PORTS 2038: HUMANS VS. AUTOMATION
A world where artificial intelligence (AI) has advanced to the point where we can compare it to a human and judge it by its actions accordingly is the scene set in the latest instalment of a futuristic port story.
In Part II of ‘2038: A Smart Port Story’, Inform’s Dr Eva Savelsberg and Matthew Wittemeier return to the character of 'Douglas' who, is using immersive drone technology combined with artificial intelligence (AI) to assess what happened before an automated machine (AGV) injures a man at the Port of Reykjavik.
Using a 'bio signature' the AI is able to confirm detailed information about the individual. “Heart rate 197, breathing elevated. Combined with his body language, it would indicate high stress and discomfort.”
The story asks us to question both the capabilities and limitations of technology and the line between automation and human responsibility.
Going back to 2036, we learn that a court case dealing with two deaths as a result of an AGV has been closed as AI is not able to be tried and no humans were responsible.
In part one of the story, readers are told that in 2032, ELAIR, the European Legal Force for AI Responsibility, tried to establish a definition in such a way that a machine could be held accountable, but following this no clear definition was struck.
As the story develops we are drawn into an anti-tech movement which criticises corporate greed and uses the internet to build support for its actions. It ultimately and ironically places its goal of bringing down automation, while demonstrating its ability to be a dangerous technology, above all else, including human safety.
Back in 2038, Douglas comes to realise that the person hit was involved the anti-tech movement, but sets about investigating as he realises they expected the AGV to stop. Meanwhile, the anti-tech movement use the accident as leverage for its aim to change legislation to make machines be held responsible for actions.
The investigation reveals adaptive code was being used in the management of the AGV, bringing with it the suggestion that the incident was the result of the system being hacked.
It is later established that adaptive code was introduced to improve productivity but this also brought unpredictability, although human checks and controls were implemented to make sure the AI doesn't understand how decisions are made that involve it.
"A team of humans reviews all this data and debates whether the logic she’s implemented to come to her decision is sound. That is done in this room so she cannot hear or learn from those discussions," it is explained.
We learn the adaptive code was switched off but are told some adaptive ability remains, leading to Douglas interrogating the AI responsible for autonomous vehicles at the port. The question of whether human life can be valued by an AI is asked to which the AI responds "an AI system should be capable of placing value on human life."
It adds: "The nature of adaption is to remain relevant when combining what we know now and our prediction of the future.”
As part II ends we learn the AI doesn't believe the AGV failed to stop due to an AI system failure but acknowledges humans are responsible for its future and that it will likely not exist in its current format in the future.