Safety & Ethical Considerations

Before we start, here are a few things worth keeping in mind whenever you use AI. None of them are reasons to avoid it. They’ll just help you use it wisely.

1. AI can get things wrong

Tools like ChatGPT and Claude are called large language models (LLMs). They don’t look up facts the way a search engine does. Instead, they write the text that’s most likely to come next, a bit like the autocomplete on your phone. And just as autocomplete sometimes changes kate to late, AI can make mistakes too. Sometimes it says something that sounds confident but isn’t true. This is called a hallucination.

For anything important, like health, legal, money or safety questions, always have a person check what the AI says.

2. AI can be biased

AI learns from huge amounts of text written by people, so it picks up the same biases people have, about race, gender, culture and more. Whenever you use an AI tool, it’s worth asking: who might this work less well for?

When you type something into an AI tool, you may be sharing it with the company that makes it. Keep personal details, health information and anything confidential out of public AI tools, unless you’ve read and understood how the company handles your data. This matters most for information that isn’t yours to share, like customer or workplace details. If you’re not sure, leave it out.

4. Stay in charge of what AI does

Some AI tools can take actions for you, like sending emails, browsing websites or running code. Think of them like a keen, helpful new assistant: decide what they’re allowed to do, and make sure they check with you before doing anything that can’t be undone.

5. AI uses a lot of energy

Building and running large AI models uses a lot of electricity. It’s worth thinking about when AI is genuinely useful, and when another approach would do the job just as well.

6. It’s not always clear who’s responsible

If an AI tool causes harm, like losing someone money, spreading wrong information or damaging a reputation, who is responsible? The company that made it? The person using it? The law is still catching up, so it’s sensible to think about this before you rely on AI for something important.

That’s the serious part done. Let’s get started!