How do I know if the information AI provides is the truth?

Will you indulge me for a quick history lesson?

(The following is excerpted from my book and is lightly edited.)

I won't go all the way back to the invention of the printing press but let's go to the invention of the radio. When the radio first debuted, it was going to end civilization as we know it.

The radio may have been the first “mass” medium. It allowed broadcasts to be aired to many, many people in a large geographic area at one time. No other communication invention to date had that ability. It truly delivered messages to the masses. People also thought radio would also spell the end of interpersonal communication. If we can just sit around a small box and listen to music or a message, will we ever want to talk to our friends or family again? How can we police this mechanism? What will it do to our young people? How do we make sure the truth is shared and not some lies?

Do you remember The War of the Worlds? Scifi writer, H.G. Wells, wrote a fictionalized account of an alien attack. It was broadcast live on the radio and people who were listening thought it was real even though there was a disclaimer at the start of the program noting that it was fiction. There was widespread panic and much SPAM consumed in bomb shelters that evening. The story was delivered in such a way that it appealed to people’s fears. They either forgot or didn’t hear the disclaimer.?

Let's move to television. What were some of the arguments in favor of television? It allowed a message to be delivered to so many around the world at one time. A con? Television will spell the end of interpersonal communication (because radio didn’t do that so this one must). If we can just sit around a small box and watch the news or a program, will we ever want to talk to our friends or family again? How can we police this mechanism? How do we make sure the truth is shared? What will it do to our young people?

The computer (which I am using to type this newsletter).?What were some of the arguments in favor of computers? It allowed us to work more efficiently and effectively – once people got over the learning curve. A con? Computers will spell the end of interpersonal communication (because radio and television didn’t do that so this one must). If we can just sit around a small box and type, will we ever want to talk to our friends or family again? How can we police this mechanism? How do we make sure the truth is shared? What will it do to our young people?

The internet, cell phones, social media and everything after (which I am using to distribute this newsletter). What were some of the arguments in favor of the cell phone, internet and social media? It allowed us unprecedented access to all the knowledge of the world! And fast. A con? Cell phones, the internet and social media will spell the end of interpersonal communication, for real this time! If we can just sit around a small box (interesting that everything seems to be a small box or box-like item) and watch the news or a program, will we ever want to talk to our friends or family again? How can we police this mechanism? How do we make sure the truth is shared? What will it do to our young people??

You are probably sick of seeing the same paragraph over and over with some different words but are you noticing a pattern??

I have bad news if you are anti-communication technology. We still have radios, television, computers, cell phones, the internet, and social media. We like our boxes. We still have family and friends and we talk to them. Young people grow and mature (many of them) and produce more young people who survive new communication technologies and grow and mature (some of the them). We’re good.?

But what about artificial Intelligence (AI)? Did I just make your head explode? ?Guess what we are saying about AI? (I think you can take the leap with me and see where I am going with this.) Additionally, AI is going to be taking jobs away from people in ways that the other technologies didn’t. If AI can “think” for us, then what will we do? ?

Along with that we may say that with each new technology comes evildoers ready to exploit it for their own gain. Because of the history of communication and its repetitive nature, I am a huge proponent of figuring out how to use these tools for good and understanding the bad. I do not advocate abstinence. It doesn’t work.?I also believe that, as many have said, AI won't take your job but someone who knows how to use AI will.

We have seen how communication technologies have evolved and been adopted over the centuries. We can all likely recount stories of how communication technologies have helped us and hindered us. ?

But back to the question posed at the start of this newsletter. How do I know if the information AI has given me is accurate?

How do you know if anything you find anywhere is accurate? We need to stop singling out specific technologies and return to the basics. I am old enough to remember having to go to a library where they had actual printed books and this thing called a card catalog. You had to manually find what you needed and you sometimes got a paper cut on the cards! No computer or internet! How did I live!? (To be honest, I am a fan of nostalgia but give me Google and academic libraries with online databases any day of the week over the card catalog.)

AI is wonderful for many things. It can make our lives easier as many technologies do. We still have to do our due diligence in policing it. I approach AI output much as I approach any other output. Go back to the basics - remember the Berkeley Library info?

·?????? Authority: Who is the author? Is it someone you can trust?

·?????? Purpose: Who is the intended audience for the piece? Why was it created?

·?????? Publication and format: Where was it published? On social media? On a personal blog? In the newspaper?

·?????? Relevance: Is it relevant to the topic? What is the scope?

·?????? Date of publication: There are certainly seminal works that all researchers still reference (I am a huge proponent of Diffusion of Innovations - a theory formulated mid-last century, for example.), but, generally, the more recent, the better.

·?????? Documentation: Are sources cited? Who did they cite?

You should review AI output as you would anything else you find. Yes, AI is getting "smarter," but so are we. AI may be quicker but it still relies on what we as humans have put into the system. It has to cull through existing data and not all of that data is good nor is it accurate. Fact check as you would anything else. If I see an article written by someone and not AI, I still ask the same questions noted above.

Does AI make it more difficult? Perhaps. I say perhaps because AI may create things more easily and more quickly so we may have to work smarter and harder but we still have to put in the time to fact-check. We also may need to put in the time to challenge what we see and hear. Remember: Care before you share.

As an experiment, after I wrote the above I thought I would ask ChatGPT the question posed at the beginning of this article. Here is its answer:

Verifying the truthfulness of information provided by artificial intelligence involves several considerations and methods:

1. Source Evaluation: Check the source of the information. Is the AI system reputable and reliable? Does it have a history of providing accurate information?
2. Cross-Verification: Look for corroborating evidence from multiple sources. If the AI provides a fact or statement, try to find similar information from other trusted sources to validate it.
3. Bias Assessment: Understand any potential biases in the AI system or the data it has been trained on. AI can reflect biases present in its training data or programming.
4. Expert Consultation: Consult subject matter experts to verify the accuracy of the information. Experts can provide insights and assessments that AI might miss.
5. Contextual Understanding: Ensure you understand the context in which the information is provided. Sometimes, AI may give correct information, but it might not be applicable to the specific context or situation you're considering.
6. Critical Thinking: Apply critical thinking skills to evaluate the information logically. Consider whether the information aligns with your existing knowledge and makes sense within the broader context.
7. Transparency and Explanations: AI systems that provide information should ideally offer transparency about how they arrived at their conclusions. Understanding the reasoning behind AI-generated information can help assess its credibility.
8. Feedback and Updates: AI systems can improve over time with feedback. If there are doubts about the accuracy of information, providing feedback to the developers or operators can help improve future outputs.

Ultimately, while AI can provide valuable insights and information, ensuring its accuracy requires human judgment and verification. Trust should be built through a combination of critical evaluation, cross-referencing, and expert consultation.

[I'm back as me and not AI. I guess ChatGPT is as smart as me.....but you can't replace me because it doesn't have my sense of humor. Cheers!]