Post

When anthropomorphized, the machine becomes the actor. And that is perhaps the main point of evading liability. It goes rogue. It decides. It escaped! The language about AI is all about perception. [If you like what you read, please subscribe! It's free!] collectivefutures.blog/the-strange-...

collectivefutures.blogThe strange lives of an irresponsible machineFrom an autonomous car killing a pedestrian to the ideology surrounding artificial general intelligence, the same pattern keeps appearing: the machine becomes the explanation, and responsibility disap...

AI is either "inevitable" - so we have to build it because someone else will... ... or it is so dangerous that only the people who understand it can safely build it. These sound like opposite arguments but they both arrive at the same end goal, putting the "machine" at the center.

The interface itself reinforces this story. AI systems reply in the first person. They apologize. They describe intentions and they appear to make decisions as if they were acting independently. They are designed to look like an actor. And then this becomes also "evidence" that it has become one..

0 reposts0 likes

On top of it of course stis a larger ideology TESCREAL, techno-optimism basically bundles of cyberlibertarianism mixed with fascism. Workers, energy, water, infrastructure... become “context” for the spectacular question: What happens if the machine becomes superintelligent?

But the boring questions are often the politically important ones. Who owns the compute? Who approved the deployment? Who gets paid to label the data? Who authorized the system to access someone else's infrastructure? Who benefits? Who bears the cost? What alternatives were available?

Machines do not go rogue. People make decisions, and then we describe the consequences as if the machine made them. Once the machine becomes the explanation, the people who made the decisions disappear. Systems have no alibi. People do. And responsibility always has a name.