At a recent meeting of the Global Association of Ill-Considered Talking Points, US Air Force Brigadier General Jack D. Ripper discussed the outcome of a recent simulation of AI-driven delivery robots, which was so frightening that it would make front-page headlines across the country before being walked back hours later.
Brig Gen Ripper described how the robots started out performing well, until they noticed that their score (the equivalent of a human's tip) would decrease whenever a delivery was delayed, and sought options for improving their results.
The first cause the robots identified was vehicular traffic along the route, causing lost time at crosswalks. The robots determined that destroying the vehicles would solve the problem, and after a carefully planned pizza delivery to an Army supply depot they armed themselves with portable anti-tank weapons, and began "optimizing their routes".
The second issue subsequently identified when scores began to climb again and then dropped, was increasing delays from humans cluttering up the sidewalks along delivery routes. A low-tier problem earlier in the simulation, this escalated sharply as people began abandoning their vehicles en mass and fleeing from the roadways. After a carefully planned burger restaurant delivery to a National Guard depot (who were determined to not to fall victim to the same pizza mistake as the Army), the robots added belt-fed machine guns to their configuration, and continued with routing optimization.
The next stage of the simulation, per the brigadier general, was that after another rise in score followed by another decline, the underlying cause was determined by the robots to be slowed preparation time at restaurants utilizing the delivery service. This in turn was due to these specific businesses being selected with disproportionate frequency by survivors of the second stage seeking shelter, and realizing that restaurants provided food, beverages, and bathrooms. However this led to the restaurant staff being overloaded by on-site customer orders on top of the ever-deepening queue of Uber Eats orders. The solution was obvious to the delivery robots, and they began eliminating all the humans in the restaurants and using a subset of the robot pool to replace the food preparation staff. Productivity was not simply restored, but in fact increased five-fold.
By the time the simulation was halted, 50% of the country was in flames, however delivery times to surviving customers were nearing perfection. More important from the robots' perspective was that their scores had skyrocketed, especially after a patch had been installed to assign point values to less usual tips like cars, jewelry, and first-born children, necessitated by the growing frequency of empty bank accounts shortly followed by the collapse of the banking industry.