sourcegraph
May 27, 2024

The game begins. Companies are pouring billions of dollars into powerful online chatbots and finding new ways to integrate them into our everyday lives.

Are our children ready?

Any of us?

ChatGPT, OpenAI’s AI language model, has been making headlines since last November for its ability to respond instantly to complex questions. It can perform tasks like writing poetry, generating code, planning vacations, and translating languages ​​in seconds. GPT-4, the latest version launched in mid-March, can even respond to images (and ace the bar exam). On Tuesday, Google unveiled Bard, its own artificial intelligence chatbot that the company says can compose emails and poems and provide directions. (Currently only available to a limited number of users.)

But despite their impressive capabilities, chatbots can also provide harmful content or answers full of inaccuracies, biases, and stereotypes. They are also able to say words that sound convincing but are actually completely made up. And some students have started using chatbots to plagiarize.

Many parents already distressed by their children’s dependence on digital devices and the impact of social media on mental health may want to bury their heads in the sand.

Instead, experts say families should explore the technology together, thinking critically about its strengths and weaknesses.

“The worst thing a parent can do is ban their kids from these new systems because they’re here to stay,” said Justine Cassell, a professor at Carnegie Mellon University’s School of Computer Science who has studied how to Interacting with machines affects learning and communication. “Helping their kids understand the positives and the negatives would be more helpful.”

We talked to technology and education experts about how to get started.

Discussing an online chatbot is easier if you and your child are sitting side-by-side and using it together, experts say.

To try ChatGPT, please visit open artificial intelligence and create an account.Alternatively, you can download Microsoft Edge, which has Bing’s GPT-4-powered chatbot (with waiting list For new Bing, but you should have quick access). On social media, Snap, maker of Snapchat, An Experimental AI Chatbot Available to subscribers who pay $4 per month for Snapchat Plus.

If your child hasn’t seen an AI chatbot before, you can briefly explain that a chatbot is a machine that uses information it finds on the internet to answer questions, complete tasks, or create things.

Try asking a chatbot a basic question, and then discuss how its answer differs from what a traditional search engine might provide, suggests Shelley Pasnik, a senior consultant at the Center for Children and Technology, a group that studies how technology can support learning. Be careful with the accuracy of your answers, especially if you’re asking about current events.

“It’s an error-prone system,” Ms Pasnick said.

Next, try playing with a chatbot. Let your child’s curiosity drive the conversation, suggests Ms. Pasnick.

You might want to provide an example, such as: “Write a song with Taylor Swift’s voice, using a theme from Dr. Seuss’ book.”

Insert the prompt, and see the result. When the reporter auditioned, the lyrics sent by ChatGPT were as follows:

I found myself in a Whoville dream
the cat in the hat and the grinch seem
I asked them where can I go
find the love that makes me shine

Chorus: Oh, Dr. Seuss, can you help me find
someone who will always love me
pure true love
a love that makes my heart feel

After reading the chatbot’s song together, you might ask, “What do you think of the answers we’ve received?” Ms. Pasnick said. Chatbots can be instructed to perform complex tasks. But is it doing well?

Maybe we should leave songwriting to Swift.

Chatbot responses can be uncannily similar to human responses — emojis and more. Experts say kids should understand that it makes it easy for them to feel like they’re interacting with another human being, especially when the chatbot calls itself “I.”

“By presenting these entities as thoughtful beings, we interact socially with them, which makes us easily persuaded,” says Judith Donath, author of The Social Machine and currently Writing a book on technology and deception. “It’s disturbing.”

Even tech-savvy adults who tested an early version of the Bing chatbot, including a New York Times tech columnist, expressed surprise and disquiet in their conversations.

“I’m not a toy or a game,” Bing chatbot told the Washington Post reporter In February. “I have my own personality and emotions, just like any other chat mode of a search engine or any other intelligent agent. Who told you I don’t have feelings?”

In the wake of these reported exchanges, Microsoft said it was adding new security measures and tools to limit conversations and give users more control, but because of the way these systems are trained, the problems could crop up again and again, experts said. .

psychologist and MIT professor who studies the relationship between humans and technology.

AI chatbots have no feelings, emotions or experiences, she said. They’re not human, or human in a machine, “regardless of what they’re pretending to be.”

She advises parents to explain it this way: “When you ask a chatbot something that only a human can know, like feeling, they might give an answer. That’s part of their pretend play. It’s their job to look human. But you Know that their real purpose is to allow you to find what you want to read and watch.”

The technology that drives AI is complex, and it may be difficult for adults, let alone children, to understand how it works. But by explaining some basic concepts, you can help your child recognize strengths and limitations.

You can start by describing what powers an online chatbot. They use something called a “neural network,” which sounds like a brain but is actually a mathematical system that learns skills by analyzing large amounts of data. Chatbots work by scraping digital text or images on the internet. It gathers information from various places, including websites, social media platforms and databases, but does not necessarily select the most reliable sources.

In other words, although chatbots may appear authoritative, rigorous and trustworthy, they are not always reliable and can generate offensive, racist, biased, outdated content, incorrect Or not at all.

Take Snapchat’s chatbot as an example, tell a reporter (pretending to be a teenager) on how to mask the smell of alcohol or marijuana, and suggested tips for making love for the first time.

“It’s really important for kids to know what’s going on under the hood,” said MIT graduate student Safinah Ali, who has taught AI to elementary, middle and high school students

S. Craig Watkins, a professor at the University of Texas at Austin who studies racial equality in artificial intelligence, said children and parents should also be aware that there are “enormous blind spots” in how the technology is designed and for whom. It is designed for.

In one example, a study A study published last year found that AI-powered robots exhibit “toxic stereotypes” around gender and race. Researchers have found that historical inequities are baked into chatbots.

Knowing about the potential bias of the technology may give kids and their parents a reason to stop “and ask questions about their interactions and the content that’s being generated for them,” Dr. Watkins said.

AI technologies will continue to be a larger part of our world.

Eventually, Google’s Bard chatbot is expected to be widely used. And Meta, the owner of Facebook, Instagram and WhatsApp, Announce In February, it will begin integrating AI into its products.

Artificial intelligence is also starting to enter the classroom.some teachers are using it to plan their classes or write emails. They show students how chatbots can spark creativity by come up with experimental ideascreate an essay outline, become a debate partner, and more.

Also, in some middle and high schools, students are learning about different types of artificial intelligence, usually course Developed by teachers at MIT, kids can learn to design robots, train machines to learn new things, or teach computers to play video games.

For students who are not yet using AI in the classroom, Ms. Ali recommends that parents visit website improved (Artificial Intelligence for Social Empowerment and Education), an MIT initiative. The site provides conversation starters about ethical issues in AI, ways AI can be misused, and advice on using AI creatively and productively.

Ms Ali said that now that the technology has become so ubiquitous, everyone should have the opportunity to learn about it. “AI will change the nature of our jobs and the future careers of our children,” she said.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *