A Latino pool boy? Really? That's just lazy writing.

Named Giancarlo no less.

Not that hard to get a default judgment in an unopposed arbitration. Assuming the arbitration demand and service are properly done and there is a semi-viable claim, I don't see why the filer would not prevail. The arbitrator's judgment can then be confirmed by a court with VERY few exceptions.

It doesn't work like that that. Uber can do new contracts with no arbitration provision (setting them up for workplace class actions), but that would not render the old contracts invalid or eliminate their arbitration requirement for claims arising under the old contract.

For any claim filed in arbitration, Uber can move that it should be before a court (e.g., because the arbitration clause is invalid or the claim does not arise under it). But if they drafted the clause they would lose the vast majority of such motions and the ones they did not lose would not invalidate the clause. (Unless the basis on which they 'won' is affirmed by a state or regional appeals court. But that's a big stretch.)

The theory behind it isn’t that it makes prosecutor’s lives easier, it’s that it shows the criminal has accepted responsibility for his/her crime.

The real theory is STFU and plead guilty colored person because holy **** this system would grind to a halt if you people actually asserted your rights.

Serial season 3 is all about this phenomenon. Highly recommended

Yeah I've often wondered what would happen if every person facing charges in the US just decided to plea not guilty. No way the courts could accommodate them. I guess that's why prosecutors over charge to force the prisoner's dilemma.

Lol.

https://en.m.wikipedia.org/wiki/Cognitive_dissonance

So AlphaZero is a beast and I look forward to it ruling over all humans.

http://science.sciencemag.org/content/362/6419/1140

The only way it will find enough data to "learn" about humans is from social media. We are ****ed.

Meh

It will learn from history books and art/literature that humans sometimes have very dark inner thoughts that generally remained private until social media was invented, and understand that's not the "real" social us.

Previous AIs turned into misanthropic racists. One of these will kill us all.

Yea that's setting all the alarm bells off for some freaky sh*t goin' on.

Wonder no more. This strategy has spectacularly worked multiple times over the last ~120 years.

We in the IWW used a "jam the jails" strategy during our free speech fights of the 1910s. During the civil rights fights of the 1960s several groups, notably including the Freedom Riders, again "jammed the jails". Over 600 arrestees at the 1999 Seattle WTO protests used "jail solidarity". They each asserted their rights to a speedy and individual trial by jury. The the relevant courts had something like ~30 of these kinda trials over the course of the previous year... now they were looking at ~600 in a month. Instead, all these ~600 defendants collectively bargained to have all charges dropped. Also... there have been contemporary calls to action for those facing the immigration courts, and are able to do so, to insist on every one of their due process hearings. If it reached critical mass, this would completely bring that system to a stand still.

Yes it did. Although I'm a developer I don't know much about AI, but presumably this was just primitive/badly written AI?

If AI is going to kill us all, which is quite possible, I'd guess it's more likely to come about as a result of seeing us as expendable and unproductive, unintelligent biological units ripe for harvesting for some other purpose.

AI learns from data.
The data reflects the fact of racism
Hence the AI suffers from systemic racism even though it's not racist

IIRC this was a chatbot that got trolled by 4chan and now just says racial slurs. I think you can still find it online.

I was under the impression it just went through twitter and became racist organically by trying to mimic what it found. I don't remember a trolling angle, that would be somewhat comforting.

In 20 years people trolling at 4chan are going to use some off the shelf AI software to make 3D printers print biological weapons. Can the nanobots save us?

So, bad AI that for whatever reason assumed a greater prevalence of racism than is actually the case.

Sure, if you only look at twitter, 4chan and stormfront you'll get a highly racist, destructive bot, but several thousand years of human literature puts 20 years of online "literature" in some perspective.

In 20 years hopefully (though also sadly) online anonymity will be a thing of the past and 4chan scum will be arrested if they do that.

Not sure what you mean but a well written AI could learn an amount of racism that is exactly the case.

It's going to be a far more advanced AI that can learn the concept of racism and adjust for it in the data. By then our concern might be it being organicist

That would be just awesome if the police have the authority power and discretion to decide who to arrest based on what they post online.

If the AI pessimists are wrong, we delay some incremental progress out of an overabundance of caution.

If the AI optimists are wrong, it's the end of humanity.