You're wrong here. Testing has shown that people who don't understand what the technology does can abuse it to prove any failure rate and reason that they want. For example, give it a database of criminal mugshots, set the match criterion very loose, and then be aghast that comparing images of a legislative body results in a large number of hits. Or start with a database of ethnically homogeneous images, set the match criteria tight, and then be aghast that comparing a set of images of different ethnicity doesn't result in any matches.

If you're not going to speak English, and don't understand what you are saying in whatever language you are speaking, please refrain from suggesting technical ideas.

No, they actually aren't, unless you have something doing DNS via HTTPS, and then you've got HTTPS encrypting it all already. 'DNS' is not "the web". "The internet" and "the web" are not synonyms.

Don't be so ignorant.

That's what HTTPS already does.

This is about DNS, not "web traffic".

There is never a never. Vendor lock-in means it is very expensive to change, not that it is impossible. If you buy a Motorola radio system, for example, you are locked in to Motorola -- until you spend the money to replace it.

If you put all your data into the AWS bucket, you are locked into AWS -- until you spend the money to get it all back out of AWS.

I was asked to investigate the costs of moving our data into AWS. The storage was really cheap. The open-ended costs were accessing the data outside of AWS, and that killed the project.

It's actually stupider than that. The alleged positive version was "greg says his car is blue." Note that this changes the verb (said/says) from past tense to future (currently says and continues to say), AND the car from "was blue" to "is blue".

Name two.

Then these would be contractors.

Are they? Do you have hard numbers to back that up?

Yes, depending on the factors I cited.

It would be, but first you need a quantitative level of "evidence" before you can say such a statement is true or false. "I am 25% NA" is quantitative and can be determined to be true or false based on other evidence. "I have NA heritage" is a qualitative statement. Is 15% enough? 1%? Being abducted at birth and raised on a reservation?

As a fact, you are wrong. If burning a flag is 'free speech', then using money to buy flags to burn is also free speech.

Again, incorrect fact. Were you correct, please explain the increase in advertising prior to an election. That's a clear demonstration of "scaling speech", and if it didn't work it wouldn't happen.

Liberals always question the legitimacy of elections they don't win. "Solely publicly funded" creates a huge taxpayer liability, or Constitutionally unacceptable limits upon some of the most important free speech. Who picks which candidates get "public funding" and which don't? How much money do they get? How can limiting the amount of speech someone can pay for not fail the first amendment "congress shall make no law" test?

There is no such thing. There's no money in doing it. The remuneration comes from furthering certain agendas, or pandering to an audience and getting more ad revenue. Kinda like network news these days -- it's goal is increasing viewership ("eyeballs"), not educating anyone on anything. They can't, because there isn't enough time to cover any story that way.

In /. we call them clickbait.

So you want not just "fact checking", but "interpretation".

Here's a test: "I have Native American heritage." Fact or fiction? It all depends on what percentage of DNA you set the level for "heritage" at, and if you exclude "lived in the culture" as having a heritage.

Now, I'm fascinated how this story is all about Facebook, when we now see a major presidential candidate saying, out loud and in public, that if Facebook won't prohibit her from doing something unethical in her Facebook ads that she's going buy ads with unethical content. "If Facebook won't control my behavior, then I'm not going to control it myself..."

Proof by assertion. Fact not in evidence. They don't have a demonstrated history to base this claim on.

So your argument is that "any response" is better than "the correct response"? If all you are considering in "better" is how fast the response happens, then that's what you are saying.

A wrong answer provided in 10ms is still the wrong answer.

So an explicit statement that the wrong answer fast is better than the right answer slower.

Something appears in front of you suddenly. Sometimes the right answer is not to immediately (10ms) slam on the brakes. Being wrong faster isn't how I want my "self-driving car" to operate.

Then thank goodness for companies like Facebook that can protect the ordinary citizen from all that stress by banning speech that might confuse them. Will nobody think of the ordinary citizen?

Early years is not long term. "Early years" with a lot of corner cases that cannot be handled means less acceptance, because the limits limit the usefulness.

Nobody is talking about "early years" when they talk about self-driving vehicles becoming standard.

Exactly analogous. Air travel is safe because of the regulation, long experience with corner cases, and often, regulation that prohibits operation in the corner cases. Commercial air travel, while well accepted for long distance travel, is not well accepted for daily commuting from home to office.

Everyone who wants to hand-wave away the problems always says that. Not only is it wrong, it's short sighted.

It is those "corner cases" that will drive consumer and regulatory acceptance. Refer to the Boeing aircraft that has had just two, two crashes due to a certain corner case issue and yet the entire fleet is still grounded.

Also, there is a term in telecom: "the final mile". Getting data all the way across country is trivial compared to being able to get it the final mile to the home. "Self driving" implies "from door to door", especially in something like an Uber where the passengers go "from door to door" and not "from a bus stop somewhere to the side of the main highway somewhere close to where they want to go". Solving the final mile problem will have to happen, and that's the hardest part.

Of course, the "if ever" part may also depend on consumer demand. Not just consumers who want "door to door", but consumers who don't want self-driving cars. If the system winds up depending on all traffic being intelligent self-driving vehicles, then mixing in the human drivers will mean "if ever" doesn't happen. If voice control systems like Alexa depended on getting rid of manual control, I dare say there are enough people just in /. who would never allow an Alexa in their presence to prevent voice control from being a thing.

No, because it is COMMON SENSE.

Companies are developing and delivering "AI" systems that need to identify "threat" vs. "non threat" in video, or respond to commands in normal speech from humans. This AI needs training to work well. After initial training, the results need to be evaluated to make sure the system is doing what it is supposed to do, and to identify areas where improvement is necessary.

Why is improvement necessary? BECAUSE PEOPLE ARE PAYING FOR THIS SERVICE AND EXPECT IT TO WORK RIGHT.

The only way to do this is to monitor the inputs and outputs, and that takes a human or two.

IF YOU DON'T HAVE ONE OF THESE, that's fine. Nobody really cares how smart you think you are, and it doesn't change the fact that the people who DO want this technology expect it to work. They want it to do more better, and they really don't care what you think about their decision. They really don't.

Can we please stop this? We know every company that provides AI voice or video recognition has people monitoring the system for validation and improvement. Every damn one of them. And they aren't big evil invaders of your privacy because they do -- you don't have one so they aren't listening or looking at you.

No, you do NOT remember when everyone said they weren't going to let such things into their home, because everyone didn't say that. Everyone YOU KNOW might have said that, but your limited set of contacts doesn't mean nobody wants this. You can't even claim that "everyone on /. said they' weren't going to use these", because not everyone on /. has said that. The fact that a large number of people post about how smart they are not to have these and how stupid everyone else is doesn't make an "everyone", either.