These days, when you pick up a newspaper, newsmagazine, or online news publication, you rarely come across information about artificial intelligence known as AI. For some of us humans reading, we may be AI too.
AI is scary. It makes the observer think of Frankenstein, a story that creates life and fears what is created. A superhuman with superintelligence and superathletics who created what HR now calls “Language Skills” I recommend the original The last where the Doctor chases his creations through the snow The scene is worth reading.)
Given things like this about modern AI, most people are wondering if they should take a moment to brake before it spirals out of control. And we’re not the only ones who think so.
Sarah Myers West, managing director of the nonprofit AI Now Institute, said:
Polls show that an overwhelming majority of individuals in the tech industry believe that substantial government regulation should be considered to prevent AI from being used for malicious purposes.
We are now hearing from Dr. Frankenstein himself. The Times reported last week that AI research pioneer Jeffrey Hinton had quit his job at Google “to allow him to speak more freely about the dangers of the technology.”
“Right now, as far as I know, they’re not smarter than us, but I think they could be soon,” Hinton said.
According to The Economist, creators are concerned about capabilities they “didn’t know would be developed” when the foundations of these systems were laid. And I thought HAL was scary.
What’s changed from previous versions of AI to today? “The first wave of modern AI systems that emerged ten years ago relied on carefully labeled training data. I was able to learn things like transcribing audio.
“Today’s systems don’t require pre-labeling, so they can be trained using much larger data sets taken from online sources.” is.
How much fake do you think there is on the internet, and will it affect AI? Questions abound.
One thing AI can’t do is figure out what’s true and what’s fiction. Even real people have a hard time distinguishing between truth and fiction in what they read, and must rely on the reputation of various news sources to determine whether something is believable. It often happens.
No AI technology we’ve heard of takes this into account. As a result, in order for AI to be the best, most responsible, and honest, we humans need to be the best, most responsible, and honest. If past performance is a prologue to future performance in the realm of truth on the Internet, it’s a tall order.
It’s more complicated than this, but the old adage “garbage in, garbage out” applies here. However, it is progressing at an extraordinary speed, taking in and throwing out garbage. It can run faster than Dr. Frankenstein’s creation.
