How Did I Spend My 500 Days Writing About Artificial Intelligence?

Original article was published by Alex Moltzau 莫战 on Artificial Intelligence on Medium


How Did I Spend My 500 Days Writing About Artificial Intelligence?

Looking back on writing every day for 500 days about AI

It seems like an eternity since I decided to write an article on artificial intelligence (AI) every day for 500 days. My friends thought I was crazy, and others said it was impossible. It has been a nice journey, with some heavy and many good days . Now I have to look back on what I have learned, where we are in the present and where I should go next.

Why?

Before the first day, one of the reasons I started writing was a random conversation with a weapons engineer building autonomous weapon systems. I had previously been involved in sustainability, so when I heard about the rapid changes in AI, I was worried. At the same time, it struck me that this technology could be used for something else.

1. Starting a crazy project

The first 100 days of writing were a lot about mapping definitions of AI, security and climate considerations. It was a lot about getting into a routine, writing every day and at the right time. Before I got into good routines, there were many late nights, and my wife was not always happy with my new project. After a hundred days, it became clear to me that nations around the world had begun to make their own strategic plans to build capacity within AI.

2. Becoming engaged

From day 100 to 300, I decided to go through all the national and international strategies for AI in the world. So, I read the strategies and ended up talking to many companies and researchers who are developing AI. In addition to this, I chose to contribute at an input meeting when the Norwegian strategy for AI was developed. Eventually I was also invited to conferences to talk about the topic. It could seem as if the whole world was fighting in an international race for advanced AI.

3. The pandemic

Then came COVID-19. When COVID-19 came into being, digital infrastructure became even clearer as a critical infrastructure in our society. Communicating digitally, and for parts of the population to be able to work digitally, has been important for our society. For the first 300 days, I had seen how little focus there was on the physical effect of AI in supplier networks in terms of climate. Everything that is digital has a physical footprint. It is said that the entire data center industry emits as much carbon as the world’s air traffic. Therefore, I decided, days 300 to 350, to focus on how physical digital infrastructure for AI develops, and the consequences of our increased technological consumption.

4. The social consequences

At the same time as COVID-19, there were major protests due to the enormous social inequalities we see in the world. Black Lives Matter could be seen abroad and here in Oslo — at the same time many became unemployed. So I wrote about social inequality and AI from day 350 to 400. IBM, Amazon and Microsoft chose to press the pause button on the use of AI for recognising faces in the police for a year, possibly after realising that avoiding discrimination was not possible at the time.

5. Application and summarising learnings

Day 400 to 450 I started working more actively with language processing at work and thought that rounding off in a practical manner could be good. The last few days I am casting my net wide and I am trying to reflect on the days that had passed.

I look back now and think that AI is already used for far more than weapon technology. We need a more conscious relationship with the use of AI in general to think about nature and people in interaction.