2 ms·
As a serial entrepreneur and founder in the technology realm my past matches his claim exactly. The first founded iteration was spent observing and learning wh
by bokohut 4y ago
As a serial entrepreneur and founder in the technology realm my past matches his claim exactly.
The first founded iteration was spent observing and learning what not to do and over time grew to around 100 employees at peak. The second founded iteration I re-architected and re-coded the entire platform in about four years automating out all possible employees before going live which peaked at 7 employees under my ownership. And as they say "Third time's a charm". ;)
From my own personal past experiences, and YMMV, people are the problem. Computers only make mistakes that the software allows which comes from the human writing the code of course until GPT3 starts coding:) Computers do not know of opinions, politics, emotions, illness, image, etc and if businesses were to track and quantify this as to the impact of said business then many eye opening considerations would occur. Just because "It has always been this way" does not make it correct or, from my direct experiences, efficient. Per Edison: "There is always a better way."
I am the exception however as I have been doing real hands-on work since about age seven and have always been the first person to jump into anything to get my hands dirty learning whether that involves ICE engines, electricity, plumbing, digging, and of course anything tech related. I love to learn and have a passion for solving the hard problems which is my downfall but I don't do it for the money since that cannot buy health nor happiness.
- treis 4y agoHe's right about automation being good but misses the dichotomy between capturing value and creating value. Instagram was worth billions because they won in a market that is lucrative and tends towards winner take all. To use a somewhat bad analogy, you don't create a billion dollars of value by winning the powerball lottery.