OU blog

Personal Blogs

Asoka

Fatigue, Machine Learning, and Mass Extinction

Visible to anyone in the world

I feel unwell today. And the body is creaky. Fatigue is challenging. It often defeats me. I really didn’t want to get out of bed this morning. It can be a mission sometimes. Cooking felt like a chore. Did manage to rustle up a meal in the end. It is always a relief when I get that out of the way. I don’t have to worry about cooking again or eating for the rest of the day. I try to eat before 3 p.m. and fast until the next morning.

I am studying machine learning at the moment. Did a recap on Python programming, got lost somewhere in arrays and tensors, and the many ways these can be accessed — before I had to stop and have a rest. I will try to manage it in small doses. Learning new stuff can be painful, especially when it comes to computers. 

There’s a lot of hype about A.I. in the news, but it isn’t what we think it is. We like to attribute human characteristics to it, but it is just a machine, an advanced autonotom. It gets fed lots and lots of training data, and if it makes mistakes, it adjusts itself according to some set parameters. It keeps doing this until it gets better at what it does and makes fewer mistakes. But it takes a long time to get there and a lot of training.

The scary thing is people are giving power to these machines to make important decisions. AI is very efficient at specialist tasks and can do them really fast, but only those specific tasks, and it still can make mistakes, it isn’t 100% right all the time. 

If an AI was to try and live a day as a human being or even as an earthworm. It would struggle and fail. It has no experience of what it is to be a human.

 I guess hype sells. It is the next Silicon Valley gold rush. 

While AI does have the potential to be a useful tool, to become dependent on it to run society would be a huge costly mistake and a disaster waiting to happen. Not because AI is evil and wants to take over the world, and not because it wants to destroy humanity, it has no concept of good or evil. It just obeys instructions and does what you ask it to because it is a mindless machine. 

It’s the way it solves problems that may be dangerous. The solution it comes up with may be unexpected and not what one intended. A.I. has no experience of being a human or what it feels like to have a body. So its solutions can be a bit quirky, and unlike anything a human would have thought up. 

It may also worsen social inequality due to inherent bias in the data it is trained on. As well as hallucinate and make mistakes. Not to mention the huge amount of electricity that is needed to run these machines, and the materials used to make them, how that is harming the environment. The truth of the matter is they are designed to make corporations wealthy, not really about making society or the planet better. 

AI is not what we imagine it is when we interact with it. It is not like us, it’s a bunch of algorithms and artificial neurons, and nowhere near as advanced as a human being.

Maybe one day it may get that advanced, who knows, but it is not there yet. 

There are computer chips now in development that use human lab-grown brain cells, called neuromorphic chips, which could have the potential to become sentient. They are being designed because they are more efficient at using electricity. But it is rather creepy, and I am surprised scientists were even allowed to do this. If these part biological machines do become sentient, it would be cruel, as the corporations that made them will claim ownership over them and of course, deny they are sentient. It raises all sorts of ethical quandaries.

Still, humans may not survive long enough to see really advanced AI, especially with the way modern consumerism is consuming the planet. Greed is insatiable, and all this industrial pollution is not just affecting the wildlife, it is affecting us too, our biology. We are part of the natural world, what kills ecosystems, also kills us.

 ...


Permalink
Share post