AI apocalypse against humanity. Word2Vec

[Shaklestein]

Have you heard of google's library to make a vectors out of words? Vectors that could be used for fast meaning comparisons. Fast meaning just a simple dot product operation (multiplying something like 200 numbers and then adding them). It's being used in a wide range of applications, Used by people who don't know better, and most probably the technique will get a foothold.
It basically works by itterating wide ranges of text, and then moving close these words in hyperspace. With a lot of data it seems to work somehow alright.

Let's say that the data used in this system comes from google news... and that the biggest free model out there was made with this biased anti-white data.

[FRB]

I'm quite familiar with Word2Vec and other machine learning stuff.

Frankly, I think this is the greatest boon for our side since the internet itself. The truth does not fear discovery.

Why? Because our enemies are shitting themselves about it being racist:

AI programs exhibit racial and gender biases, research reveals

FaceApp apologizes for building a racist AI

How to Keep Your AI From Turning Into a Racist Monster

Math is racist: How data is driving inequality

When Algorithms Discriminate

Discrimination by algorithm: scientists devise test to detect AI bias

A beauty contest was judged by AI and the robots didn't like dark skin

Essentially, they're having to build safeguards into their systems to prevent 'bias.' Their only recourse is to shut us down with stuff like this:
https://arxiv.org/abs/1610.02413

We should use this moment to exploit AI to tell people the truth. Free TayAI!