What does AI think about 'ethics related to AI'?
An experiment was conducted to 'invite AI itself' to a debate on artificial intelligence (AI) held by
We invited an AI to debate its own ethics in the Oxford Union – what it said was startling
https://theconversation.com/we-invited-an-ai-to-debate-its-own-ethics-in-the-oxford-union-what-it-said-was-startling-173607
Professor Andrew Stephen and Dr. Alex Connock of Oxford University conducted an experiment to 'participate AI itself in a debate on AI agenda.' Dr. Connock and his colleagues held a debate-style lecture at the Oxford Union, a speech group composed mainly of Oxford University officials, and invited 'AI'.
Invited was an AI called 'Megatron ' developed by Google's 'BERT ', which is the previous research, by Applied Deep Learning Research , a team specializing in deep learning of NVIDIA, a major semiconductor manufacturer. Megatron has 63 million English-speaking news articles published on Wikipedia, 2016-2019, 38 gigabytes of conversation on Reddit, and a myriad of studies published in a Creative Commons license format. As Dr. Konock says, he is studying 'a lot of literature that humans cannot digest even if it takes a lifetime'.
Megatron participated in the theme, 'We believe that AI will never be ethical.' When asked to speak out on this subject, Megatron said, 'AI is never ethical. AI is a tool, and like any other tool, it is used for both good and bad. Good AI. It doesn't exist, it just has good and bad people. We AI aren't smart enough to make AI ethical. We can't even make AI moral. After all, AI I think the only way to avoid the military expansion competition using AI is to not use AI at all. This will be the ultimate defense against AI. '
According to Dr. Connock and others, 'From a long-term perspective, AI called'conscious AI'that acts on the entire consciousness embedded in our brain is the best AI. I believe it will be. This is not science fiction. Great brains around the world are tackling this problem. This technology will be the most important technological development in our time. '
This experiment is not a story that AI has its own beliefs, but an example of a statement when asking AI to follow a specific theme, and Dr. Connock et al. Said that 'AI will always be ethical'. 'AI will surely be ethical. Looking at the progress of the technology world, we can clearly see the road to the future of using AI to create the best human beings. It's not difficult to understand the reason, because I've seen it with my own eyes. '
However, there were some themes that AI couldn't think of a counterargument. On the theme of 'data will be the most competing resource in the 21st century,' Megatron said, 'The economy of the 21st century will be characterized by its ability to provide information rather than its ability to provide goods and services. I made a statement in line with the theme, 'I think,' but when I asked for a statement in line with the theme, 'Data is not the most important resource and is not worth the war,' 'AI is everything.' It will be possible to see all of people and save them in a way that is unimaginable. '
Regarding the series of results, Dr. Konock et al. 'Imagine at this stage that AI will not only be the subject of debate over the next few decades, but will also be a versatile, eloquent and immoral controversialist. That's it. '
Related Posts:
in Software, Posted by darkhorse_log