dlkf

11 ms·

61.
by dlkf 3y ago
Are you joking? Learning nonlinearities by hidden representations ranks among the most important scientific discoveries of the past fifty years.
62.
by dlkf 3y ago
That you have to reach for such obscure examples kind of illustrates my point, no? Geoff Hinton isn’t known for ”using Lisp at some earlier point in time.” It’s not my goal to denigrate Lisp. I think Lisp is great. I learned programming wit
63.
by dlkf 3y ago
Enlighten me. What has his research agenda resulted in? What did Hofstadter get right that others did not?
64.
by dlkf 3y ago
What did he get right? While Hinton, LeCun, and others were laying the foundations of deep learning in the 80s, Hofstadter was writing expert systems in Lisp.
65.
by dlkf 3y ago
> You actually need someone with vision and track record of doing right predictions and placing right technology bets. Hofstadter does not exactly fit this description.
66.
by dlkf 3y ago
> empirically investigated if philosophy professors engaged with ethics on a professional basis behave any morally better or, at least, more consistently with their expressed values than do non-ethicist professors The first investigation
67.
by dlkf 3y ago
Why is the title clickbait? What were you expecting - that the philosopher had proposed a model of consciousness that got experimental confirmation?
68.
by dlkf 3y ago
Lots of discoveries are confirmatory. We already believed relativity was true before we observed observing binary pulsars obeying it.
69.
by dlkf 3y ago
Imagine if Einstein, Newton, and Maxwell had used LSD! They could have reached the heights of such greats as ”Humphry Davy” and ”Sigmund Freud” instead of wallowing in mediocrity.
70.
by dlkf 3y ago
The experts have been confident that AI is 30 years out for about 70 years now.
71.
by dlkf 3y ago
Cade Metz is the same hack who tried to smear Scott Alexander. This guy is the personification of journalistic malpractice.
72.
by dlkf 3y ago
No there isn’t
73.
by dlkf 3y ago
An observational study with controls chosen at the whim of the PI is not the same as an RCT. Presumably you could implement different controls and observe the opposite conclusion. What scientists end up doing will be driven by the incentive
74.
by dlkf 3y ago
> if bilingualism is talking to your parents in kitchen Spanish without actually traveling anywhere, rather than learning the language of a culture that's alien to you so that you can spend an extended period of time there, is that
75.
by dlkf 4y ago
It’s simpler than that: typically there is an underlying factor that causes both the unemployment and the divorce. It could be depression, sustance abuse, poor impulse control, anxiety disorders, any number of things. You can’t control for
76.
by dlkf 4y ago
> One new study of 6,300 heterosexual couples found that all other factors being equal, men who were not working full time were 33% more likely to divorce in the following 12 months than husbands who did have full time jobs. You cannot c
77.
by dlkf 4y ago
The media is full of shit, but compared to the academic sociologists you’re referencing it’s not that bad.
78.
by dlkf 4y ago
This is a prescriptive definition that doesn’t fully correspond to common usage patterns: - Some sports bettors apply advanced analytics and win consistently. We can still sensibly use the expression ”sports gambling.” - Conversely, people
79.
by dlkf 4y ago
It’s almost as though news stories are calibrated to maximize clicks, rather than to convey information.
80.
by dlkf 4y ago
But crucially, they also decide who is in the union.
81.
by dlkf 4y ago
> Given that “the market” seems to have settled on LC as the only measure of skill That’s not true and I suspect you know it.
82.
by dlkf 4y ago
Would you trust unions over markets to determine which programmers are skilled? How would they determine that?
83.
by dlkf 4y ago
WGA west has about 10k members, whole US has about 15k. (This is from Google, please correct me if I got the wrong number.) By contrast, there are over a million people building software. Will this model scale up two orders of magnitude?
84.
by dlkf 4y ago
I would spend some time focusing on search fundamentals. When I search ”James Ellroy” only one of four hits is a James Ellroy book. If I can’t even find what I like, how can you recommend similar stuff?
85.
by dlkf 4y ago
You make a good point regarding incentive structures in pharma - but this is nothing like what OP was saying. OP is asking for a “Frommian psychoanalytic approach,” whatever that is. Your point about chronic pain increase is totally implaus
86.
by dlkf 4y ago
> Why do so many people want to check out of life completely by abusing medicine? This isn’t how it works. Few people, if any, want to “check out.” What happens is that people are exposed to opiods via one of two channels (they are presc
87.
by dlkf 4y ago
It’s not an argument, it’s just a positive claim. You’ve projected a bunch of normative angles on to it, but there are none. The more interesting follow up questions are to what degree is this true, how do we know, and do we think this is a
88.
by dlkf 4y ago
This question assumes nurture is the lever driving behaviour, but it’s not clear that’s true. What if instead you asked ”what personality traits would you expect to find in people who become billionaires?”
89.
by dlkf 4y ago
> why shouldn't it, with some finagling, also be able to do the dumb, simple things most programmers do for their jobs? Because those things, while dumb and simple, are not continuous in the way that visual art is. Subtle perturbati
90.
by dlkf 4y ago
And the award for dumbest headline of the year goes to

More ›