Occam's Razor (Marcus Hutter) | AI Podcast Clips

[@guafito]

the dislike is probably from someone that got cut by the Occam's Razor

[@Dayvit78]

I think the longer it stays at only 1 dislike, the more your explanation is the true explanation.

[@neolynxer]

It's from the Occam. He wants his razor back.

[@DeadInsideDave]

tbf i think someone just finds that tie a bit harsh imo

[@calebjaymes9710]

100p on a stack skrilla

[@DoctorLazertron]

Awesome tie

[@vincent_hall]

This works for theses as well: Get to the point quick good thesis. A thesis that is very long is generally thought to be not very good while a thesis that is very short is usually assumed to be very good.

[@jcinthailand]

Lex, thanks for bringing interesting subjects and guests to us unwashed masses. Seriously your interviews are some of the best I've seen.

[@JHBG1971]

This podcast is the exemplification of its topic. So good.

[@49prodog]

I came here by accident, but man, so many smart people in this world...Blows my mind!

[@brandonroberts13]

"it could have been different" is the biggest asumtion one can make.

[@crassflam8830]

Occam's fallacy (and sometimes occam's begged question)... This is definitely not the most important principle in science (all due respect to the venerable guest). The concept has value, but it's not even about simplicity, it's about elegance (which is more like *pound for pound* predictive power). In truth, reliability and accuracy of prediction is the be all and end all of science. physicists didn't accept general relativity and then quantum mechanics because they thought them simple, they were forced to accept unintuitive and complicated models specifically because that's what experimental evidence demonstrated. It's fallacious to assume that simpler explanations are therefore better (we need to check). And when it comes to formulating new models, we often have no idea before hand how many variables or components will be required to approximate (predict) some phenomenon. We should create our models based on what we see, not based on some expected simplicity.

[@gdn5001]

The issue is that any finite amount of data will generally underdetermine the model, and we need ampliative principles like explanation and Occam's razor to select the preferred model.

[@mariog1490]

This is really great. Ockham’s razor seems to be something science invokes pragmatically, but I’ve seen no objective and functional models of simplicity. Another problem that Ockham’s razor has is individuating entities. There is no canonical way of individuating entities, so how is it done? And more importantly, how do I individuate using Ockham’s razor? This process is rather presupposed by the razor. I mean, this is sloppy logic. Is baseball one thing? Is it many things? And how are you determining that? Can I get an objective model of this? You also have a problem with the epistemology itself that’s a perforative contradiction. If science has a reductionist ontology, then you have a problem of the science that’s done at these higher order functions. If you posit the reduction is real, how am I doing science at these higher levels which give me a truth of the reduction base. And if you say “no no, those higher order functions are real” then I posit to you that you have a complex ontology rather than a simple one.

[@cameron4814]

@7:36 my theory: why was symmetry chosen as a signal that represents sexual attractiveness? because our neural networks (brains) are compressing information, and, the more symmetrical things or faces appear, the easier it is for a brain to process, compress, store, and recall. further reading: is DNA an object oriented language? thanks to rotational symmetry, evolution must only write the code for one leg of the octopus.

[@Neavris]

What? How about there's 2 sexes, female and male... Our brains don't compress or store either, it adapts. It's called neuroplasticity, google it.

[@r-gart]

@@Neavris this has nothing to do with what he said

[@Neavris]

@@r-gart I guess my explanation wasn't simple enough. There are 2 sexes, obviously one will be attracted to another. What are the signs of health? Lack of defect. Like I said, there's no such thing as compression or storage in the brain's biology. It's adaptation. Gosh, just think about it.

[@cameron4814]

​@@r-gart soo.... it kinda seems like neavris missed the point. My above posted theory is just a different way to look at the compression theory in the video. The plasticity concept neavris might be referring to is sorta explained in the Implicit Association Test. kzbin.info/www/bejne/nqOcqXSZjNtqb5I

[@cameron4814]

​@@Neavris ​the idea i was trying to describe with this theory is maybe somewhere in Ray Kurzweil's book en.wikipedia.org/wiki/How_to_Create_a_Mind. it seems like what you are trying to demonstrate is "implicit Bias" explained in this video: kzbin.info/www/bejne/nqOcqXSZjNtqb5I

[@rkstr9965]

Occam's Razor is leverage of principals

[@SafeTrucking]

It's the principle of principals :)

[@JustinHappenstance]

weights and biases

[@wz-hdn8851]

Lex 2 words..Hydro Flask, And a question how does a Colo guy get ten min on the phone with you?...or better yet come to Colorado!

[@jaketapper1953]

Perhaps SAI is just a bunch of simple solution embedded in a complex program.

[@TactileTherapy]

I just released my second novel and the theme of Occam’s Razor is present throughout

[@AbcDef-tm7mv]

link?

[@TactileTherapy]

Abc Def www.amazon.com/Tactile-Therapy-Adjunct-Durell-Arrington-ebook/dp/B0856TCRMC/ref=mp_s_a_1_2?keywords=durell+arrington&qid=1582836831&sr=8-2

[@user-lr8ul1fl2z]

heurism!

[@sherrivonch6231]

Neat

[@burkebaby]

Lex, thanks for bringing interesting subjects and guests to us unwashed masses. Seriously your interviews are some of the best I've seen.