Do you trust everything you read online? How about the content generated by chatbots? If your answer is a resounding ‘Yes,’ then let me introduce you to an unexpected voice of caution: Google’s Vice President, Debbie Weinstein.
In a recent interview with BBC, she advised people to fact-check content generated by Google’s Bard AI using Google Search. A bit ironic, isn’t it? A VP at Google, suggesting that we need to double-check Google’s own AI product.
Weinstein considers Bard more of an “experiment” for “collaboration around problem solving” and “creating new ideas”. However, she was clear that Bard wasn’t intended as a resource for “specific information”. Bard’s homepage states ‘it has limitations and won’t always get it right,’ but doesn’t repeat Ms. Weinstein’s advice to double-check results via Google Search.
So, does this mean that we should start treating our friendly neighbourhood AI chatbot with suspicion? Well, yes and no.
Generative AI does have a track record of hallucinating, which means they sometimes generate completely false information. A couple of New York lawyers learnt this the hard way when they used ChatGPT in a case, presenting what turned out to be “fictitious legal research”. Ouch!
However, it’s a little disconcerting when a company’s VP essentially says, ‘Don’t trust our AI too much.’ Especially when the same AI is being marketed as a ‘launchpad for curiosity’; a conversational, context-providing tool for looking up stuff.
And here’s where it gets even more curious. Google recently showcased how Bard could enhance Google Search, from providing in-depth results on a topic to even creating a fitness plan. Now, wouldn’t these features require factual information? So, is Ms. Weinstein implying that this update is all for naught because it uses Google’s AI tech?
The contradictions are puzzling, to say the least. If we’re not supposed to use the chatbot for important information, then why is it being added to the search engine to further enhance it? Why implement something that’s apparently untrustworthy?
In the end, it boils down to this: Trust, but verify. Whether it’s AI-generated content or human-written, a healthy dose of skepticism can go a long way.
#GoogleBard #Chatbot #FactOrFiction
https://www.digitalinformationworld.com/2023/07/fact-checking-bard-ai-google-vps.html