “To be a good CEO people have to want to work for you”
“What's the fundamental thing that you have to be able to do in order to be CEO? People have to want to work for you. That's it. Because if nobody wants to work for you… you can't actually do anything.”
Ben urges founders to consider the reasons people actually want to work for a person:
“Do they have good vision? Do they care about my interests and not just theirs? Are they somebody that I can get behind? These are the questions people ask of themselves.”
He points out that TV generally portrays CEOs as ruthless, but that’s actually the opposite of what you want to do because nobody wants to work for a person like that.
Video source: @UofCalifornia (2010)
Update for the small number of residents who normally have their waste collected from an alleyway:
Significant snowfall coupled with continued freezing temperatures leads to alleyways full of snow. Residents who normally receive alley service should bring their material to the curb to allow Sanitation Workers to access and collect their material.
“Because it is easier to monitor something with which you are familiar, if you have a choice you should delegate those activities you know best.” - Andy Grove, “High Output Management”
https://t.co/DtjVOnJXNl
Anthropic just released the receipts on a fear everyone’s been hand-waving.
52 junior engineers learning a new Python library. AI group scored 50% on comprehension tests. Manual coding group scored 67%. That’s a 17% gap on foundational skills, and debugging showed the steepest decline.
The productivity trade looked even worse. The AI group finished only two minutes faster on average, and that difference didn’t reach statistical significance. Several developers spent up to 30% of their time just composing queries.
Here’s what actually matters: they identified three failure patterns that predicted sub-40% scores. Fully delegating code to AI. Starting independently but progressively offloading work. Using AI as a debugging crutch without building understanding. All three share a common thread: removing the cognitive struggle that produces learning.
The high scorers (65%+) did something different. Some generated code first, then asked follow-up questions to understand what they’d produced. Others requested explanations alongside the code. The fastest group asked only conceptual questions, then coded independently while troubleshooting their own errors.
The gap between “AI makes you faster” and “AI helps you learn” turns out to be enormous. And most workflows are optimized entirely for the former.
@lydiahallie We are not seeing Cowork appear for users in our Claude team plan (standard seat or premium seat). Is this intended? https://t.co/TpOyreNQjP
This ITA software deck from 2003 reminded me of the great public hiring puzzles that ITA used for their software engineering roles around that time. It is understandably no longer on the ITA software site but Wayback Machine has them. I wonder how many Gemini 2.5 Pro could solve in one-shot.
https://t.co/Gkd4MVqpb0
https://t.co/3txgFieusA
“Because it is easier to monitor something with which you are familiar, if you have a choice you should delegate those activities you know best.”
- Andy Grove, “High Output Management”
@doodlestein@abacaj What were your prompts for this? I typically ask for mermaid diagrams but this it looks like both OpenAI models and Google models are getting quite a bit better with LaTeX.
Gemini 2.5 Pro really is impressive...
"Can you create a simple 3d car simulator with Three.js in a single HTML? Please add clouds, mountains, a road, some trees and a train going around. Make sure it works on mobile."
Much of the wisdom for managing humans applies to managing AI as well.
“Because it is easier to monitor something with which you are familiar, if you have a choice you should delegate those activities you know best.”
- Andy Grove in “High Output Management”
People with deeper software knowledge are better vibe coders.
This feels obvious. Demand for engineers might decrease but claiming Computer Science is no longer useful is plain wrong.
This piece by @gruber on Apple’s botched “Intelligence” has me completely floored. I’ve read Daring Fireball for 23 years and he has never come close to this level of criticism about the company. It reads like an obituary, and I agree with every word. https://t.co/KF15x417AG
The AI skills gap isn't about coding or prompt engineering—it's about systems thinking.
Companies scrambling to hire LLM experts are missing the point. We need people who understand both the business process and how AI can transform it.
With Hantavirus in the news right now due to recent tragic news of Gene Hackman's wife, I am reminded of the stories from @threatdetective of the first identification of a Hantavirus outbreak in the United States in the early 90s. Mark went on to serve as an EIS officer with the CDC and later was a part of https://t.co/kjHFACUuN5.
In 2022 Mark was interviewed for a TED podcast and spoke about his experience of doing this essential research: https://t.co/APaOeZho3P
I can't believe they've just cancelled the Epidemic Intelligence Service program at CDC.
My father was an EIS officer:
https://t.co/fqv3JcZTax
@Farzad_MD's thread below gives you a sense of the kind of people in this elite program to train the best & brightest epidemiologists.
I wish more people were taking seriously the possibility that @ezraklein and the leadership of the AI labs are raising: that AGI is a real possibility in the near future.
You don't have to buy it yourself, but leaders & policymakers need to consider the possibility it is true.
one famous photographer I know once confided, “it’s really just taking a shit load of photographs and then picking the best one.” Reminded me that quality is often a function of quantity of options, time and tools for creation, and taste.