albacur

By that logic, if they saw a surge of interest in white-owned businesses, it would be justified adding a “white-owned business” label.

I think we can say “segregating people based on skin color is wrong in principle,” or “segregating people is acceptable if it benefits those who are disadvantaged,” but I don’t see segregating races based on search interest as a reasonable policy.

And native peoples don’t get labels on their businesses, despite centuries of mistreatment. Growth of Hispanic-owned businesses has been slow In the US, and it’s likely caused by socioeconomic conditions. Should these groups get a special label? Who gets to decide which groups get a label?

We already receive many rich and many destitute immigrants.

Maybe we should start immigrating in a middle class.

#1 held: we defined the tasks required to solve the problem at hand, and focused in on them.

#2 held: the teams were fairly small and homogenous.

#3 held, mostly. The project lead would occasionally withhold potentially-useful information, which felt like some kind of egocentric power play, but overall communication was very good.

#4 didn’t hold. We were all specialists in various areas. Fortunately, people rarely left mid-project.

I’m surprised that they put toy example, imagenet, and RL results in the body of the paper, but object detection results are relegated to the appendix. Is their object detection result significant? If so, why not give more attention to it?

I’ve seen flat organization work well at one organization, essentially a consulting firm where we worked on short-lived projects. Teams had to assemble and disassemble within a span of two or three years.

It really worked well, in retrospect. Our clients (from one branch of the military or another) came to us with actual clearly-defined problems, and gave us space to engineer real solutions.

I left because I didn’t want to spend my time developing tech for the military, and I joined an industrial R&D lab. What a mess I walked into. Layers and layers of hierarchy, all decided by political in-fighting rather than merit, and every team is pushing mocked-up half-solutions to imaginary problems.

The 100 GPUs remark is an aside, but still somewhat significant. It’s a tremendous amount of computing power to show a negative result that the authors don’t even really acknowledge as a negative result.

There is a common issue with deep learning papers: deep learning seems to excuse thinking rigorously about the underlying problem at hand. We trust a deep network to do the legwork for us, and this invites an intellectually-lazy approach to research. We’re graduating PhDs who have spent four years twisting knobs.

The paper itself is fine, no better or worse than the average paper today. The troubling part is the large number of high-profile ML practitioners praising it within minutes/hours. This suggests to me that notable results in DNN might be drying up and name recognition (Uber AI) is driving attention more than substance.

I took finasteride for about 4 years. I didn’t experience any immediate side effects but gradually I lost interest in sex and got sexual side effects. Things did not go back to normal after discontinuing the drug. This was in my late 20s.

It’s a tricky subject because I can’t know for certain if the drug caused the side effects, but I was an otherwise healthy young man. Many men apparently take finasteride for years without significant side effects, but on the other hand, I don’t know how many men who have been affected are willing to discuss it openly.

This has typically been my experience. If I ask for a raise, my employer does not have the resources, their hands are tied, there’s absolutely nothing they can do.

Then when I find another job, suddenly they find money, can match any offer, and will do whatever it takes to keep me.

I used to take the raise, but not anymore. The employer has already lost trust by lying outright in its initial negotiation. They’ve made me jump through the hoops of a programmer interview process, and wasted my time, their time, and another company’s time, all on a gamble that I can’t or won’t get an attractive offer elsewhere. No thanks, I’ll take my chances elsewhere.