Tripping up AIs and other security systems to make them fuck with your enemies will be the new norm by the middle of next year or so.
Quote Tweet
fascintimidating. krebsonsecurity.com/2017/08/twitte
1
1
2
Conversation
It's already common as hell, but as Facebook and other social media platforms gradually implement more AI systems it will become ubiquitous.
1
2
Especially as the geniuses thought it was a good idea to *simultaneously* roll out automation of customer service responses.
WCGW?
1
1
Everyone talks about Facebook and Google as monolithic entities of tremendous power. While they are, they are also incredibly fragile.
1
1
The implementation of AI is pushing tech outside the realm of best practices, etc. etc.
Intelligence services are quick on the uptake.
1
1
While this is going on in the background, the AI frontier means disruption techniques available will soon multiply.
Quote Tweet
I’m going to be really pissed off if the smart contract revolution gets interrupted by a North Korean nuclear EMP attack on the American net
1
1
We'll know how to mitigate some attacks by year end, with several band aid fixes before then. But the means of attack will keep multiplying.
Replying to
I don't think people realize yet just how badly we've fucked up in letting something that teaches itself new things loose on the internet.
1
It's not that we'll get Doomsday; but it will constantly develop new "social" engineering vulnerabilities. Impossible to keep up.
1
Regular code has bugs, 0 days exploits known vulnerabilities backdoors ladi-dadi-fucking-doo.
This code can change its mind about stuff.
1
The amount of ways in which this can be exploited without even knowing anything about hacking... let alone if you understand the code.
Replying to
The best outcome would be one of Twitter, Facebook, or Google being effectively destroyed by these bots so an upstart can solve it right.