[March/2024 | Tech] Breaking Smart – how software is eating the world – Venkatesh Rao

Carlota Perez – installation vs deployment phase of technology

铁路革命的副作用 – 城市化,超市连锁,肉食普及


Hackers ethos vs credential based ( 学位)

Credential based – the org man


Pragmatist vs Pureist – Pragmatist wins (自下而上)

Purist is visioning, single architect make sure everything in the punch card is right the first time;

Pragmatist is hands on iteration – trial and error


Authoritarian high modernism – 根除任何不符合意识形态的自上而下的规划 – 人定胜天的去处社会环境中的不安定因素

Well planned suburb vs slum

作者认为hacker用这些unplanned factor来高效产生作品

Democracy in programming – reconciling conflicting authoritarian views

Innovation – ideas illegible to all but few (contrarian beliefs that works)

Agile dev – combine decisiveness in authoritarian and pluralistic vision in democracy

Increased info availability & lower friction makes an industry hacker friendly



Abundance mindset in software development via trial and error

Purist driven by architect with scarce mindset (limited computing resources)

Traditional rough consensus means creating alignment between competing autocrats

Rough concensus actually means path of most fertility – lazy evaluation/optionality – avoid constraining future degrees of freedom

Waterfall model – Authoritarian – vision to implementation

In agile model, execution is ahead of vision – creating serendipity. Waterfall is the other way around


Perpetual beta – release early release often (RERO) – 作者在此尬吹


Software Fork – if you disagree, then fork it – 没有民主化的不同意见的死锁

Worse is better – most important is allowing fast experimentation

Jeff Bezos’s two pizza rule (dev team no more than a dozen)

IBM OS/2 developments – interchangeable programmer – estimated ten lines per day – if more than ten lines considered as irresponsible


Luddites in 19th century – 科技浪潮下的反抗者


spillover effect – steam engine tinkered & improved by hundreds of minor innovations after patent expire

Collective invention phase – 共同把蒸汽机改良

Book “Levers of Riches”

Pastoralist 看不到科技革命给社会带来正反馈循环

另一种 – 举国体制的堆科技 – 短期见效但长期过多institution 而阻碍创新 – 例如阿波罗登月


对于1950 会飞的汽车的向往 – 作者认为也是因为当时世博会的宣传造成的田园牧歌式的愿景 – 被当时1950的aerospace tech发展主导


Tendency to view adaptation as degeneracy

Car centric to smart phone centric, urban transportation that’s a shift

Technology doesn’t limit human agency, but it becomes the expression of human agency, empowers the human, essentially by technology

Power is zero sum by definition, and the public signage in multiple language situations are a example of such battleground

Bits > Atoms – virtualization / containerization

Software eats Hardware – transient atoms, permanent bits

在atomic的世界 – 更加零和竞争,bits的世界不那么竞争 – 点播频道的切分 vs. 推特思想的共享

作者认为互联网是一种解决问题的新方法. 旧方法: 定义问题,获取资源,用资源解决问题. 新方法(Serendipity): 在解决问题的思路中沉浸,然后快速迭代发现新思路,新思路效果好的话重仓出击.

Low stake experimentation – e.g. EDA in data science

Open source movement

Extractive institution – not just to solve problem and extract wealth, but to extract wealth indefinitely – 现金流 (某种意义上吸血制度的既得利益者)

此类既得利益者越多,新的解决问题越难(盘根错节的利益关系) – e.g. 美国医疗 – 最终状态是政令无效化

拟人化的怀旧主义会让政府拯救某些entity – e.g. American Middle Class, 尽管其可能已经被时代所淘汰

Gall’s law – complex systems harder to patch up if not working

Free as in beer vs. Free as in speech – 免费 vs. 自由


If you related to a coworker outside of worker (e.g. tennis buddy) rather than just colleague, this weakens the authority of the organization. 

没有自由的信息流的地方,占领组织内部高地(innermost sanctums)是好的; 有自由的信息流的地方, 在新领域的边界是好的


Leave a Reply