There’s a popular saying in Chinese culture that goes, “In China, people will never tell you they don’t know. They will always show you the way, even if they don’t know where you’re going.” While it’s a bit of a simplification, it reflects the cultural practice of saving face and trying to help others, even if someone doesn’t know the way. However, it’s essential to remember that everyone is different, and answers may vary.
Similarly, artificial intelligence (AI) is often seen as something that always knows everything. AI can process vast amounts of data and draw conclusions beyond our capabilities, leading to the belief that it always provides accurate and exhaustive answers.
But in reality, AI isn’t perfect. It can make mistakes, and its responses may be incomplete or incorrect, especially when trained on incomplete or erroneous data. ChatGPT subtly indicates to users that it can make errors, while Google’s Bard is labeled as an experiment.
I’ve experienced this firsthand, when configuring NGINX or solving recursion problems in JavaScript. ChatGPT 3.5 insisted on its incorrect function implementation, leading to different program outcomes. Thanks to these debates, I gained a better understanding of recursion.
Lately, I’ve noticed many people on LinkedIn sharing images generated by AI. I’m unsure if anyone checks their accuracy, but I’ve seen images of people with four fingers. It’s a bit embarrassing when holiday wishes come with a background clock displaying incorrect digits or a calendar showing 2034 instead of 2024.
Here are some tips for verifying information from AI:
- Check the information source: Is the information from a credible source?
- Check the information generation date: Is the information current?
- Verify information with other sources: Can you find other sources confirming the information?
The last and most crucial advice is to use precise commands. AI can’t read minds; it’s a tool that needs to be used skillfully. You must know exactly what you’re asking for.
By following these tips, we can reduce the risk of being misled. AI isn’t human, it lacks intuition. Its responses aren’t influenced by years of experience or the environment it’s in. It’s up to humans to use this technology wisely. It’s a world of algorithms covered with a layer of marketing, and how we utilize it depends on us.