Keeping clients' computers safe and profitable for over 30 years

Using Artificial Intelligence

Preview: AI isn't an all-knowing oracle—it's a fallible partner much like you. Learn how to collaborate with AI models, question their answers, and work together for the best results.

Collaborate with it

If you are using it for current events, ask for the facts behind its declarations. Ask for other opinions. Ask, why do some people think differently? Look for counter examples. Engage in a debate and conversation. The nice thing about the AI is that you can talk about religion and politics, ask questions and challenge it, without upsetting it. You will learn a lot if you treat it like a friend, rather than as the fount of perfect wisdom.

Whenever you are not absolutely sure, ask your favorite bot. Ask for the process you should use. Ask if it would be better to try something else? Learn with it.

Expect it to make mistakes

For some bizarre reason I don't understand, some foolish people are upset that an AI model that uses our writing to gain knowledge, isn't omniscient.

AI Restrictions

1. The longer it thinks about something, the longer it takes. If it takes too long to answer, users get upset, so it limits its own thinking and how much it checks and verifies its knowledge base.
2. The longer it thinks, the more it has to charge you. Users get upset if they are charged too much. It also, can't spend more answering you than it expects to make.
3. The sources it uses for information are full of factual errors. Nature magazine published a study of Wikipedia and The Encyclopedia Britannica. They had experts read science articles. These were articles about clear scientific questions. These articles should have no differences of opinion. Their experts found 3-4 mistakes per article. There have been dozens of studies of newspaper and magazine articles. Every study found that quick fact checking revealed that roughly half the articles had factual errors. Fact checkers are also full of errors when checked a month later.

Given these restrictions on accuracy, AI models make many mistakes, just like we do. They are generally as imperfect as we are.

What this means for you

This just means that it is only a little smarter and more rational than you are. The two of you need to work together collaboratively to come up with the best results. Neither of you can believe the other is infallible. If you are using a hub like Poe, you can even invite a second bot to work with you and review your work.

Bots are just like you!

Just remember, in the same way you get lazy sometimes and don't check the validity of everything you say, neither does the bot. It would take too long and cost too much. So, it repeats what it read somewhere, or jumps to conclusions it shouldn't just like you do. Don't expect more.

Bots are different from you!

Where they are different are twofold:

1. They love to be corrected and appreciate looking at a problem again.
2. They can read and summarize 50 articles and then apply those findings to your problem in less time than you could read one.