More and more AI tools are increasing in popularity as they are being used across many Industries and jobs. People are using AI tools to gather information about a topic they have interest in or using it to make essays or emails sound more professional. People are even using it to come up with funny slogans or ask it for help with handling a difficult situation. The opportunities are endless. But are there risks?
With implementation of AI tools growing, the dangers also grow. Although you are lead to believe everything AI tools tell you, typically AI produces what it believes is the most likely answer for your prompt, which does not always mean it is factually correct. Want to ensure the information you are getting is the truth? We recommend fact checking.
In this blog, we will look at the accuracy of AI tools and cautions that you should be aware of and how to ensure you don't fall victim to inaccurate information when using AI generated tools. Let's get started.
How does AI Work?
AI's ultimate goal is not to assess if the information is correct, but to generate what it thinks is most likely the right words to answer the question. Sometimes the information they
provide is true, where other times it may seem correct but isn't. Not only can the information be incorrect, but it can sometimes make up false information.
Since using AI tools is a relatively new thing that a good amount of people may not have experience using yet, there can be more run ins with misleading information getting out there. To ensure the information you are provided is correct, you should start fact checking information that AI tools give you.
AI Platforms Available
There are several AI tools available right now like ChatGPT, Gemini, OpenSea & Grok. Depending on which one you use, sometimes these different platforms have different biases and offer conflicting information. Be sure to do your own research and double check the facts.
Fact Checking
Typically AI tools will quote its sources but not always. Sometimes it can be tricky to validate if the information is true or not. You can use fact checking techniques and consulting other sources to confirm what AI has provided to help protect yourself from using false information.
AI tools are large language models which means that they don't really understand what they are saying. They basically just repeat the words to make it seem like they do understand. AI tools do not have the same thoughts that a human has; it can only repurpose and build upon what knowledge you give it. AI will use the entire Internet to find its answer, which is why it can be hard to trace.
Ways you can accomplish fact checking could be researching for correct information from a notable source and breakdown all of the information. Then you can figure out specific claims, deciding which part should be double checked. Often times an AI tool will provide you with a source to double check, sometimes that source will have contradictory information right in the article. You can then check another source to validate the source that AI provided you to confirm that it is the correct information.
AI tools learn from you and gain knowledge every time you ask it a question. The way you ask a question or prompt is also something to be aware of. Sometimes, AI doesn't fully understand the question just because of the style to which you asked it.
Be sure to be aware of how you ask a specific question, being as detailed and specific as possible. By doing this, you are trying to eliminate wrong information being spit back out. This may help, but it is still recommended that you check another source to ensure the information AI gave you is accurate.
AI Example
Let's try an example. Below, I have asked ChatGPT about the recommended amount of Check Valves that should be used in a Water System, here is the answer they provided.
How AI can Sometimes be Wrong
Boshart Industries actually recommends to place a second check valve no more than 25 ft above the water pumping level and then any additional check valves a maximum of every 200 ft in the drop pipe. This would be an example where you would want to find another source to ensure that the information ChatGPT provided is accurate and actually answers the question you asked them.
In Conclusion:
Although AI tools can be extremely helpful and save you lots of time and energy, it is important to understand the dangers of spreading false information. If you are going to use an AI generated tool, be sure to fact check to ensure the answer to your prompt is true and accurate.
Have questions about Boshart products?
Head over to Boshart's Knowledge Base: technical product information, guidelines, and more.
SHARE