blurps

> into the trash bin?

They did not implement the winning solution as is, but a lot of good stuff came from the winning teams, including techniques (such as SVD) that are in use as of today (or maybe a few years back).

> just a nonsense engine

No, it is safe to assume the recommendation engine of Netflix is close to the state-of-the-art. A lot of money and talent went into it.

This only works when you don't win. You have to upload your code, including model training code, they won't accept a trained model binary.

Winners have to submit code most of the times. There are sporadic computer vision competitions where it is possible to hand-label (such as geo-int competitions where people reverse engineered the flight paths), but vast majority it is impossible to hand-label. Nobody becomes a grandmaster without applying world-class machine learning skills.

The "cheater" in this competition is both a world-class data scientist and a reverse engineering hacker. Heck, people used to write papers about how they crawled the ground truth. Now they see their name in the newspaper.

The Netflix Prize winners regularly compete on Kaggle. One of the top 10 Netflix Prize winners joined Kaggle early on (he got a job out of it). Kaggle competitions are a game. Most notorious players already have a good job.

You are using the same tactic I described: State surveillance has the potential to turn us all into Nazi's, so state surveillance is inherently bad. How can I ever hope to support the good of state surveillance for national security when I have to defend the Stasi? I can't. You (God)win.

It is intellectually dishonest to pose the effort of governments to combat illegal immigration as a targeting of people based on race. It is a sensitivity trap that shuts down our rational faculties. Yes... all religious extremists are also religious. Most immigrants have a different race. Yes, being an undocumented immigrant makes you vulnerable. But deserving of coddling, because they happen to also share a protected variable? No.

Great informative article, but at times the posing is naive or supporting a certain narrative.

For instance: The government goes after Islamic terrorists and Mexican illegal immigrants. So what to complain about? Mix cause and effect to pose the illegal immigrants as targeting based on race. Or pose the extremist religious as vulnerable Muslims, targeted on religion. Remove their status as far away from the cold-hard reality, so you can claim racism and prejudice and use that to cripple surveillance. But the suspect is a "citizen"! All weasel words to stay away from the real dirt. Let's go after the atheists for counter terrorism, so we don't discriminate in the eyes of the ACLU...

Then the "ineffective against terrorism"-claim. This can't be supported, because no clear numbers/cases are known. Then conflate laws against terrorism with actual actions against terrorism: Sure, no law directly contributed against countering terrorists, but the surveillance certainly did. Thanks NSA for passing along information to my country many times, so we could capture really dangerous people before they could strike.

Anti-terrorism is also a red herring for surveillance efficiency. Surveillance predominantly used for other national security purposes, such as counter intelligence, foreign intelligence, border safety, crime fighting, and even economic espionage.

Sure, posing any immigrant as a bad hombre, potential MS-13 gang member, selling your children fentanyl-laced drugs, is ridiculous. But the other side of that same coin is posing bad people as vulnerable undocumented citizens who are being targeted on race. It does not do justice to the situation on the ground.

Thanks. I'd never considered adaptiveness of a theory before and I like it. Perhaps adaptiveness at times emerges spontaneously from simplicity / regularization. But it is good to keep it in mind from the very start.

Just like one can't disprove solipsism, one can only brush it aside as pragmatically useless. Just because a theory is internally consistent or even accurate, does not mean it is a good all-around theory: simplicity and elegance counts too. And having to agree on who is the solipsist among all figments is inelegant and an impossibly complex task.

> According to the most extreme form of this view the only way by which one could be sure that a machine thinks is to be the machine and to feel oneself thinking. One could then describe these feelings to the world, but of course no one would be justified in taking any notice. Likewise according to this view the only way to know that a man thinks is to be that particular man. It is in fact the solipsist point of view. It may be the most logical view to hold but it makes communication of ideas difficult. A is liable to believe ‘A thinks but B does not’ whilst B believes ‘B thinks but A does not’. Instead of arguing continually over this point it is usual to have the polite convention that everyone thinks. -- Alan Turing