In all likelihood, 2024 is going to be the year of AI. With chatbots dominating the headlines in 2023 and alternatives like Microsoft’s AI-powered Bing and Anthropic’s Claude popping up everywhere, it’s safe to say that AI — especially AI chatbots and generative AI. — will be front and center this year.
If you’re hesitant about AI, you’re not alone. In 2023, the Pew Research Center found that 52% of Americans were excited about AI becoming a part of their daily lives.
Speaking from my own experience, I was incredibly hesitant to build a toy with AI at the beginning of last year. However, once I did, I felt a bit more comfortable using AI chatbots.
One thing that helped me become more comfortable with generative AI — which creates text and images by drawing on vast amounts of data — was making sure I asked it well-formed questions. , or asking for hints. Like learning how to ask the right Google search questions to get usable results, “prompt engineering” is a skill — and an art — of creating detailed and focused prompts to actually get generative AI models. Do what you want them to do.
When Google first came out, you didn’t become an expert at crafting effective search queries overnight, and the same would be true of quick engineering. Creating productive cues for creative AI will take time to master. If you want to get started with instant engineering and learn how to use AIs in your life a little more effectively, you’re in the right place.
I’ve tested the AI ​​use case on a couple of different AI generative models to show how to get you started on the path to becoming a competitive prompt engineer, no matter which AI model you’re using. are For more, here’s what to know about AI on your phones and how Adobe is thinking about AI.
Common questions won’t cut it enough.
The first thing to know is that writing short, broad tips won’t get you the results you want.
If you ask an AI “How can I become fluent in German,” for example, you’ll get a much less focused and efficient answer than if you ask the AI ​​”I’m a college-educated adult.” I am an English major, moderately fluent in Italian, and have taken one semester of college-level German. I practice my German vocabulary and grammar for about 15 minutes every day. I How can I become fluent in German?”
To help me become a better AI prompter, I decided to feed the generative AIs information about something I know well, so I can predict what my prompts will be. How clarification and clarification resulted in better answers.
A popular prompt — and as a runner it’s one I’m particularly interested in — is to get help from generative AI to create a training plan for a marathon or half marathon.
To get started, OpenAI suggests asking ChatGPT “Help me train for a half marathon.” As I expected, this hint makes sense as a starting place, but it won’t yield specific results. For example, ChatGPT gave me general tips for training such as setting goals and getting proper nutrition and hydration. Note: For this test, I used Red Ventures’ (CNET’s parent company) version of ChatGPT. And I compared the results to Anthropic’s Claude, who took the same general approach — no specific training program.
After using the basic hint OpenAI gave me some basic advice to run.
While these aren’t bad places to start, they won’t get me to the finish line of a marathon, which will require a specific program based on my fitness and experience, how long I have to train, I typically How I like to train and even when and where I will run my races.
Limiting your prompts can make a big difference, depending on which model you’re using.
For help with a quick fix, I turned to CNET’s sister site ZDNET for help. ZDNET suggests adding more details to help guide creative AI.
I tried ZDNET’s advice next: “I’m a beginner runner and have never run a marathon before, but I want to complete one in six months. How can I prepare for a marathon?” In this example, I’m giving the AI ​​some more context to work with, such as a timeframe and experience level for training.
I think that’s a little more helpful.
Unfortunately, with ChatGPT, the reaction wasn’t that different. ChatGPT gave me a bit more specific advice, like focusing on completing a marathon instead of setting a personal record and joining a running community for motivation while training. Impressive, maybe, but still missing the mark.
When I gave Claude the same prompt, I got something else. He gave me a pretty bare bones training program, suggested monthly training goals and even possible exercises I could incorporate into my training. Close to, but not yet fully developed training plan.
The more details, the better
It was time to take my prompts to the next level. In my next shot at an extended tip, I included lots of details like: race and training date, age, gender, road conditions and terrain for the marathon and training.
Now we are going somewhere. Now booking London flight.
Success! What I got back was a good jumping off point for a real program that I could see myself using (with some tweaks of course). ChatGPT recommended a pretty detailed program breaking down specific training goals and the pace I could incorporate into new workouts by month. Claude broke the training plan down into weeks, giving me specific workouts I could focus on each week, mileage goals and how often I should run each week.
I would just note that these programs weren’t quite right for me yet and how I would train, but they were close and I would have a strong starting point for my training. The proof is in the prompt — the more details I added, the more details I got back from ChatGPT and Claude.
And one last tip: Whether you’re looking for a marathon training program or help preparing a meal, it’s always best practice to critique an AI recommendation and do your due diligence — like before you start. Check with a doctor. A rigorous exercise routine – before taking AI advice as fact.
Also see: What to know about photography in the age of AI and Biggest AI trends in cybersecurity
Editors’ note: CNET is using an AI engine to help generate some stories. For more, see This post.