Wohpe

We recently returned from our traditional Dolomites holiday period, and while there I eventually managed to read the first novel written by Salvatore Sanfilippo aka antirez. Most of you probably know Salvatore because he has been for a long time the leading developer of Redis until he came out from the company that now holds the ownership of the project, some years ago. He often compared the creative work of a developer with that of a novelist and I was curious about his first book.

First of all, he probably started to write that novel in 2019, and I bought it in October of 2022, largely before the ChatGPT and the AI hype. Therefore, it cannot be considered an instant book, written after some big public event, just to ride the wave. It has been written by antirez after two years of efforts and refinements.

The novel is a sci-fi book centered mainly on a couple of interesting themes, climate change and General Artificial Intelligence. Wohpe is the contraction of world hope and the whole story deals with the central idea of using advanced self-aware AI to find the silver bullet to manage the climate change that, in a distopic vision of a not-so-far future, shall cause a certain extinction of humanity.

The novel deals with the long months of development of a large neural network that is so big to reach the fictional threshold of self-awareness and become so intelligent to possibly find a final solution to the climatic catastrophe.

I will not give more details about the plot to avoid too much spoiling for people interested in reading the novel themselves.

Fun fact: Wohpe is also the name of the spirit of Peace for the Lakota people - a Sioux tribe - and that could be not so casual after reading the book.

About the book

I found the writing style in some way very similar to that of classic sci-fi novelists I like, such as Isaac Asimov. If you read the classic Foundation trilogy, you can recognize the same slow times of development, with sudden improvements in the narration until the climax of the novel.

A style I personally like, but a lot of people dislike, as well. That motivates probably some negative comments I read about the book, considered too slow or full of unnecessary details that render some parts of the story heavy. I found instead the book well-balanced and of course with the reasonable amount of technical details about AI technologies that one could expect by antirez. We are nerds, after all.

I found much of the story well-centered and convincing about a possible future for AI applications that are currently in their very beginning.

The story focuses on AI applications, which are currently in their infancy, is both compelling and thought-provoking. It presents a convincing and potentially realistic future for those technologies, which I found to be a fascinating aspect of the book. Even, antirez's undoubted technical competence ensure a well-constructed scenario for their applications, with the right weighting of details and no improbable or too imaginative ideas about the reality of neural networks.

Even in a distopic context, the book is positively optimistic, maybe too much for my inclination. I would like to share antirez's faith in human capabilities of dealing effectively with events that largely go beyond the knowledge and understanding of the average individual, as well as the possibility for single individuals to be relevant for humanity fate.

Let me discuss now the two main topics of the book and my personal ideas about them.

Artificial Intelligence, the next revolution

The hypothesis of a theoretical network size limit (or number of parameters) for self-awareness it too simple to be true, but a fascinating one, indeed. But for that, most of the applications of AI as described in the book are practically possible in the next future, maybe even before the next 10 or 20 years term: in some contexts, things will change heavily in maybe less than 5 years, in my vision for the next future on those regards.

The idea that the AI shall improve effectively the quality of life of human beings, after a period of adjustments and a mixture of hope and fear for the new wave, is a credible scenario. We already saw the same for Internet in the latest dozens of years, another technology that changed the lives and jobs of a large part of people in many countries. I predicted the Internet revolution in the beginning of 90s, I'm quite sure that even in this case, I will be right: if an excessive regulation (driven by human fear of novelties) will not dismantle the whole thing at the very beginning, which could be one of the possibilities in this crazy world.

Climate change, the next challenge

I am more worried about the implication of the climate change in the medium term for humanity. We already are missing the inital goal of limiting the increase in average temperatures under 1.5°C before 2050, and I seriously doubt that our models can predict effectively changes in our daily life that we will have to deal with in the next few years.

We seriously could face big problems in many parts of the world a lot of time before a Wohpe can help us. And on those regards we are not brilliant in predicting disasters and acting effectively. We are generally and simply regretting - disaster after disaster - for what we did not do instead.

Here in the Mediterranean area, we already are experiencing hot summers with weeks of continuous high temperatures and humidity, reduced rain (even during the winter) in many urban areas, as never seen until a few years ago. Unfortunately, for the average citizen, it seems a problem of someone else, with even a sense of nuisance for the few initiatives adopted to limit the worldwide CO2 production.

Under those regards, we have not a better attitude than that of the silly folks shown in the Don't Look Up! movie, just waiting for the meteorite.

I don't know if an AI can help us with all that. For sure, I would prefer that people generally be aware of the problem and all act as a consequence even in their daily lives.