Posted on 07/11/2015 11:38:58 AM PDT by Lazamataz
A few weeks ago, Google researchers announced that they had peered inside the mind of an artificial intelligence program.
What they discovered was a demonic hellscape. Youve seen the pictures.
These are hallucinations produced by a cluster of simulated neurons trained to identify objects in a picture. The researchers wanted to better understand how the neural network operates, so they asked it to use its imagination. To daydream a little.
At first, they gave the computer abstract images to interpret like a field of clouds. It was a Rorschach test. The artificial neurons saw what they wanted to see, which in this case were mutant animals dredged from the depths of damnation.
(Excerpt) Read more at washingtonpost.com ...
I agree with your take on that.
Still not enough; a plan has to come into view or else it just trips all over itself and goes nowhere slowly.
I thought we were going by Abstergo now.
I have a thesis that 99.99999999999999% of the AI-self-rewriters will crash and burn.
It's that 0.00000000000001% that keeps me up at night.
Ack! Her Heinous and the Webbspawn in one horrifying graphic belched from the nether regions.
Wow, google’s AI saw Web Hubbel?
Okay. Then you don’t think the Left would stoop to this.
I don’t join you in that assessment.
So... I accidentally allowed a polymorphic smart code to escape quarantine.
Nothing bad will happen.
Just.. stuff.
Honest.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
No big deal if one of these algorithms is placed in charge of redistributing wealth, right?
Touche sir, Touche.
Don’t know why you disagree; I speak from a technical viewpoint.
By not feeding it any data about Democrats of COURSE the result is focused on whatever the data is focused on. Just like the results of a relational table join are tied to the parents. Those who ran this knew this. Of COURSE it is biased.
Actually, it was “TRANSPORTATION”.
Don’t know why.
Did you see the article on /. a day or three ago about the "programs" that can fix buggy code in a binary executable without the need for symbols or source code? IBM has been seeking that for years. I believe it was MIT that wrote the "re-writer". I think eventually what will be called "AI" won't be made from scratch. It will be imported into 1's and 0's from a mad scientist figuring out how to "map" a human brain. Then there will be "will" and we might be able to stop it if we ask it what the definition of is is.
When you post something to me, I generally reason that it is to inform.
If we both agreed that humans had manipulated the process, why tell me humans manipulated the process, unless you wanted to explain something away?
That’s the way I interpreted your post to me.
It seemed as if you were trying to say it was all a big misunderstanding, because humans didn’t ask the computer to do Democrats.
Although I didn’t think I had to say that in those exact terms, what else did you think I was thinking?
So, Dark, you’ve been giving the Google programmers coffee, have you?
Are you sure they are not just pseudo-artificial intellectuals?
You sound full of yourself, like yourself is the judge of what is going on.
Try expanding your horizons a bit, ok?
So on top of silly restating the obvious posts, you add this to your portfolio.
Doubling down are you?
Sadly, yes...
Possibly.
bfl
We are so screwed.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.