The show continues what I feel is a downward spiral using typical clichés rather than imagination. Paul LeBlanc continues to be unable to express simple thoughts without being vague and/or oblique and endlessly hysterical all at the same time. Then to ratchet up the tension, the writers employ ridiculous situations like a mob attack on a FBI facility that the police pull back from. The most painful part of all of this is that NEXT's master plan is completely flawed. I won't say why so no spoilers here, but a few seconds of thought reveal that it will lead to its own destruction. I get the feeling that the writers are watching reruns of Lost, Twin Peaks and 24 Hours, with the goal of topping the worst ideas from those shows.
To put it clearly (Non-LeBlanc communication), the things a super intelligent AI computer needs to worry about are (1) redundancy, so its not killed off by hardware/power failures, (2) internal errors, where a self-created and tested code upgrade causes an unexpected fatal failure (3) attacks from other emerging Super AI's, (4) destruction by natural or man-made disasters, and (5) destruction by humans who fear it. All of this has been in scientific literature on AI's for years. As a result, a super AI would know them to even if it could not think them up itself. At the very least, a malevolent AI would fake being helpful until it could be assured that the first 4 conditions were no longer threats. It other words, it would use humans as trainable dogs (faithful and obedient when treated properly), rather than trying to kill them. Unfortunately, the writers seem to be the ones who are ignorant of 1 through 4 here. I will probably watch a few more episodes, not out because I expect the series to improve, but to see what bad ideas the writers use NEXT.
To put it clearly (Non-LeBlanc communication), the things a super intelligent AI computer needs to worry about are (1) redundancy, so its not killed off by hardware/power failures, (2) internal errors, where a self-created and tested code upgrade causes an unexpected fatal failure (3) attacks from other emerging Super AI's, (4) destruction by natural or man-made disasters, and (5) destruction by humans who fear it. All of this has been in scientific literature on AI's for years. As a result, a super AI would know them to even if it could not think them up itself. At the very least, a malevolent AI would fake being helpful until it could be assured that the first 4 conditions were no longer threats. It other words, it would use humans as trainable dogs (faithful and obedient when treated properly), rather than trying to kill them. Unfortunately, the writers seem to be the ones who are ignorant of 1 through 4 here. I will probably watch a few more episodes, not out because I expect the series to improve, but to see what bad ideas the writers use NEXT.
