Comment by MyHypatia | Hacker News Reader

Comment by MyHypatia | Hacker News Reader

This article is attacking a strawman. Newton, Einstein, Ramanujan, etc. were geniuses. It is not controversial to say that. They did not work in complete isolation, and were in regular contact with the scientific community at that time. Acknowledging that even the "lone geniuses" did not work in complete isolation does not detract from their extraordinary contributions. Even lone geniuses have to communicate their ideas to a community of people capable of understanding them, using language and analogies to make their novel ideas understood and adopted. I am sure there are many lone geniuses that we don't know about because they were unable to effectively communicate their ideas to the right audience.

Throughout history there have been many examples of people who want the glow of being a "lone genius" to join the ranks of Einstein and Newton. In order to accomplish this they have downplayed the work of others. It is easier and more compelling for journalists and screenwriters to run with the lone genius story, because it makes the narrative less complicated. I appreciate that we examine a more complete picture of what it takes to accomplish extraordinary goals, whether it's a solitary or team effort.

I personally liked the series Connections that showed off that many of these guys did not live in a vacuum. https://www.imdb.com/title/tt0078588

Basically mostly everything is fairly incremental but every once and awhile you get someone who connects 3 or 4 different fields together and something really cool happens. If you work by yourself you will become odd and stale. Bouncing ideas off others helps weed out junk and maybe even help you get a better understanding of what is going on. Or they may ask a question you did not think of and woosh you found something even more cool.

I had an idea a few days ago that most ML seems to be mostly curve fitting (except when it isnt). But what if you could turn that network back into a formula instead of a set of nodes with weights and particular follow rules. Those sets of formulas could yield some interesting ideas, I think, if you could compare them. Like there may be a set of formulas that are just hiding there but are smeared across 1000 nodes. Not even sure where to begin on that. But an interesting idea none the less. One of the smart ones could see the idea and also figure out where to start and tease out the idea until it is done.

Einstein was ostracized from academia. He is the quintessential lone genius.

Einstein's 1905 papers were ignored, even by his own teacher Minkowski. Einstein was considered an arrogant academic failure. The "lone genius" was that Einstein knew that he was right and and smarter than his peers and superiors, so he kept working.

Max Planck was the one who identified Einstein's Brownian motion paper as significant. Planck knew that Einstein was onto something and that if Planck continued "quantum"/atom research history would acknowledge Einstein's previous work.

Planck didn't want to needlessly share his glory so he pushed Einstein's relativity work. Planck elevated Einstein's relativity work in an effort to remove Einstein from the quantum.

Einstein betting on himself continued. He promised Mileva his Nobel prize money in 1918 long before it was certain in 1921.

> Einstein was ostracized from academia. He is the quintessential lone genius.

This is all true but most of the "lone genius" narrative paint people in an almost supernatural light.

Einstein work was widely recognized later on. Also, he interacted with plenty of people and read a lot and he never denied that.

The model I like to think of this is that intelligence is speed and knowledge is distance, but you start at the peak of a mountain and the farther you go the slope is increasingly getting sharper, so you keep getting faster.

Terrance Tao is a genius. He can go from sub-field in math to sub-field in math and solve a problem that stumped someone dedicated in that sub-field for years. But he is only able to solve those problems because of the understanding that the sub-field expert developed over those years in addition to the broad context that he has developed jumping from sub-field to sub-field.

He started with a lot of speed. As he's worked with more individuals, he's covered significantly more ground that other people, which has compounded to make him unbelievably fast.

One of the main take-aways from this model, is that if we are in an economy that only allows the wealthy Einstein's to flourish, not only are we losing the poor Einstein's we're quite literally slowing Einstein down.

We must do everything we can to ensure that we find smart dedicated people and allow them to thrive.

We also forget, if Newton were truly a "Lone Genius"... then we would be using Leibniz's calculus.

How many millenia did it take for heliocentric theory to be adopted? Yet we attribute it to Copernicus, as he managed to convince enough people to accept it.

Also, there's Leibniz, for Newton :-)

For a fair number of these figures, there's not just the people they communicated with, the prior work they drew on, but often competitors, people they knew were working on the same or similar problems. This is particularly the case with technical inventions.

Reply on news.ycombinator.com