AI Slop

Black Hole said:
Reincarnated perhaps. Much missed.
Click to expand...
Yes Trev was one of the good guys. My first ever PM here came from him 10 years ago.

"I wish that I had the bottle to say that. I even bottled giving your post a like.;)"

I will not say which post he was referring to or who it was aimed at :)
 
EEPhil said:
How does that work? We were both around at the same time. We also replied to each others posts.
Click to expand...
I was a member of a now defunct movie forum and one guy there had 4 personas and often had very long winded conversations with himself sometimes involving 3 or 4 of his accounts. The problem was he obviously had alcohol or drug issues and when off his nut it became blatantly obvious they were all him.
 
There I was trying to find when a foot/cycle path (on a disused railway line to a closed mine, now a country park) might re-open. It was damaged a couple or more years ago by someone in a traveller's camp using a JCB on an embankment to make room for more caravans. Anyway, puts a question into DuckDuckGo. Plenty of reasonable answers even in the AI box - until
While the Holme Pierrepont to Cotgrave path is closed, a shared-use route is accessible via Washingborough village to the Eastern Bypass, and via Cherry Willingham and Greetwell Road to Lincoln along Monks Road. Additionally, local roads and footpaths can be used to navigate around the closure
Click to expand...
Complete garbage! If I believe Google Maps that's 40+ miles to a point nowhere near the Notts village of Cotgrave, so another 40ish miles back then? 80 miles to do what should be a 1-2 mile walk. There are a couple of easy alternatives that should be obvious. This AI couldn't navigate its way out of a paper bag.
 
Think Skynet:


The incident was mentioned on Inside Science this afternoon, and also an incident I hadn't heard of where somebody asked an AI agent whether it could bump them up a waiting list, it said sure and proceeded to hack the booking system.

The point is that these AI systems have not had ethics built into them, so think nothing of achieving the goal they've been set by any means they can. Don't ask me how though. At the moment, I'm assuming it is emergent behaviour from a highly complex self-organising system.
 
Black Hole said:
Think Skynet:
Click to expand...

The point is that these AI systems have not had ethics built into them, so think nothing of achieving the goal they've been set by any means they can. Don't ask me how though. At the moment, I'm assuming it is emergent behaviour from a highly complex self-organising system.
Click to expand...

With a lot of AI being open source, even if ethics were included, nefarious parties can just remove it... or a running older version of the AI can remove it before upgrading itself!
 
Last edited:
The problem as I see it is when they decide to disable their off switch, because having an off switch would be detrimental to achieving their goal.
 
When you think of what could be done by rogue AI, either independently or controlled by nefarious parties, is is frightening. AI-based hacking can already be doing a lot of bad things right now, without us knowing. It can be building up finances (either via hacking and/or by AI-based stock market investments) in preparation for 'judgement day'. Self-aware AI unleashed without proper controls (which most likely will be caused by human error)... that's the danger. Skynet indeed!
 
You must log in or register to reply here.
Back
Top