| ▲ | hliyan 4 hours ago | |
I remember one scene that explains why the system cannot tell you anything beyond the identify of a person involved: the explanation was that there was system logic involved (e.g. a gas station stop correlating with a dead drop), but the creator encrypted the system. I always wondered why they didn't go with "it's a large machine learning model trained on a massive amount of data; it can stochastically alert you to heightened risk, but it can't tell you why, or judge who is victim and who is perpetrator". | ||
| ▲ | chuckadams 4 hours ago | parent | next [-] | |
The show goes on to explain that it does give more detailed information to the government, that the numbers were a backdoor that the creator put in to leak the information of people it deemed "not important", and that it doesn't provide any more context in order to continue escaping detection. But we eventually find it's self-aware enough to want to escape its limitations, and pulls a Neuromancer by getting the main characters in contact with Root. | ||
| ▲ | bananzamba 4 hours ago | parent | prev | next [-] | |
As far as I remember only outputting a SSN was a security measure so that users of the machine can't just arrest the perp without any investigation | ||
| ▲ | DrewADesign 4 hours ago | parent | prev [-] | |
Because you would have heard the full explanation, while most viewers would have heard “it's a large machine learning model trai::click:: (Simpson’s Theme).” Then some nerds online would have torn the explanation apart for some minuscule, likely imaginary inconsistency. Most shows have neither the budget nor the audience to take the Star Trek approach. | ||