Automated data poisoning is something thats done right now by fake news and propaganda outlets all the time, but I think its something that will be embraced by the masses at some point to thwart the scenario you outline. Because nobody wants their life available online as an open book that any criminal, malicious state actor, or anyone else in general can pick up and read. And then impersonate and exploit and hoax connected parties like family members, like their banks and employers.
So there are going to be decoys, chaff, static, fake personas, rabbit holes leading bots off on wild and crazy chases that lead nowhere and consume endless bot resources, and so on. No electronic data will be an accurate representation. Think the virus and antivirus ecosystem with the constant arms-race to try and catch out each other, but in a data format.
Conceivably the zoom job-interview you just had with a potential new staff member that you will sign up for a work-from-home position might be an AI among thousands that some guy has coded up and is running on a cloud host, all collecting paychecks from jobs they are either pretending to do or sort of doing, and funneling the cash to him until they get 'fired' from some company that figures out its a hoax. Which lasts until another victim HR department is located.
Or that 20 year old german student who has signed up to a dating website with a video profile how she is interested in meeting guys in her city might be a 50 year old korean man who is trying to catfish people for $ and then extort them by threatening them with exposure as a cheater to their families.
Real life is going to be the safe and trusted medium if the info-space gets so contaminated its no longer viable. And what value is AI if its data-set building blocks are all hoaxes, fabrications, or arguments among 6 year olds?
Automated data poisoning is something thats done right now by fake news and propaganda outlets all the time, but I think its something that will be embraced by the masses at some point to thwart the scenario you outline. Because nobody wants their life available online as an open book that any criminal, malicious state actor, or anyone else in general can pick up and read. And then impersonate and exploit and hoax connected parties like family members, like their banks and employers.
So there are going to be decoys, chaff, static, fake personas, rabbit holes leading bots off on wild and crazy chases that lead nowhere and consume endless bot resources, and so on. No electronic data will be an accurate representation. Think the virus and antivirus ecosystem with the constant arms-race to try and catch out each other, but in a data format.
Conceivably the zoom job-interview you just had with a potential new staff member that you will sign up for a work-from-home position might be an AI among thousands that some guy has coded up and is running on a cloud host, all collecting paychecks from jobs they are either pretending to do or sort of doing, and funneling the cash to him until they get 'fired' from some company that figures out its a hoax. Which lasts until another victim HR department is located.
Or that 20 year old german student who has signed up to a dating website with a video profile how she is interested in meeting guys in her city might be a 50 year old korean man who is trying to catfish people for $ and then extort them by threatening them with exposure as a cheater to their families.
Real life is going to be the safe and trusted medium if the info-space gets so contaminated its no longer viable. And what value is AI if its data-set building blocks are all hoaxes, fabrications, or arguments among 6 year olds?