How to use chatbots, like ChatGPT, in your daily life and work

In today’s newsletter, the fourth in our five-part seriesI’ll try to convince you that large language models are already good for a wide variety of tasks — and they’re getting smarter every day.

In the previous edition of this newsletter, my colleague Cade Metz wrote that AI large language models (LLMs) can be erratic and unreliable, giving false information and acting strangely towards users.

I’ve been using AI tools like ChatGPT almost daily for several months now, and I’ve seen them spit out a lot of wrong answers. (And don’t remind me of my scary encounter with Bing.)

But I’ve also seen these AI programs do amazing things – feats of creativity, flexibility, and efficiency that took my breath away. And I fear that, in an attempt to drown out the AI ​​hype, skeptics are missing out on what is so revolutionary – and potentially disruptive – about this technology.

Great language role models write poems and screenplays. One of the latest, GPT-4, scored in the 90th percentile on the bar exam and had top scores on a number of advanced placement tests, including AP Biology and AP Art History. And the potential productivity gains for workers are enormous. (In one study, programmers who used GitHub Copilot — an LLM for coders — completed a task 56% faster than programmers who didn’t.)

Here are six things the AI ​​is particularly good at right now:

Explain concepts at multiple levels of difficulty

I’ve been using ChatGPT as my personal tutor for a few months now whenever I’m curious about a new topic. If it’s, say, fractional reserve banking, I’ll start by asking ChatGPT to “explain fractional reserve banking at the high school level.” (Or, if I’m still confused, “college level.”)

ChatGPT and other AI chatbots work well with these types of prompts, using their skills to natural language processing, as long as you don’t ask about very recent events or extremely obscure topics. Having an endlessly patient, no-cost tutor to guide me through new topics feels like a superpower. Used correctly, they can be amazing teaching tools.

Writing and constructive criticism

I always write my columns (including this newsletter) myself. But in the last few months, I hired ChatGPT as my assistant.

When I get stuck, I often paste in a few sentences and see if that can spark any ideas. If I try to narrow an argument, I will ask him to poke holes in my reasoning.

I’ve found Bing to be particularly good for editing because it can search for things on the internet. (One of my favorite writing hacks is to ask her to read Strunk and White’s “The Elements of Style” and suggest edits based on the book’s principles.)

You can use similar tricks to use AI to improve your emails, cover letters, or whatever writing you’re working on. As a reminder, you should always carefully check the work of the AI.

Break out creatively

AI can also be a good tool to boost your creativity. Recently I was trying to come up with questions to ask a podcast guest. I pasted the guest’s bio into ChatGPT and asked him to give me “10 thoughtful and incisive interview questions” for that person. Of the questions he generated, most were pretty good, and I ended up using a few on the show.

Ethan Mollick, a professor at the Wharton School at the University of Pennsylvania, recommends using AI to overcome writer’s block or get started on difficult projects. A person who wanted to start a bakery, for example, could ask ChatGPT for a list of all the necessary equipment and permits.

“It’s very easy to get distracted from a task by a difficult challenge,” Mollick wrote in a recent blog post. “AI provides a way to give you momentum.”

Repetition for real-world tasks

I also use ChatGPT and other AI apps as a kind of rehearsal for offline tasks that I find unpleasant or difficult.

When I had a difficult conversation with a friend, I asked ChatGPT to participate in a role play. “Pretend to be my friend and react as you think my friend might react,” I told him. I then held a fictionalized version of the conversation. It didn’t make real conversation painless, but I felt more confident knowing I had practiced with a chatbot before.

Of course, AI chatbots cannot replace human friendships. But they can be a sort of on-demand sounding board, offering us basic non-judgmental feedback and advice.

SparkNotes for Everything

One of the most powerful capabilities of AI language models is to quickly summarize large amounts of text.

I’ve used AI programs to “read” long articles for me and summarize dense academic papers. Usually they get the broad strokes, even if some details are missing.

This power really shines when you combine it with other AI abilities, like transcribing audio and video recordings. Suddenly, it becomes possible to condense a two-hour podcast into bullet points, or to obtain a concise summary of a work meeting while you’re still in a meeting.

If you get tired of reading this newsletter, you can ask ChatGPT or Bing to summarize the rest for you, or rewrite it as a flippant limerick. I promise I won’t be offended.

Code, even if you don’t code

People are already using AI to write code, whether they know programming languages ​​or not.

These models were not designed for writing software. But when trained on large amounts of Internet text, including content from coding sites like GitHub and Stack Overflow, they learned to code – a phenomenon known as emergent behavior. Now if you type a prompt like “Make me a Chrome extension that translates text from any website into pirate language” you might grab the code for an app that really works.

Over the past few months, social media has been filled with examples of surprised people that use AI tools to build websites and apps faster than any human could – at a tiny fraction of the cost.

And after?

All of this raises some pretty obvious questions about the future, such as: if LLMs help non-writers write well and non-coders code well, what other kinds of powers will they unlock? And how many jobs will these new AI capabilities eliminate?

Tomorrow, Cade will be back to wrap up this series by looking at where all this progress might be headed.

Your homework

What other chatbot skills can you discover?

Think of a task that takes up a lot of your brain every day, like a tedious chore at work or making a weekly dinner plan to feed your family. Then work with a chatbot to see how well it can work.

(Last week, The Times spoke with people who use ChatGPT to have more productive conversations with their teens, cook up imaginative bedtime stories, and write love letters to their spouses. And Hard Fork, the podcast from The Times which I co-host, featured stories of listeners using AI to negotiate their rent, understand medical outcomes, and affirm their gender identity.)

As always, share your best, worst, and most interesting results in the comments.

Quiz

Question 1 of 3

What has Kevin’s experience been using AI tools like ChatGPT?

Start the quiz by choosing your answer.

Glossary

Natural language processing: Techniques used by large language models to understand and generate human language, including text classification and sentiment analysis. These methods often use a combination of machine learning algorithms, statistical models, and linguistic rules.

Emergent behavior: Unexpected or unintended capabilities in a large language model, enabled by models and the model’s learning rules from its training data. For example, LLMs trained at programming and coding sites can write new code. Other examples include creative abilities such as composing poetry, music, and fictional stories.

Click here for more glossary terms.