Microsoft, Google, Amazon and tech peers sign pact to combat election-related misinformation::A group of 20 leading tech companies on Friday announced a joint commitment to combat AI misinformation heading into the 2024 elections.
Microsoft, Google, Amazon and tech peers sign pact to combat election-related misinformation::A group of 20 leading tech companies on Friday announced a joint commitment to combat AI misinformation heading into the 2024 elections.
Concept is, end user reports misinformation, fact checkers at the company take the reported information and check it against the database. If it’s listed as false in the database It gets squelched and the AI gets a little tuning to make sure that it stays squelched. If it’s in the database and it’s true the user is informed that it’s not false information. If it’s not in the database, That’s when it’s dicey. Does the team of people moderating the posts make the call, does it go to another team to be classified. At what point do you block it? If one details wrong if two details are wrong if half the post is wrong. Do you squelch mostly true? Or do we just get disclaimers everywhere for 6 months.
I’m mostly puzzled by how this would be carried out when the vast majority of information seems to be discretionary, interpreted, perceived, opinion. Like the statement I just made ;)
Facts either are or aren’t.
Misinformation is vary more challenging because it’s usually derived from an event that was a fact, but the interpretation, analysis, significance, etc is based on the person’s bias.