Saturday, 25 March 2023

Valuable Humans in Transit by qntm

Anthology of short SF Stories. Strong on SF and Technology concepts, and Philosophy. Not so much on the literary side I guess. Not that I mind this so much because I really enjoyed reading these stories. All of them are thought provoking. 

Lena - about the first consciousness to be successfully uploaded on a computer simulation. It touches on the implications for the original human, and the copies that are made of it. This was my favourite story (1). 

If you are reading this - a man meets with an old astronomer who had previously found a message from outer space. 

The frame-by-frame - it follows an autonomous car computer and how its various functions decide action when a person walks into the road. 

The Difference - it might be about one of those AI apps which are popular nowadays or it might be about a real human being asking for help. This was my third favourite story (3). 

Gorge - a spaceship finds a strange planetoid inhabited by nanobots. Cripes does anybody remember google this is people chatting about Google People. Not the strongest story I think. 

Driver - another story of uploaded consciousness. This time uploads are used to control large numbers of other virtual images. 

I don't know Timmy, being God is a big Responsibility - scientists create a simulation of reality. It's so real and accurate one can see earth develop like it did in real life, and humans and civilisations. One can zoom in and see the scientists themselves working on a simulation within the simulation. This was my second favourite story (2). 

A Powerful Culture - about pollution and clean use of resources and humans from another dimension. Valuable Humans in Transit - AI trying to save humanity. Shares third place with The Difference (3).

Monday, 20 March 2023

The Poe Clan by Moto Hagio.

Short Vampire stories or as Hagio calls them "Vampirnella". The main character is Edgar, a 14 year old who is transformed into a vampire by the leader of a Vampirnella clan. Because of this, he became immortal but will never grow up. Edgar is an interesting character. Sometimes lonely, sad and frail but others cruel and sadistic. He loves his sister Marybelle and his friend Alan and would do anything or everything for them. Themes to think about are immortality, loneliness, life and death. The stories start from Edgar and Marybelle's transformation, in the 18th century, to stories in the 19th and 20th centuries. I think my favourite is one in which a group of people gather in the early 20th century to compare stories about a strange boy called Edgar who is mentioned in various writings and paintings dated since the 1700s. They want to know if this Edgar (and Marybelle) is the same person.

In our own image. Saviour or Destroyer? The history and future of Artificial Intelligence by George Zarkadakis.

From prehistoric Australopithecines to Modern Humans to theories about the future of AI. This book is dense with information. Not difficult to read but I took my time to digest it. I enjoyed very much the discussion on how humans developed social language (social interactions, tool making, hunting) and then, thanks to some mutation of gene FOXP2, developed general purpose language. Zarkadakis says that we developed General Intelligence from General language. General Intelligence gave us high level consciousness. Through history we have referred to this as soul, mind, self-awareness, subjective experience (qualia). Zarkadis discusses how through the ages philosophers, thinkers, scientists have thought about us and our intelligence. This explains why we think the way do about AI nowadays: as if AI will be like us, self-aware... However in reality this won't happen, we are subjective beings but subjectivity is not logic and therefore it cannot be coded. 

... and the Moravec Paradox, Moravec says that it is comparatively easy to make computers exhibit adult level performance in intelligence tests but difficult or impossible to give them the skills of a 1yo when it comes to perception or mobility. In other words coding cognition is easy but coding sensing and action isn't. What this means is that it is (highly) intellectual jobs which will be taken by AI, not manual ones. Janitors from the future will be better off than lawyers.