How much of the tech world is bullshit?

10 points by keeneyes ↗ HN

7 comments

[ 4.1 ms ] story [ 29.3 ms ] thread
I'll take "tech world" to mean the modern computer software & web milieu, made possible after ~1992 by Moore's law effects and commonly-available internet access for households. Vs. municipal water supplies, steel plows, electrical power, antibiotics, weather satellites, and Lotus 1-2-3 for MS-DOS are technology, and changed the world, but...

Hmm... A really new area of economic activity (vs. farming, mining, transportation, ...), often supported by huge economic bubbles, frequent indifference to viable business models, a now-large workforce often commanding pay 3x or more what "normal economy" workers make and often expecting annual pay growth >10% higher than inflation, generally underwhelming contribution to human welfare (shinier things to click on every year is not a basic human need, nor does it seem to overall improve human society that much), ...

I'll say 95%.

I think (but have no data to prove) that most is not bullshit. I assume that, because building something costs money and nobody is going to spend money if it has no use. Except for scammers and/ or overhyped technologies that are not actually necessary but used because managers like buzzwords like crypto, meta verse, AI ...
Consider human nature, and The Red Queen's Race.

I recall the incredible churn and waste of computer printer hardware ~~1985, when things were moving from daisy wheel printers to 9-pin printers to 24-pin printers to 300dpi laser printers to 600dpi laser printers. Sure, 600dpi laser output is more flexible & nicer than daisy-wheel output. But between all the effort of the replacement cycles, and working equipment with remaining service lives totaling many millions of years ending up in landfills... And let's face it - the main motive behind it was a zero-sum human nature "my tail fins are longer than yours" Red Queen's Race.

I ran a ML program to get an answer for this. I wanted to make sure that there would be proof of this experiment, so I used blockchain. I also made sure I used a turing complete language. The result of this analysis turned out to be 42%.
Roughly 99%,which has consistently been the failure rate of tech startups since Y2K.