Artificial Intelligence

AI Ain’t.

Today, AI is mostly a statistical model build using a set of training data and it can go spectacularly wrong if the training data set is flawed.

• Facebook Apologizes After Its AI Labels Black Men As ‘Primates’

• IS THIS A WOLF? UNDERSTANDING BIAS IN MACHINE LEARNING

AI doesn’t actually understand what the books, videos, etc that it recommends contains, it’s just recommending based on what other users you’ve been clustered with liked.

Then you have things that go wrong with semi-autonomous and autonomous algorithms:

• Flash crashes caused by High Frequency Trading (HFT) algorithms

• CIWS had a penchant for following the target tow cable and engaging the tow aircraft

• Microsoft’s Racist Chatbot Revealed the Dangers of Online Conversation

• Adversarial attacks: How to trick computer vision

We don’t need actual sentient AIs to create a nice mess and bizarre behavior, just some poorly trained black box algorithms build on faulty training data. Other things, like chat bots, can be intentionally exploited and/or manipulated (aka adversarial attacks).

• Avogadro Corp: The Singularity Is Closer Than It Appears

Bot nets and troll farms can be rented to push engagement, appearance of wide interest, fake clout, and so on and so forth. This is already rampant on social media and creates a huge cloud of uncertainty. Can also be used to manipulate content creators who are suffering from audience capture, create artificial influencers.

Troll farms are at least semi-automated and properly executed can get people to turn out in real life and do silly things.

AI created, and bot net laundered, fake news (local and otherwise).

Cyberpunk 2077’s SECRET Character - Mr Blue Eyes & His Dark Story Explained!

• https://youtu.be/QQf-9aIgTBI