|
AI Advancements Producing More False Info
AI has been a trending topic for several years now. It's to the point where almost every organization in every industry had heard about some way that AI will be able to assist in your work. A way that it can take something over entirely. But there's still a need to be careful with how much it is allowed to do and a need for oversight by humans.
A recent article published by the New York Times talked about a computer program that's beginning to grow and used an AI for their customer support and marketing purposes. This AI announced a new policy change that told customers that they could only use the program on one specific device. This made the clients very angry and the company a target of online posts bashing them.
The major problem, however, was that this policy didn't exist. It was false information that the AI was guessing was going to exist and that wasn't checked (or at least caught) by a human before that fake info was spread to clients.
The article goes on to discuss how the newer models of many open text AI programs have added more power and have ultimately used that to cause even more errors in factual reports (called hallucinations because the AI is "seeing" facts that aren't there).
This may even be something you notice on Google search - many times when you search now there will be an AI generated response summarizing things for you, but this box often contains inconsistencies that, when you click into the sources it's pulling from, are about another subject, an event from prior years that already happened instead of the new/upcoming one, a different product number from the one you searched, etc.
The incident is a good reminder to temper expectations for AI. It's something that sounds great and very well could be beneficial to implement in spots for your organization, but it should not be viewed as an end-all solution and requires human oversight to weed out the false information it comes up with and should not be trusted to be 100% accurate at any time.
If you have any questions regarding AI or other IT topics for your organization, please feel free to reach out to our offices and we'll look to assist as best we can.
| 
