Radim

37 ms·

301.
by Radim 8y ago
Very frustrating for companies who actually do use machine learning to solve a problem! Hard to rise above the background "AI noise" of well-funded marketing. And people are already fed up with all the AI marketing bullshit; &quot
302.
by Radim 8y ago
Left to its own devices and somewhat bizarrely, capitalism evolves strategies that include "social stability". Long-term prosperity becomes a part of the equation, with despotic wealth gaps an obstacle to that. Nobody likes their
303.
by Radim 8y ago
It's a cowardly way of saying "I want that guy's money, but am too weak to take it myself". The fundamental problem with such redistribution schemes is: Who gets to decide who's weak and who's the villain? Cert
304.
by Radim 8y ago
Surprised nobody mentioned Slate yet: (probably) not as powerful, but free, open source, self-hosted and fully within your control. https://github.com/lord/slate What would be the main differentiator of DeveloperHub vs
305.
Elon Musk at Joe Rogan: AI, SpaceX, the Future(youtube.com)
5 points by Radim 8y ago|0 comments
306.
by Radim 8y ago
Note that like the article says, memory-mapping (mmap) vectors from disk already works in Gensim: https://radimrehurek.com/gensim/models/keyedvectors.html The low memory footprint in LMDB is interesting. There wer
307.
by Radim 8y ago
In the general sense of "evolution" (moving between states): absolutely! That's pretty much what physics is. For an extreme example, see Wolfram's theory in New Kind of Science, a (controversial) hypothesis about how law
308.
by Radim 8y ago
The tricky part is that your first "physics" (human models) is also a part of the physical world. The human concepts and equations affect what humans do, which changes the physical world. There's a bit of a feedback loop. Con
309.
by Radim 8y ago
This smells of false dichotomy. Does a suicide terrorist choose to die for his individual "good" (sic), or for his ideology = the "group good"? In reality, an "individual" and his social context are inseparable
310.
by Radim 8y ago
I wouldn't single out "sentient beings" as any kind of exception, just because we're aware of that "homunculus between our ears" whispering to us at all times. Our self-awareness and knack for introspection thr
311.
by Radim 8y ago
What makes you say that? Yes, it is a goal. In fact, it appears to be THE goal, whether our human perspective likes it or not. Societies follow the same optimization principles as everything else. If the topics of self-organization, similar
312.
by Radim 8y ago
Thanks. Then we're in disagreement :-) For me, people are a part of the whole system, following the same rules as everything else in the universe. No special exceptions, no "get out of jail free" card. (by the way, it's
313.
by Radim 8y ago
The answer is literally the part after the comma you left out. Entropy maximization.
314.
by Radim 8y ago
I am familiar with both; what's the connection in particular? That patterns organize along least-energy principles and propagate by extracting resources from surrounding energy gradients is almost a tautology. I'm genuinely curiou
315.
by Radim 8y ago
I never heard of this term "detectologists… a blend between archaeologists and detectorists", but it makes sense. There's such a hunger for understanding other cultures and our history. On a this-is-how-people-actually-lived
316.
by Radim 8y ago
People fall into rituals and procedures for reasons that are not necessarily obvious to them. Why should we expect to be able to determine what is and isn't "bullshit"? Wouldn't you need to know where the whole endeavour
317.
by Radim 8y ago
Fascinating presentation! Joel is such a meme factory :-) Has anyone done a "rejected memes" section at the end of a technical talk before? On topic: I applaud the effort to set things straight, but I'm afraid that ship has s
318.
by Radim 8y ago
That sounds really fascinating. Where could an interested layman read more about your (and related) work? I find it very hard to find material in that space, between the pop-culture "Check out These 10 Things The Ancient Sumerians Did!
319.
by Radim 8y ago
That I can agree with. It is your earlier posts where you say "Money … should … not even try to touch meaning" and "The question is not if these jobs make money … but if they contribute anything to the society" that rais
320.
by Radim 8y ago
"Valuable contribution according to coldtea" is not a thing. And without irony, I'll add thankfully . What is a good vs bad contribution to society is played out over periods of time and using millions of independents actor
321.
by Radim 8y ago
False dichotomy. Money is the medium through which societies (attempt to) measure such contributions and meaning. And this includes abstract and indirect ways, like art, fun, future potential, etc.
322.
by Radim 8y ago
Isn't that the point though? Europe has experienced first hand not only "national socialism" (nazism), but also another type of socialism — communism. Still legal though, and going strong in Europe. It's a grim compariso
323.
by Radim 8y ago
For a short while, maybe. "Excellent talent" means a strong inner drive, it takes something like an obsession to put in the grind to get there. The feeling of contributing and creating and playfulness and attacking new frontiers i
324.
by Radim 8y ago
Unlikely. No "talent" would agree to waste their life in such a scenario, worrying about "being found out for not contributing to the bottom line". That's kinda what makes them a talent — almost a tautology.
325.
by Radim 8y ago
Try walking in the employer's shoes: "can this person learn what we need them to learn?" is notoriously difficult to assess. While the risk of making a bad hire is tremendous (deadly for smaller companies), the upside's
326.
by Radim 8y ago
Is it common to refer to ancient American civilizations as "US history"? Here in Europe, US is understood to mean United States, the country (declaration of independence and all that jazz). America(s) is a separate concept.
327.
by Radim 8y ago
Understandable, but also somewhat proportional? There's 250 years of US history vs thousands of years for many European countries. Of course you need more time to cover that. On topic: Iran is an absolutely stunning country. Second to
328.
by Radim 8y ago
Speaking as both a ML researcher and an applied ML business owner (one who hired stared at one point — hi Piotr :), I respectfully disagree. The replication crisis in "AI" may not be as bad as psychology (I wouldn't know),
329.
by Radim 8y ago
Probability of propagation is just a proxy for consuming energy gradients more efficiently.
330.
by Radim 8y ago
"As a species we seem to be a cancer on the planet." This platitude is actually interesting in this context. It connects multiple threads: * One of the previous major extinctions was the Oxygen Holocaust [1], caused by photosynt

More ›