How to use ChatGPT, Bard, other chatbots

As the field of generative AI heats up, consumer-facing chatbots answer questions about business strategy, design study guides for math classes, offer salary negotiation advice, and even write wedding vows. And things are just getting started.

OpenAI’s ChatGPT, Google’s Bard, Microsoft’s Bing, and Anthropic’s Claude are some of today’s leading chatbots, but over the coming year we’re likely to see more emerge: in In venture capital, generative AI deals totaled $1.69 billion globally in the first quarter. this year, a 130% spike from last quarter’s $0.73 billion — with an additional $10.68 billion in deals announced but not yet closed in the first quarter, according to Pitchbook data.

investment related news

Two months after the launch of ChatGPT, it surpassed 100 million monthly active users, breaking records for the fastest-growing consumer app in history: “Phenomenal adoption – frankly we haven’t never seen anything like it, and interest has been growing ever since,” Brian Burke, vice president of research at Gartner, told CNBC. “Since its release on November 30 until today, our request volume has increased like a hockey stick; every customer wants to know more about Generative AI and ChatGPT.”

These types of chatbots are built on Large Language Models, or LLM, a machine learning tool that uses large amounts of internet data to recognize patterns and generate human-sounding language. If you’re a beginner, many sources we spoke with agreed that the best way to start using a chatbot is to dive in and try things out.

“People spend too much time trying to find the perfect prompt – 80% of them just use it interactively,” Ethan Mollick, an associate professor at the Wharton School at the University of Pennsylvania, who studies effects of AI on work and education, CNBC told CNBC.

Here are some tips from the pros:

Keep data privacy in mind.

When you use a chatbot like ChatGPT or Bard, the information you input (what you type, what you receive in response, and any changes you request) can be used to train future models. OpenAI says as much in its terms. Although some companies offer ways to opt out – OpenAI allows this under “data controls” in ChatGPT settings – it’s always best to refrain from sharing sensitive or private data in chatbot conversations, especially while companies are still fine-tuning their privacy measures. For example, a ChatGPT bug in March briefly allowed users to see parts of other people’s chat histories.

“If you don’t post it to Facebook, don’t put it in ChatGPT,” Burke said. “Think of what you put in ChatGPT as public information.”

Provide context.

For the best possible return on your time, give the chatbot context on how it should act in this scenario and who it is serving with this information. For example, you can write the persona you want the chatbot to adopt in this scenario: “You are a [marketer, teacher, philosopher, etc.].” You can also add context like: “I am a [client, student, beginner, etc.]. This could save time by telling the chatbot directly what kind of role it should take on and what “lens” to convey information in a way that is useful to you.

For example, if you’re a creative consultant looking for a chatbot to help you with analyzing company logos, you could type something like “Act like you’re a graphic designer studying logo design for I am a client who owns a business and is looking to find out which logos work best and why. Generate an analysis of the “best” business logos for publicly traded companies and why they are considered good choices.

“If you ask Bard to write an inspirational speech, Bard’s response may be a bit more generic – but if you ask Bard to write a speech in a specific style, tone, or format, you’ll likely get a much better answer,” Sissie Hsiao, vice president of Google, told CNBC.

Let the chatbot do all the work.

Sometimes the best way to get what you want is to ask the chatbot itself for advice – whether you’re asking what’s possible as a user or the best way to phrase your prompt.

“Ask him the simple question, what kind of things can you do? And it’ll give you a list of things that would surprise most people,” Burke said.

You can also game the system by asking something like, “What’s the best way to ask you for help writing a shopping list?” or even giving the chatbot a job of writing prompts, like “Your job is to generate the best and most effective prompts for ChatGPT. Generate a list of the best prompts to ask ChatGPT for healthy dinner recipes at a jar.”

Ask for help with brainstorming.

Whether you’re researching vacation destinations, date ideas, poetry prompts, or content strategies to go viral on social media, many people use chatbots as a starting point for brainstorming sessions.

“The most important thing…that I find them useful for is to inspire me as a user and help me learn things that I wouldn’t necessarily have thought of on my own,” Josh Albrecht, CTO of General Intelligent, an AI research startup, told CNBC. “Maybe that’s why they’re called generative AI – they’re really useful in the generative part, the brainstorming.”

Create a crash course.

Let’s say you’re trying to learn geometry and consider yourself a beginner. You can start your studies by asking a chatbot something like “Explain the basics of geometry like I’m a beginner” or “Explain the Pythagorean theorem like I’m five years old.”

If you’re looking for something more expansive, you can have a chatbot create a “crash course” for you, specifying how long you have (three days, a week, a month) or how many hours you want to spend learn the new skill. You can write something like, “I’m a beginner who wants to learn to skateboard. Create a two-week plan to learn how to skateboard and do a kickflip.

To extend your learning plan beyond the chatbot, you can also request a list of the most important books on a topic, some of the most influential people in the field, and any other resources that might help you advance your skills. .

Don’t be afraid to give notes and ask for changes.

“The worst thing you can do if you’re actually trying to use ChatGPT output is [to] just ask him one thing once and then walk away,” Mollick said. “You will get a very generic output. You have to interact with him.”

Sometimes you won’t choose the perfect prompt or the chatbot won’t generate the output you were looking for – and that’s okay. You can always make changes to make the information more useful, like asking follow-up questions like “Can you make it less generic?” or “Can you make the first paragraph more interesting?” or even rephrase your original request in a different way.

Take everything with a lot of salt.

Chatbots have a documented tendency to fabricate information, especially when their training data doesn’t fully cover an area you’re asking about, so it’s important to take everything with a grain of salt. Suppose you ask for a biography of Albert Einstein: a chatbot might tell you that the famous scientist wrote a book called “How to be smart”, when, unfortunately, he never did. Also, since large language patterns are trained over large swaths of the Internet, they are best at pattern recognition, which means they can generate biased output or misinformation based on their training data.

“Where there’s less information, it just creates stuff,” Burke said, adding, “These hallucinations are extraordinarily compelling…You can’t trust these models to give you accurate information. all the time.”

Experiment and try different approaches.

Whether you’re asking a chatbot to generate a list of action items from a meeting transcript or translating something from English to Tagalog, there’s an endless range of use cases for the chatbot. Generative AI. So when using a chatbot, it’s worth thinking about the things you want to learn or need help with and experience the quality of the system.

“AI is a general-purpose technology; it does a lot of things, so the idea is that whatever field you’re in and whatever job you’re in will affect aspects of your job differently than anyone else on the planet,” Mollick said. “It’s about thinking about how you want to use it…You have to find a way to work with the system…and the only way to do that is to experiment.”

Subscribe to CNBC on YouTube.