Write something
Pinned
🔥
2d •
📊 Nutrition Tracker Part 2: Teaching ChatGPT Context Using Project Instructions
In Part 1, we started with a simple experiment. I gave ChatGPT a basic food and activity log and asked: How did I do today? The answer was generic. Why? Because ChatGPT did not know enough about me. It did not know my: - Goal - BMR - Daily calorie target - Protein target It did not know what “good” meant for me. That was the first real lesson: - A generic chat produces generic answers. - The next step was creating better context memory using Projects. Step 1: Create a Dedicated Project Instead of starting a random new chat every day, I created a dedicated ChatGPT Project for nutrition and activity tracking. Why does this matter? Because a regular chat is just a conversation. A Project is more like a focused workspace. It gives ChatGPT a place to keep the work organized around a specific purpose. For this project, the purpose was simple: 🍽️ Track food 🚶 Track activity 🔥 Estimate calorie burn 📊 Summarize the day 🎯 Help me understand whether I am moving toward my goal To create the Project: 1. Click Projects 2. Click New Project 3. Name it Nutrition and Activity Tracker Step 2: Add Basic Project Instructions Next, I gave ChatGPT a few simple Project Instructions. Nothing complicated. Just enough context to make the answers more useful. Something like: You are my nutrition and activity tracking assistant.My goal is to lose body fat while maintaining muscle.My estimated BMR is 1,800 calories per day. When I enter food, estimate calories and macros.When I enter activity, estimate calories burned.When I ask for a summary, show food calories, protein, carbs, fat, activity burn, net calories, and whether the day is trending toward a calorie deficit. Keep summaries concise. That small set of Project Instructions changed the quality of the output because ChatGPT now had context. To add Project Instructions: 1. Open the Project 2. Click the three dots 3. Select Project settings 4. Add your instructions The Same Food Log Became More Useful
Pinned
🔥
7d •
Nutrition Tracker Part 1: Understanding LLM Generic Answers
Series: Learn ChatGPT by Building a Nutrition and Activity Tracker In the introduction to this series, I shared how I started using ChatGPT to build a personalized nutrition and activity tracking system. Today, it tracks food, activity, workouts, steps, estimated calorie burn, bodyweight, and trends. It also helps generate dashboards, calculate custom metrics, analyze patterns, and provide feedback based on my goals. But it did not start there. It started with a simple food log. The Experiment: I opened a new ChatGPT conversation and entered something like, today I ate, and entered my food. Breakfast: - 2 scrambled eggs - 2 slices wheat toast Lunch: - Turkey sandwich - Small apple Dinner: - Grilled chicken breast - 1 cup white rice - Mixed vegetables Snacks: - Protein bar - Handful of potato chips Activity: - 7,500 steps - 30-minute walk Try it. Simply copy and paste the list above into a standard chat, and follow with the prompt, "How did I do today?" In my case, the response was reasonable. ChatGPT told me I had a fairly balanced day, included some protein, got in some movement, and made some generally healthy choices. That was fine. But it was also generic. So I asked a follow-up question. Was I in a calorie deficit? And ChatGPT essentially said, I do not have enough information to know. The model did not fail. It simply did not have enough information. At that point, ChatGPT had no context, and therefore did not know: - My weight - My height - My goals - My calorie target - My protein target - My BMR So the response was very general. It could comment on the food. It could estimate calories. It could provide broad health advice. But it could not tell me whether the day was successful against my actual goals. A generic prompt produces a generic answer. A personalized system produces a contextual answer. At this stage, I was NOT using: - Projects - Project Instructions - Memory I was simply chatting with ChatGPT.
🔥
4d •
Quick Quip Part 1: Why Humanoid Robots Feel Like Form Over Function
Okay, I am going to say something I may be completely alone in thinking. I keep coming back to the same question. Why are we so obsessed with making robots look like us? I understand the argument. The world was built for humans. Doors. Stairs. Tools. Handles. Factories. Warehouses. Kitchens. Vehicles. So the logic makes sense on the surface. Build a robot shaped like a human and it can operate inside the world humans already built. Fair point. But then I look at a humanoid robot pushing a lawn mower (illustrative) through tall grass, and I cannot help but ask: Is that really the best solution? A purpose-built mowing machine seems far more logical than a human-shaped machine using a human-designed tool to perform a machine-friendly job. A humanoid robot mowing the lawn looks futuristic. But the boring machine built specifically to cut grass may actually be the better answer. And that just makes me pause and shake my head. Because this is where humanoid robots start to feel more like form over function. They look impressive. They photograph well. They feel like the future we were promised in movies, cartoons, and science fiction. But impressive is not the same as useful. And familiar is not the same as optimal. Maybe humanoid robots are necessary because we are trying to automate environments that were designed around people. Maybe they are a bridge technology. A way to bring robotics into homes, businesses, factories, and job sites without rebuilding the world around the robot. That is possible. But I still wonder if we are making a deeper mistake. Maybe we are not just designing robots to solve problems. Maybe we are designing robots in our own image because we still struggle to imagine intelligence, labor, and usefulness without putting ourselves at the center. That is the tension. We say we want machines to do the work better. But then we keep making the machine look like the worker. That is a very human thing to do. And maybe that is exactly the problem.
3d •
Anyone with the $20 per month Claude plan?
What are you uses If you have claude pro the $20 plan? Have you ever exhausted tokens on the pro plan? I was told the freemium claude plan is barly functional.
1-30 of 432
skool.com/ai-bits-and-pieces
AI lessons you can read in under 3 minutes and apply in everyday work and life.
Leaderboard (30-day)
1
+100
2
+44
3
+35
4
+32
5
+21
Powered by