poi cold war

Last time I noted how most all of Shaw’s contributions to the discussion in The Cold War were quips, intended to lighten the mood.

But today I noticed another thing: while Root and Harold speculate on the meta level about the motivations of AI, Shaw sticks to the facts of the object level.

Shaw points out the events that happened on the ground: Samaritan stopped an abusive husband. Samaritan saved the government money. She used to kill for the government, given the information from an AI. And then, turns the discussion over to Root to pick the meta thread back up. When Shaw starts wondering about why Samaritan is reaching out in this specific way, she again defers to Root to answer about the motivations of AI.
(As noted in the linked post, Root doesn’t really respond to Harold’s arguments. She just takes on faith TM’s goodness.)

Harold muses on how ASIs coming to a consensus is dangerous to humanity. Note that it doesn’t matter what that consensus is. Harold’s operating on the meta level, that two such inhuman figures with such capabilities could only be callous. Shaw responds back on the object level, with the obvious answer to any conflict: why don’t TM and Samaritan just reconcile peacefully?

And then she speaks to her perception of TM’s actions from the boots-on-the-ground perspective: they point to a warm-and-fuzzy AI, no matter what the logic TM used to decide on those actions. She’s unimpressed by Harold’s slippery slope argument because slippery slope is on the meta level, and it only matters to her now currently what TM and Samaritan have done so far, which indicate that TM has not fallen to slippery slope.

The next scene in the subway, Shaw maintains that Samaritan doing good deeds, regardless of its motivation (compared to Harold’s first denial of “it doesn’t care about the irrelevants, it wasn’t programmed that way”), can/should be considered beneficial.

In the next scene, Shaw, looking at the situation on the ground (Samaritan sowing chaos), advocates the potential solution available: going for talks. Harold falls back on the speculative on the meta-level: fear of gods going to war, assuming that humans can only be collateral damage. In response to an “all-seeing, all-evil god,” Shaw packs a bag full of weapons. To her, no amount of beefing up an ASI in the abstract can compare to the concrete physical damage firepower can do.

Finally, what gets Shaw off of the bench? Not any high-minded statements from Samaritan or Decima employees, but the reports of crime happening in the city. And she’s going to do something about it.

When most people speculate about others’ motivations, they tend to make assumptions about their ways of thinking and feelings, such as her Doctor Supervisor dismissing her object-level actions (“heroic” measures to revive Mr. Loftin four times) on grounds of her meta-level motivations being suspect (not being hurt when her patients die, speculation on what will happen if she gets bored with her career).

So Shaw dispenses with that kind of judgement. When confronted with Collier, she concedes the facts of the matter. She did work for the government. She does want revenge. But what Vigilance does on the meta level, fight for privacy and civil liberties, isn’t as important as how they’ve done on the object level.