Slime molds & worm brains

Claire L. Evans: I like to get dirty when I'm gardening. I don't use gloves, I get dirt under my fingernails and carry it into the house on my shoes, I get splinters and I don't care about insect bites.   

That's a matter of definition. Our understanding of technology today is very one-dimensional, technically characterised by the history of computer science as an industry. But ‘computing’ is much more than that. Every living being does it.  

The act of computing, or rather information processing, can be applied to almost anything, everything and every living thing that acts dynamically in the world and thus shapes reality.   

Think of a worm. It is a simple organism, but in order to survive it must process and store information from its environment—things like temperature, the presence of food and predators. When it makes decisions based on this information, this is also a form of ‘computing’. We have tried to create computer models of these basic functions of a worm’s brain—it’s extremely complex, and takes enormous computational power. But a worm does it with hardly any energy at all.  

Ant colonies use algorithms that work in the same way as our internet data routing systems. Slime moulds find the most efficient paths through uneven terrain. Our network design can learn from this. The more we learn about such organisms, the clearer it becomes: Much of what we think are human inventions have existed in nature for millions of years. If we put our ego aside, we can learn an incredible amount from them.  

Nothing special, we never talked about that kind of thing. Although I do believe that children need some kind of spirituality. So I invented my own rituals because I knew there was something bigger. I prayed to my grandparents' dead dog, even though I didn't even know what “pray” meant.   

I was an only child from an immigrant family in the US. My mother was French, my father British. I was more of a loner. I didn't get on badly with other children, but I liked being on my own. I made up stories or sometimes just stared at the ceiling. This boredom was almost meditative. I still like being alone today.    

As a sci-fi fan, I used to think AIs were humanoid robots with consciousness. When I worked with them myself, I realised that it was just complicated mathematics. I tinkered with AI by chance at a particular moment of transition, when it was still primitive and interesting. You had to intervene as a human to get something useful. I found that thought comforting.  

I think human intervention is still needed to make something interesting. Today, however, these tools present themselves as if they can do everything with little user input.   

Yes, I've seen how pointless things have been done with these tools. I've seen evil being justified, but I've also seen wonderful things. So I'm not a technophobe. I see it more like language.   

Historically, we've had more time to understand such huge technological changes, such as photography, which took a century to get rolling. We’re still working on questions such as: What is a photograph, what is reality? Today, these changes are happening rapidly and constantly – too fast to categorise and understand.    

The appeal of AI back then was how quirky and unpredictable it was. Whether you were generating images, text or sound, it always felt like it was coming from a computer, with that unique soft accent. A tone of voice that somehow sounds wrong. But that accent has disappeared. It's as if the computer has learnt our language much more fluently. 

One of the first big AI machine-generated images to emerge in 2015 was Deep Dream by Google. It generated these images that were absolutely crazy and disgusting. It was completely unlike anything a human would ever create, which I found endlessly captivating. Today, AI just generates boring stock photos.  

No. The name is a bit embarrassing – YACHT sounds like luxury. But it comes from a dilapidated building with the sign ‘Young Americans Challenging High Technology’. My partner, who founded the band in 2002, was so fascinated by this enigmatic but also symbolic scenario that he named the band after it.   

Yes and no. Not as Luddites, but in dialogue: How do we use tools to make art that is bigger than ourselves? We made an AI album in 2017. The lyrics, the cover, the videos and the merch were also created with AI. That was a one-off experiment. After that, we no longer wanted to be an ‘AI band’. Now we're in a ‘post-AI’ phase and are going back to being more human.  

No, never.

I don't find it interesting. When it was still rather bad, I found it exciting to figure out how machine learning interprets input that it doesn't understand, how it messes up all these variables and presents you with a guess about what it thought you wanted. These were ideas that were just bizarre and very inhuman, but poetic in their own way. The dialogue between the human and the inhuman is what really appealed to me about working with AI. And now these tools are just too good.   

Recognising real dangers through fiction – and taking countermeasures. I'd rather read about a fictional dystopian future than actually experience one.  

Tim Maughan describes a scenario in the US that brought the production of iPhones back to the United States due to a trade war. The only way to make this economically feasible is to have prison inmates build our phones. It's also a love story about an inmate trying to hack into a mobile phone in this factory to send a message to his sweetheart.   

Badly, but I'm working on it. Instagram is not good for me, and I only download it when I need to post something for work. Then I delete it again. In the meantime, it's no longer important to me to reach an audience of millions at some point. I'd even prefer to work on my article for months, get paid and have it disappear into a drawer without ever being read. I want to learn new things through my work, and the audience is secondary. 

Absolutely! The best life hack for creative work: never think about what others want, otherwise you'll go crazy.  

Simply because I couldn't find any information about them anywhere. So I started looking for their names – often they were only in footnotes or as the fifth author of some paper. The more I talked to these women, the clearer it became: There was this huge untold story. Women who were there from the beginning, who made crucial contributions and then were ignored. I see the book as a kind of corrective.  

Before the web, when the internet was used more for academic or military purposes, there were no search engines. If you wanted to find information, you had to call the Stanford Network Information Centre. There was a woman on the phone who was the only person who knew what resources were available where. She was Google before Google existed. For 20 years, the Internet had a phone number and if you called, a damn smart woman would tell you where to go.  

Often. Early female programmers? All women, because programming was considered ‘secretarial’ work until men recognised it as valuable. The network information centre? Indexing information was considered unimportant, like filing work – until search algorithms became valuable and Google built an empire on them. In every tech era, there have been women who invented things and were ousted as soon as their inventions became ‘valuable’. This is a lesson for all fields.  

That's the interesting thing about AI: it is based on historical data sets, on the entire corpus of human creation. This is an opportunity to examine our past: Who are we? What does this data represent? How can we prevent ourselves from repeating the same problems? Sure, the amount is overwhelming – but much like with our history, we can analyse it or stumble blindly into the future.