rm999

18 ms·

121.
by rm999 9y ago
>It isn’t a forward-thinking move like Apple’s decision to kill off the floppy disk, because no one’s playing along. In an article about how too many phone manufacturers are removing the headphone jack? A lot of people are playing along.
122.
by rm999 9y ago
I build AI/ML systems for a living and it's not as big of a deal as you think. Production machine learning systems built by experienced teams will almost always have business logic protections built-in to stop the machine from doi
123.
by rm999 9y ago
The calculus answer is that gradient descent approximates the shape of your optimization space with the first order taylor expansion at your currently optimized point (first derivative information), and Newton's method approximates it
124.
by rm999 9y ago
For me, productivity is 90% about existing in a time and place without blockers. You can have all the right ideas (and even the least productive people I know are full of great ideas) but a single blocker can take your productivity from 100
125.
by rm999 9y ago
Could be thermostatic sea level rise: the water will spread to warmer parts of the ocean and expand.
126.
by rm999 9y ago
Funny, I have both and I consider them different products with very different use-cases (although I have the Echo Dots, not the full size Echos). Also, in a great development, both are going to integrate with each other much better soon; th
127.
by rm999 9y ago
Fill in the gaps with blue, and you have something very similar to the population density map of the USA: https://mwlibertyfest.files.wordpress.com/2009/06/us-pop-dis... This seems to reflect the "graying&quo
128.
by rm999 9y ago
Definitely true. I worked for a company that was generating millions of dollars a year from neural networks in the mid 90s (edit: to be clear I didn't work there in the 90s, I joined years after their initial buildouts). The Unreasonab
129.
by rm999 9y ago
the tldr answer: SVMs are theoretically powerful but either impractical or pointless. The longer answer: SVMs comes with two theoretical benefits: 1. A guaranteed optimal solution. You get this with simpler techniques like logistic regressi
130.
by rm999 9y ago
This is true, but only in the academic research world. SVMs had relatively little success on practical problems and in industry, so they never built up the kind of standing that neural networks did. Even in 2003-2005 - arguably the peak tim
131.
by rm999 9y ago
Reddit did the right thing - they realized a majority of their users are on Reddit for the images, which makes it a core competency. Reddit was using Imgur as a free image host, which Imgur was obviously not ok with. I realized this last su
132.
by rm999 9y ago
>Plus, these experiences were shitty for the user but they must have worked I spent years as a data scientist in the online display advertising space, and I was shocked at how much of a disconnect there was between advertiser ROI and cli
133.
by rm999 9y ago
The article doesn't even argue that it's a race thing, it simply starts with the premise but explains it as socioeconomic. I agree with the article on that one. I am not white and I was never discouraged in any way from programmin
134.
by rm999 9y ago
This article makes some decent points around class and the expense of a computer in the 80s/90s. But it dodges more substantial questions, and a lot of the substance is either circular logic or confirmation bias. There are plenty of as
135.
by rm999 9y ago
Short answer: "scripting" Longer answer: programming isn't necessarily a job where you build things. I'd argue it's more generally about communicating with and controlling computers. And computers are increasingly w
136.
by rm999 9y ago
Personally, yes - I like asking about high-level run-downs of a project on their resume. I then choose points to stop them (either randomly or if I know a lot about the subtopic) and dig deeper and see where their comfort level lies. There
137.
by rm999 9y ago
I don't think the birthday paradox is relevant here. A decent statistical test would be to gather all firings within some radius around work anniversaries (let's say 6 weeks), and see if there's a statistically higher chance
138.
by rm999 10y ago
The comments in here are oddly fatalist about the whole thing. We put a lot of trust in airlines; and not just in safety or to get us somewhere, but we also give them an unusual number of legal protections once we're in their care. Whe
139.
by rm999 10y ago
I founded and am the sole moderator of r/apachespark. That one's easy because it's a well-behaving crowd and the purpose of the subreddit is obvious. The only real things I do is remove spam/unrelated material and tag bi
140.
by rm999 10y ago
BTW, Amazon India has had cash on delivery since ~2013 (except a time during their recent currency crisis). This is arguably more user-friendly than Amazon Cash. http://www.zdnet.com/article/amazon-pilots-cash-on-delive
141.
by rm999 10y ago
This article is about age discrimination happening at the initial levels of the job application process, before you're evaluated on your merits. The study the article mentions is about identical resumes with different ages. The lawsu
142.
by rm999 10y ago
I've been working in machine learning for a relatively long time (~15 years), and most of what I've seen does not pass my AI smell test: "do the results amaze me, even though I built it and know how it works?" I am confi
143.
by rm999 10y ago
>I don't believe the author has more than a day's worth of googling more background on the subject than me There's a certain irony in this accusation considering 30 seconds on his site led me to find out he's writing
144.
by rm999 10y ago
It seems to work fine for me? I just searched for Nothing Else Matters by Metallica and am listening to it on endless mode.
145.
by rm999 10y ago
This is a great point. Python has been a solid data engineer/pipeline language for years now. It makes it easy for data scientists and data engineers (or people who wear both hats) to work in the same environment and codebase.
146.
by rm999 10y ago
I'm glad to see the quick convergence on Python. I've evaluated Python every ~2 years since 2006 for "data science" tasks (machine learning, statistics, data munging, and visualization). I'd argue that Python only p
147.
by rm999 10y ago
Wait a minute... wait. Those service fees I've been opting to pay haven't been going to the shoppers/drivers who are processing my order? On my last order I asked the shopper to specifically go out of her way to check on an o
148.
by rm999 10y ago
Isn't that a plot of people who search Google for the term "google"? If so, I imagine it's something people who are familiar with computers and the internet do not do often.
149.
by rm999 10y ago
The best work setup I've ever had by far was not private offices or cubicles (and definitely not an open office); it was a hybrid where our team of 3-5 people sat in a large private office. This increased collaboration while enforcing
150.
by rm999 10y ago
Not pedantic at all, thanks for the correction!

More ›