Seventeen years ago, Apple rolled out the first smartphone. A computer AND a phone AND a camera all wrapped up in a gadget the size of your palm. Despite the high price tag, the smartphone lured many of us into Apple stores to buy one for ourselves. And then, eventually, our kids started asking for one too. When social media really started picking up, and our kids had grown attached to their phones, many of us found ourselves in over our heads.
Now that large language models (LLMs) like ChatGPT, Gemini, and Copilot are in the world, our kids will want to try those out too (if they haven’t already). But unlike social media, we can get ahead on our knowledge of AI chatbots and generative AI before our kids. Doing so will keep them safer. With an AI chatbot, here are 4 ways parents can pave the way for AI safety.
1. Learn as much as you can about AI and chatbots.
LLMs like ChatGPT, Gemini, and Copilot have been eating up everything on the internet for years. This means they’ve been trained on a variety of sources. So, while they’ve digested everything from Shakespeare and Maya Angelou to various news sources, they’ve also been learning from Reddit posts and tweets on X. Because they’ve been trained on content you and I might also have created, an AI chatbot can get things wrong or “hallucinate.” But it can also sound very human.
It’s also important to know for AI safety that anything you put into a chat can be reviewed or used by the companies that built the chatbots. Companies like OpenAI (ChatCPT), Google (Gemini), and Microsoft (Copilot) use our input to better train an AI chatbot and improve its technology. So, obviously, don’t divulge your social security number. And be sure to talk with your kids about sharing personal information with a chatbot too.
2. Download and get familiar with a chatbot yourself.
It’s like interviewing a very smart friend before introducing him to your kid. You want to make sure you know each other a little bit before welcoming him into your home. And it’s not hard to get started. Once you’ve signed up with Gemini, for example, it’s easy to start asking it questions. It almost feels like you’re texting with another person. But keep in mind that AI is not a person and “There is no evidence this technology is sentient or conscious” writes technology reporter Cade Metz. It doesn’t have feelings or emotions and isn’t any more humanlike than a rock. It’s just been trained to put conversational sentences together. This is something your kids should know for AI safety.
You can ask a chatbot to “write song lyrics in the style of K-Pop” or help you come up with a “strength-training plan tailored for runners.” On Copilot, using Chat GPT-4, you can also “create an image” using OpenAI’s DALL-E technology. Getting familiar with everything these powerful AI tools can do can help prepare you for what your kids will likely want to do themselves when they get the chance.
3. Share what you know about generative AI with your kids.
“Have you heard about ChatGPT at school?” I asked my son, pulling out my phone. He nodded. When I asked him to explain, he said it was “like a robot on your phone that can answer questions.” Great start. I wanted him to know it was more than just a search engine, so I did my best to explain generative AI. “It uses the internet to find answers, but it creates new stuff too,” I began. Together, we spent some time asking ChatGPT questions. We found out it doesn’t have a favorite MLB player but could tell us about certain players and their chances of making the Baseball Hall of Fame. My son thought that was cool.
If you want to establish AI safety and responsibility in your home, start talking to your kids about AI now. Be open (and age-appropriate) with what you know. Our kids are going to grow up alongside AI whether we like it or not. Talking about AI with your kids while they’re young will help establish trust around tech as they grow.
4. Take AI chatbots and their power seriously.
As Uncle Ben told Peter Parker in Spider-Man, “With great power, comes great responsibility.” It’s our responsibility as dads to educate our kids about the power of generative AI before they have free access to it at school, with friends, or on their own devices. As you play around with an AI chatbot, you’ll find they’re fun, but also addicting. Show your kids the neat tricks you can do with them but start setting ground rules early.
Downloading the big companies’ chatbots like ChatGPT, Gemini, and Copilot are a good way to get started. Other big-names with free options for the public include Claude, Mistral, and Perplexity. Because tech is changing so fast, there will continue to be new developments weekly. Let’s stay informed so we can teach kids how to use tech in a safe and responsible way.
Sound off: What are your main concerns about living in a world with AI chatbots?
Huddle up with your kids and ask, “What question would you ask a chatbot?”