Dem
51
GOP
49
We and many others have been warning about the potentially huge and harmful effects of AI in 2024. Most AI systems are trained by feeding them the Internet. They absorb all the information there and "learn" from it. However, there is a popular saying in the IT community: "Garbage in, garbage out."
Case in point: Amazon's Alexa. When asked who won the 2020 election, Alexa is saying it was "stolen by a massive amount of election fraud." It cites Rumble, a right-wing Twitter competitor, as the source. Alexa also says races were "notorious for many incidents of irregularities and indications pointing to electoral fraud taking place in major metro centers." It got this from Substack. It also says Trump won Pennsylvania, citing an Alexa answers contributor.
Even though Alexa is spewing total garbage it found on the Internet, Amazon is promoting the device as a reliable source of election information to its 70 million users. When The Washington Post article about Alexa was brought to Amazon's attention, the company claims it fixed the problem. So we now have a situation in which devices to which millions of people turn to for information get fixed only when a national newspaper points out that it is spewing harmful nonsense. What happens when no newspaper catches false information some AI system is spewing?
Voice assistants like Alexa, Siri, and others are only as good as the information they can find on the Internet, a fair amount of which is completely bogus. None of this software understands what it is doing. It can only repeat what it found, and has no way of telling whether an article on Rumble is any better than an article in the Post. But many users think that the information has been gathered or is approved by Amazon, Apple, Google, or whatever tech company built the gadget or wrote the software, so they believe it. They don't understand that "I found it on the Internet" is not proof that it is true. All we can say is "buyer beware." (V)