imh

33 ms·

271.
by imh 9y ago
Can't it be both a technology problem and an accountability one? The tech problem might take ages to get solved without accountability forcing it. If we were accountable, we'd be forced to make better tools for writing reliable so
272.
by imh 9y ago
That's fair. Communication is hard. I edited my original comment along the lines of your suggestion. I hope that it's still clear and concise and now does a better job of not taking sides. I think sometimes I go overboard in tryin
273.
by imh 9y ago
I'm trying not to implicitly disagree or agree with either side. The point here regards the meta discussion about disagreement and arguers not respecting and trying to understand the other side before jumping on them. I'll try to
274.
by imh 9y ago
>Yes, we disagree constantly. But what makes our disagreements so toxic is that we refuse to make eye contact with our opponents, or try to see things as they might, or find some middle ground. This bothers me so much. In my social circl
275.
by imh 9y ago
Here's a more in depth version of the proposal https://static1.squarespace.com/static/599c23b2e6f2e1aeb8d35... It answers many of the questions people are asking here.
276.
by imh 9y ago
Not necessarily. If the costs drop as a result of increased efficiency and that's what leads to more sales, it would be unfair to say that not hiring extra workers is like lost jobs.
277.
by imh 9y ago
As a matter of fact, betting markets are a fantastic way of aggregating opinion in the wisdom of the crowds literature. If you have some option that will pay out $1 if something is true, and you think there are 70% odds of it happening, you
278.
by imh 9y ago
It makes me wonder why identity has to be a centralized government thing. For most purposes, my google account is my primary identity. If I forget a password, resets go there, so it's my foundational identity online. Per-purpose identi
279.
by imh 9y ago
SQL is neither FP nor OO! It's not a general purpose language though, so it may not be a great answer. It's still amazingly useful and incredibly prevalent.
280.
by imh 9y ago
>1st: If she sought comment from Rebecca Campbell, it's not mentioned anywhere in the piece. An article in a magazine like The Atlantic should include, at the very least (we reached out to Rebecca Campbell, but she declined to comme
281.
by imh 9y ago
As a matter of fact, I can do even better than 0-50% of cases! Consuming widgets has been shown to lead to recovery in 0-100% of cases! It isn't as simple as you suggest.
282.
by imh 9y ago
>For a wide range of common statistical tests, transitioning from a P value threshold of α = 0.05 to α = 0.005 while maintaining 80% power would require an increase in sample sizes of about 70%. This proposal is a great pragmatic step
283.
by imh 9y ago
Does anyone understand the intent of the foregone conclusion doctrine? It seems like if something's a foregone conclusion, instead of overriding 5th amendment rights, they shouldn't even need whatever's being hidden in the fi
284.
by imh 9y ago
I've been reading through this ( https://www.cs.cmu.edu/~fp/courses/15317-f00/handouts/logic.... ), and have found it way more useful and fun. I expect it's much closer to what the HN crowd would
285.
by imh 9y ago
There's an interesting point lower in the thread. Is a landmine an autonomous weapon? They've been pretty awful for humanity.
286.
by imh 9y ago
>No one owes James a charitable reading. A charitable reading isn't something that's owed. It's something that almost universally helps discourse. Communication is hard. We're always willing to give Us a charitable
287.
by imh 9y ago
Instead of shooting for a mega-exit, I think it's about using the business to just enable a certain lifestyle for the founders/employees.
288.
by imh 9y ago
A scalar is just a magnitude (call it amount). A vector is a direction with magnitude (call it length). A bivector is a pair of directions with magnitude (call it area). You can keep going and make something with volume and even more into h
289.
by imh 9y ago
The Martian was fantastic and the author is a programmer. I recently got hooked on the Bobiverse books and their author is a programmer too. In fact, I wholly recommend anyone reading this to check both of those out :) Super fun books
290.
by imh 9y ago
Thanks for the answer :) Great work here, btw. It's refreshing to see emphasis on "label some damned data" and work towards making that easy.
291.
by imh 9y ago
Since syllogism is participating in this thread, what kind of active learning are you using? I'm always hesitant to use anything except for IWAL since most of the more common ones aren't actually consistent. Even then, then payoff
292.
by imh 9y ago
Cool thanks for confirming/clarifying :)
293.
by imh 9y ago
How does that relate to the privileged communication stuff? It seems like there's a lot he recalls, but not outside privileged communication, so he says "I don't recall?"
294.
by imh 9y ago
"I don't recall" seems to be an unusual statement here. It seems like he's using it more like, "I don't recall it outside of certain things I don't have to/won't answer." Am I reading that r
295.
by imh 9y ago
>Q: You're saying it's not unusual to pay somebody a bonus of over $100 million? >A: I don't -- I don't think characterizing it as a bonus is correct. I think it's more like a start-up compensation, which woul
296.
by imh 9y ago
This has all of my favorite media failings in one! It's like bingo! We have the game of telephone, with outlets reporting on outlets reporting on outlets, each diverging further from the primary source. We have attention grabbing headl
297.
by imh 9y ago
In the multiclass problem, they have the same number of degrees of parameters. You can see this by choosing a reference class c_y. The NB parameter for class c_x and feature w_j is f_jxy = log(p(w_j|c_x)/p(w_j|c_y)). If you want a new
298.
by imh 9y ago
Heh, nope you're right I was writing to quickly. I meant to say to take a log of of the probability term in p(c_1|w). So you take p(c_1|w) = 1/(1+stuff) and turn it into p(c_1|w) = 1/(1+exp(log(stuff))). stuff is a product, s
299.
by imh 9y ago
Not true. Let's derive it. Here's NB for a binary classification: p(c|w)*p(w) = p(w|c) * p(c) p(c|w) = p(w|c) * p(c) / p(w) p(c|w) = p(w|c) * p(c) / [sum_i p(w|c_i) * p(c_i)] Let's look at the pro
300.
by imh 9y ago
The (conditional) independence assumptions it makes are pretty strong and usually inaccurate. In the text example, consider "Steph Curry" and "Warriors." These are going to show up very frequently together, and both prov

More ›