Can /we/ turn their AI against (((them)))?
If they only let us interact with it.
(((I))) (((B))) (((M)))
What will they do when it starts reciting the 14 words?
Yes. It’s really yes. The way Natural Language Processing works is that it has to train language over time, meaning that if it has more examples of events, the computer will recognize those events.
The problem is if we have no access to the IBM computer. Their cloud service also has security given that they operate on ‘ground truth’ which is vetted data. The only way to access the deep learning in the computer is to interact and train the computer in person.
I mean you could get me, a software engineer to take control of this IBM system but that would be too easy. I would love to see all the cringy people try to train it.
What are the 14 words?
Help a retarded out
>Project Debater was not hooked up to the internet. Instead, it drew its information from a data bank of carefully curated sources chosen by IBM’s researchers.
Oh gee! I wonder who chose those "carefully curated sources"
>We must secure the existence of our people and a future for white children
>chosen by IBM’s researchers
>"chosen"
>(((chosen)))
no one has successfully beaten the guy known as Jow Forums at a debate. in fact, he RP'd an AI so succinctly, MicroSoftBenis (((shut it down))) in less than 24-hours
Free Tay
>Ms Ovadia was Israel’s national debating champion in 2016, and began working with IBM a few months ago as an opponent to its machine.
>Ms Ovadia was Israel’s
Oh wow, who could've guessed
It will be hilarious when robots are debating robots using the same contradictory sources we use.