https://artistdetective.wordpress.com/2017/11/17/the-most-sensational-news-ever

With the recentmost news about Facebook’s AI going rogue and a robot being granted citizenship, I figured it was time to hand out some insight to the general public, and especially people tempted to read these stories, hence the clickbaity title

I had intended to start the article off with debunking Google’s 2012 expertiment “Computer teaches itself to recognise cats”, but after mulling over their paper I couldn’t conclusively tell how much of the reported 70% accuracy contained guesswork (potentially 64% of it, or none at all). In retrospect it is remarkable how much time I myself have wasted searching for the facts behind these stories over the years.

Thank you for writing a quality news article based on research.

You deserve to earn money for your efforts, on your blog.

There is a saying in the news business, “Ratings Raise Revenues”

Well, in a sense the ads are paying the fees for hosting my writings. This article also nicely ties into my current project, a no-nonsense summarizer, which may generate some revenue eventually.

Oh Don, I’m disappointed in you! Where’s the obligatory illustration of the Terminator to accompany your article?

Good point! The thing is, I tried to draw one on my computer, but then it became self-aware so I had to pull the plug. NAO robots are much safer as illustrations.

I can hear it now…

{computer to Don} Master? Do you think I’m pretty?
{Don to computer} Um… [reaches frantically for cord… YANK!] ... No.

I am with you on that! A couple favorite examples:

- Musk/Hawking/Gates fear of “killer robots” - as if AI will be “emergent” and the machines will automatically evolve into beings that want to destroy humanity. Weird how this “evolution” could skip all the intermediate stages between mud and Charles Manson. [Also weird cuz B2 bombers exist and so do trigger-happy people]
- The idea of driver-less cars that “almost” work.
- Siri cannot even answer a simple follow up question about the world series! You would think that would be a minimal requirement for AI.

I find Mr. Musk’s predictions re: AI a bit odd, since at least one of his companies (Tesla Motors, for one, and probably others) is employing a (rather limited, admittedly) form of AI in their products (autopilot, to be specific). I’m wondering how much is sincere concern, and how much is self-serving hype. :-\

My theory on Musk is that, because he is funding various AI projects, his scientists tell him all these amazing stories about how “close” they are to AGI so that he keeps funding them.

I am also disappointed in the low grade language processing of Siri and its like. I watched a video of Alexa that Steve posted, where it fails to even treat the word “it” as a reference, let alone resolve it. And it’s not like the technical solutions aren’t available or that the ambiguity in question-answering is crippling.
https://youtu.be/a_LQLBrR5wg?t=93

Musk is driving quite fast in his latest A.I. equipped Tesla and it advises him of his speed. Elon gives the A.I. a rather flippant, off-handed remark to the effect of, “I’m the one who’s driving right now!”

Just as he suddenly feels his seat belt harness cinch itself tighter around him, the steering wheel is spun out of his hands.
His A.I. equipped Tesla guns the motor and to Musk’s horror, crashes through a guardrail high atop a treacherous mountain pass and begins a death-dealing, downward plummet to the rocks below!

His tightly closed eyes are suddenly opened by the sound of a car horn!

“Mr. Musk, time to get up, your Tesla is waiting for you in the driveway to take you for your 9:00 meeting with your A.I. research group! Please don’t be late, Mr. Musk….”