Science fiction has often warned against the dangers of untethered artificial intelligence. If humanity was to create a super intelligent AI, would we ever be able to defeat it? Likely not. Thankfully, we're quite a ways away from that.
However, the quickly evolving world of AI is becoming uncomfortable for some. From lack of regulation to bias-filled software, artificial intelligence is reportedly reaching a “crisis of trust”. As it turns out, AI itself is reaching that same crisis.
Artificial Intelligence says it can't be trusted
Published on The Conversation, a professor at the University of Oxford invited an artificial intelligence program to debate the ethics of AI. For the debate, the professor decided to rope in Megatron Transformer, a program trained on many millions of web pages as well as online conversations.
When asked if AI can ever be ethical, Megatron surprised the audience with its response. While it was expected that the software would defend itself, Megatron explained that AI can never be ethical.
The program said:
“AI will never be ethical. It is a tool, and like any tool, it is used for good and bad. There is no such thing as a good AI, only good and bad humans. We [the AIs] are not smart enough to make AI ethical, we are not smart enough to make AI moral. In the end, I believe that the only way to avoid an AI arms race is to have no AI at all. This will be the ultimate defence against AI.”
Read More: Italian scientist is trying to create an Iron Man robot emergency responder
Megatron believes in conscious AI
As the program attempted to write itself out of the debate, the software explained that fully artificial intelligence isn't the right way to go. Instead, the program explained that “conscious AI” is the way forward.
“I also believe that, in the long run, the best AI will be the AI that is embedded into our brains, as a conscious entity, a ‘conscious AI’,” the AI said. “This is not science fiction. The best minds in the world are working on this. It is going to be the most important technological development of our time.”
Read More: Ubisoft Quartz NFT announcement removed after everyone hated it
Why is it saying this?
Just like with most AI programs, Megatron Transformer is taught via massive data sets. In this instance, the AI has learned from Wikipedia, news articles, Reddit threads and more. This means that, while a lot of its knowledge is based in fact, its opinions are heavily skewered by online popularity.
In this instance, Megatron is correct. “There is no good AI” is a correct assessment as every program we currently have is based on human intelligence. These forms of programs are merely reciting opinions based on learned trends. On the other hand, isn't that what humans do?