How Elon Musk Thinks: Reasoning from First Principles.

http://www.oninnovation.com/videos/detail.aspx?video=1251

by jasonjackson • 15 years ago

94 41 15 years ago

oninnovation.com

rhnoble• 15 years ago

I didn't think much of the video, but I liked the "Interactive Transcript" widget on the right side of the video.

javert6• 15 years ago

The transcript on the right is the same as the video, and easier to read.

I think this is great overall. However, I think he's wrong at the very end. People aren't just hardware and software; they're hardware, software, and the thinking they have done (call it "free will" if you want). That accounts for a lot of the differences in people.

bfe• 15 years ago

Or it's software that keeps writing more software, thereby modifying itself and sometimes its hardware.

javert6• 15 years ago

Yes, I think that's probably more accurate. Anyway, the point is that the end of the interview is a fallacious denial of that which is commonly called free will.

abdullahkhalids• 15 years ago

Or you could listen to what biologists are saying..

javert6• 15 years ago

Well, if you're right, and we can't excecise thought independent from our genes and external influences, then all our thoughts, mistaken or correct, are determined anyway.

So who are you to say that you're right and I'm wrong? Maybe your determined thoughts happen to be flawed, and mine happen to be correct.

In other words, a complete and utter denial of "free will" (which is vaguely defined, hence my "complete and utter" qualifier) is a self-defeating argument.

Empact• 15 years ago

As far as the nature vs. nurture division, there's a good amount of new evidence from twin studies and adoption studies which back up Elon's conclusion about the heritability of traits. Bryan Caplan covers some of this in his new book "Selfish Reasons to Have More Kids."

http://www.dailymotion.com/video/xj9b1h_selfish-reasons-to-h...

radu_floricica• 15 years ago

You don't program much in lisp, do you? :) Software can be adaptable and self-writing, too.

javert6• 15 years ago

Oh, I totally agree with you. Your comment would be better directed to Elon Musk; it was his analogy. :-)

icarus_drowning2• 15 years ago

I think Musk is right to point out that many people don't reason from first principles-- basically, that they consciously decide not to think about something, but to allow other's thinking to guide their thoughts. ("It has always been done this way")

I recently began teaching in a school where students are required to write extensively in every class, including "extras" like PE, Music, or Art. I've found that, generally, students who have been subjected to that system for a number of years are far better at reasoning (and expressing that reasoning, naturally) than those who are new to the school. (It is astonishing how quickly new students, who inevitably struggle with complex reasoning at the beginning of the school year have made leaps and bounds in this area, and are far more skilled at verbal reasoning by the end of the school year).

All of which is to say: I think that this style of thinking is greatly buttressed by enhanced linguistic ability, as language is critical to higher-level thinking.

Very interesting video, and the interactive transcript was extremely cool.

jleader• 15 years ago

I'm a little suspicious of people who claim to think "everything" through from "first principles" (whatever those are).

Where do you get your first principles from?

If you get them from observing the world around you, how do you decide what parts to pay more or less attention to?

While you're deciding things from first principles (and coming up with those first principles), how do you make decisions in the mean time?

Note that in physics, computing things "from first principles" is often done to attempt to recreate a result that's already been reached by other approaches, to attempt to verify that you've got the right set of first principles.

icarus_drowning2• 15 years ago

Many people don't have first principles to begin thinking from. While you're certainly right that the quality of one's first principles is at issue in judging their conclusion, the mere fact that so many people don't even bother to try to think presuppositionally seems to be more of the point of the video, IIRC.

reso• 15 years ago

Well you have to decide on what your axioms are to start with. If you're developing a product you might decide your axioms are: "It must ship on time, its quality must be this high, it must sell x number of units", and then reason from there.

jhuckestein2• 15 years ago

You make assessments based on everything else you know (I think he mentions that at one point).

Of course this isn't unbiased, but at least you're thinking for yourself.

I don't think it's just about thinking from first principles, it's also about being able to change your first principles as you go along and understand that you need to reassess everything else that depends on it.

tedkimble• 15 years ago

Your first principles are the constituents of your worldview. Put in reverse: You worldview is the sum of your first principles.

Unfortunately, understanding your world view is a daunting task. This is why most people don't start with first principles. They have yet to know them.

dredmorbius• 15 years ago

That particular debate goes back to Descartes vs. Bacon.

Descartes was a logician. Bacon was an empiricist, Descarte a rationalist reasoning from first principles, beginning with "cogito ergo sum": I think, therefore I am.

nazgulnarsil• 15 years ago

Your reward system is hard coded. So you pay attention to the sensory data correlates that result in the highest preference satisfaction. This gets compounded until you've bootstrapped up to having a reasonable causal model of the things that affect you the most.

shalmanese2• 15 years ago

I think it's ironic that Elon Musk starts off by railing against reasoning by analogy and then ends up with analogizing human behavior using the technological metaphor of his age.

shalmanese2• 15 years ago

More specifically, it's our tendency to rely on the current technological metaphor to analogize the brain that contributes to misunderstanding and false equivalency: http://www.cl.cam.ac.uk/~jgd1000/metaphors.pdf

jhuckestein2• 15 years ago

In the end he says that after having had 5 kids he believes that nature has a much stronger impact on people's personalities than nurture. That surprised me, I wish he could have elaborated on it.

klenwell• 15 years ago

Steven Pinker's The Blank Slate is a spirited and informed articulation of this view.

MikeCapone• 15 years ago

That book is excellent. In the same vein, you can read The Moral Animal by Robert Wright, or if you want to dig deeper, check out the studies cited in those books.

chubs• 15 years ago

I think personality comes from nature, but character comes (initially) from nurture.

pshapiro• 15 years ago

If it's really a first principle then there should be only one of them. This is very important if you want to understand the world completely.

stmartin• 15 years ago

Oh the arrogance....the pride... my head hurts.

"First principles"... Must be that our interpretation of physics is THE first principles.

lutorm• 15 years ago

From the context in which he's saying it, I don't think he means it in the sense you're thinking. I thought his reference to physics was an analogy, not an exact statement. I took it to mean more "don't assume that what you hear is true, if what you conclude based on the things you know is different, you should go with that."

ThomPete• 15 years ago

Say basically ayn rands "existence exist"

That's not first principles just to be clear.

Jach• 15 years ago

I'm all down for first-principles reasoning, but I don't like that term for some reason. For some reason it makes me think it's unduly favoring deduction and making assumptions whenever possible, even though it's not. If you start from good first principles, like the axioms of probability theory, all of a sudden you get inductive inference and deductive as a special case. Yay! Nevertheless I'd still more enjoy shouting "Baaaaaaaayes!" from the rooftops than "First Principles!"

jamesrcole• 15 years ago

I think this ability is really valuable, but there's so many barriers to people getting experience with it.

To do it, you have to venture into the unknown without a map. You have to make the map yourself, and you have to figure out how to do that. You don't know how long that will take, but it is likely to take quite a while. While you're working at it you will not have a very clear understanding of exactly what your position is, and you won't be able to articulate it clearly to others. And of course you will have to go against what "everybody knows" (but really just think they know).

There's so many barriers to this that come from our social and institutional norms, where it's generally expected that you can explain what you're doing, that you can say why it is better, that you can estimate how long it will take, etc, and where it's frowned upon if you can't do these things. And where it's generally frowned upon to "have the arrogance" to "go against" what people of high-standing came up with or take to be true.

How Elon Musk Thinks: Reasoning from First Principles.

Related Stories