I read a blurb on dreyfus' AI book today and was astonished at his insight into limitations of artificial intelligence research and how early it was.
https://en.m.wikipedia.org/wiki/Hubert_Dreyfus's_views_on_artificial_intelligence

What do you think?

 

No idea if he was right but AI can go Skynet in an instant.

https://www.theregister.co.uk/2017/04/25/webroot_windows_wipeout/

Webroot's security tools went berserk today, mislabeling key Microsoft Windows system files as malicious and temporarily removing them – knackering countless PCs in the process.

Not only were people's individual copies of the antivirus suite going haywire, but also business editions and installations run by managed service providers (MSPs), meaning companies and organizations relying on the software were hit by the cockup.

https://arstechnica.com/security/20...down-as-update-nukes-hundreds-of-legit-files/

 

Sounds like a very good reason not to use Webroot tools.
Can you imagine the chaos if it had deleted a file critical to Window's booting up?

 

I read a blurb on dreyfus' AI book today and was astonished at his insight into limitations of artificial intelligence research and how early it was.
https://en.m.wikipedia.org/wiki/Hubert_Dreyfus's_views_on_artificial_intelligence

What do you think?

Hi,

Interesting reads. Lessons to learn.

My favorite phrase was, "unbridled optimism", when it came to the belief in the quick success of AI. That happens now and then in various fields when something seems so darn good that our hopes begin to overshadow our subjectiveness and we get carried off into fantasy land without realizing it. It's like a gambling problem: the payoff is so good we start to believe we will win no matter what.

The lesson to learn is to not present our negative viewpoint too strongly as we just end up provoking the proponents into resentment and thus start to appear as an enemy rather than a colleague. A gentle dissertation with focus on accuracy wins the day.

I had an experience just the other day where i had to question AI mixed with robotics. I dropped a bottle of water on the floor in a dark room and went to retrieve it whilst lazily laying on the bed and reaching out. I realized that the bottle was just beyond perfect reach but my fingers could touch it, I had to fiddel with it to try to roll it but it did not roll so i had to slide it, all this in the dark. I wondered how an AI+robot could do this. The bottle was soft plastic too and bent in a little as it was touched, so that added to the complexity.