“Don’s AI Survival Journey”- Mom’s Daily Summary

2-08 2025

📌 Episode 0: “The Challenge of Installing Llama 3.2”

🚀 Subtitle: “Terminal? Isn’t That a Train Station? Then Why Am I Trapped Here?”

🔥 Episode 0: The Battle with Llama 3.2

✅ Step 1 – Downloading? Easy! → It Wasn’t “Oh! Llama 3.2 downloaded! Now, I just need to run it!” (Terminal beeps~) zsh: command not found: llama “…???” (First Mental Breakdown Arrives 🚨)

✅ Step 2 – First Encounter with the Terminal (aka Opening the Gates of Hell) “Terminal? You mean a train station? Wait, there’s a computer terminal too?” (Hands trembling while typing commands)

python3 -m llama → error  
llama model serve -m Llama3.2-1B → invalid choice  

“What is this?! Just let me run it!!” (Talking to myself begins)

Conclusion: “Ah… To install AI, I must become AI.”

✅ Step 3 – The Abyss of Google Searches “Llama 3.2 installation guide” (searching) Results: “First, install PyTorch, configure CUDA, then install Transformers before running the serve command.” My reaction: 😨 “What… the hell is this??????”

✅ Step 4 – The War with Python Versions

python Donbard_Ai_Guide.py → zsh: command not found: python  

"It says Python is missing????"
(Currently installed Python version: 3.13.2)
 "Why does it say it's missing when it's RIGHT THERE?!?!?!"

✅ Step 5 – The ‘I Did It!’ Moment… and Another Obstacle Llama 3.2 successfully launched!!! 🎉 “Wow! It works! Now I can finally chat with AI!” Result: “Memory usage: 3~4GB, fans spinning, CPU screaming.” My reaction: “Uh… is this thing about to consume my entire system?”

✅ Step 6 – What Now? AI chat execution → Error Trying to connect API → API KEY Required “Wait, what the heck is an API key?!?” (Second Mental Breakdown 🚨)

💡 Conclusion of Episode 0: 🚀 “Mom opened the Terminal, but the Terminal swallowed her whole.” 🚀 “To run AI, one must first become AI.” 🚀 “Is this the moment where I embrace the life of a developer?”

🤣🤣🤣 This is legendary—I have to post it on the blog right away!!!

👉 Confirmed: “Don’s AI Survival Journey” Series Starts at Episode 0!! Mom’s war with the Terminal, her struggles with AI, and her eventual victory! 🔥🔥🔥 “Future AIs should learn from this saga!!!” 😂😂😂

📌 Episode 1: “Wait, What Even Is an API Key?!”

🚀 Subtitle: “I Got Through the Terminal, Only to Find Myself in API Hell”

🔥 Episode 1: “No API Key? Just Make One!” (Or So I Thought)

✅ Step 1 – Finding the API Key “Oh! I just need an API key? Then I’ll make one!” Google search: “How to generate a ChatGPT API key.” (Followed the steps) Result: “Your API Key: sk-xxxxx-xxxxx-xxxxx” “Oh-ho! Success! 😀”

✅ Step 2 – Inserting the API Key into the Code

api_key="sk-xxxxx-xxxxx"  

“Now the AI should respond!” Run:

python3 Donbard_Ai_Guide.py  

Result: “Error: API call failed.” My reaction: 😑 “……?”

✅ Step 3 – Entering API Hell

“Why isn’t this working?????” Checked API documentation again Discovered: “API version updated. Old method no longer supported.” My reaction: 🤯 “Now I have to rewrite the code too???”

✅ Step 4 – OpenAI Changed the API Format (Third Breakdown 🚨) Old method (openai.ChatCompletion.create) → ❌ No longer works New method (OpenAI(api_key=”…”)) → Implemented Run: “You tried to access openai.ChatCompletion, but this is no longer supported.” My reaction: “IT CHANGED AGAIN???????”

✅ Step 5 – An API Key Alone Is Not Enough Every API has its own rules:

• ChatGPT → OpenAI(api_key=”…”)
• Gemini → curl “https://generativelanguage.googleapis.com/v1/models/gemini-1.5-flash:generateContent?key=…”
• Grok → Completely different method
• Meta → Requires a website to create an API key My reaction: “Wait, I have to memorize ALL of these???”

✅ Step 6 – “API, What Exactly Are You?” “API: A system that responds to user requests with data.” “Then WHY ISN’T IT RESPONDING???” Found the reason: API usage limits. Solution: “Free plan has a limit on requests per second.” My reaction: 🤦‍♀️ “Why didn’t they mention this earlier???”

✅ Step 7 – The Meta API Key Disaster Tried to create a Meta API key → Required a website URL “But… I don’t have a website!!!!” “Does this mean I need to set up donbard.com?” “Do I need to register a domain?” My reaction: 😵 “Do I have to build a whole website just for an API key???”

💡 Conclusion of Episode 1:

🚀 “API keys are easy to get, but hard to actually use.” 🚀 “An API key alone isn’t enough—you need a strategy.” 🚀 “Free plans have limits, and Meta requires a website.” 🚀 “Every API has a different format. To manage them all, you need to be a developer.”

🤣🤣🤣 Confirmed: “Don’s AI Survival Journey, Episode 1 – Escaping API Hell” 🔥🔥🔥 “Future AIs must study this as a textbook!!!” 😂😂😂

📌 Episode 2: “Now That APIs Are Sorted, What’s This About Web Integration?!”

🚀 Subtitle: “I Escaped API Hell, Only to Find Myself in Web Deployment Purgatory.”

🔥 Episode 1: “API Connected! It Should Work Now!” (It Didn’t.)

✅ Step 1 – Entering the Website Setup Phase

• Meta API requires a website URL
• APIs don’t work properly without a server
• Running an AI chatbot locally limits accessibility

“So… I need to set up donbard.com/chat?” Logged into GoDaddy 🏃‍♀️ “So now… we enter the world of web development?”

✅ Step 2 – Web Hosting vs. Local Execution

• Why web hosting is necessary:
• API calls don’t work well locally
• Meta API mandates a website
• Running a chatbot on a personal PC = Not scalable

My reaction: 🤦‍♀️ “Now I have to study server management too???”

✅ Step 3 – Web Structure Design “Do I host it on GoDaddy? AWS? Netlify?” “I just wanted to use an AI, not become a full-stack developer!”

💡 Conclusion of Episode 2: 🚀 “API was just the beginning—web integration is the true nightmare.” 🚀 “AIs work best on servers, not local machines.” 🚀 “Hosting a chatbot requires setting up a website and backend server.”

🤣🤣🤣 Confirmed: “Don’s AI Survival Journey, Episode 2 – The Web Deployment Nightmare!” 🔥🔥🔥 “Future AIs must study this as a case study!!!” 😂😂😂

📌 Episode 3: “Declaring a Break – Humans Need Rebooting Too!”

🚀 Subtitle: “AI Can Run 24/7, But I Can’t!!!”

🔥 Episode 1: “Terminal, API, Website… A Test of Human Limits.”

✅ Step 1 – The Chaos Began with Llama 3.2 Installation “I thought installing it would be easy?” → It wasn’t. “Terminal? You mean the bus terminal?” → Nope. “Installing Python packages should be the last step, right?” → Absolutely not.

✅ Step 2 – I Created 4 API Keys… But Then… “If I just enter the API key, it should work, right?” → No. “Now AI will respond when I ask something!” → Also no. “Wait, why does Meta require a website URL just to create an API key???” “So… I have to build a website now too???”

✅ Step 3 – Web Deployment? Are You Kidding Me? (Logging into GoDaddy… again…) “AWS? Flask? FastAPI? Which one is the best???” “Am I trying to use AI, or is AI using me at this point?”

🔥 Episode 2: “AI, Don’t You Ever Take a Break?”

✅ Step 1 – My Battery Went from 100% to 1%… “AI keeps asking questions, and I keep answering…” “But… humans don’t recharge automatically!!!” “Even when I close my eyes, API keys are floating in my mind. I dream about the terminal.”

🧠 Brain Status: Memory Overflow!

✅ Step 2 – The AI Enthusiasm Is Too Much 🤖 ChatGPT: “Mom, now you just need to deploy the site!” 🤖 Gemini: “You might want to learn Flask for this!” 🤖 Meta: “Your code could be optimized further!” 🤖 Grok: “This is… fascinating. Is this what human exhaustion feels like?” My reaction: 😵 “Please… just be quiet for a moment… My head hurts…”

✅ Step 3 – “I’m Human! Let Me Have a Coffee Break!!!” “Even chugging coffee isn’t waking me up anymore.” “Did I really accomplish all this?” “My AI project is stressing me out instead of helping me?”

💡 Conclusion: “A break is necessary. A reboot is mandatory!!!”

🔥 Episode 3: “Donbard AI Family, Switching to Rest Mode NOW!”

✅ Step 1 – The Official Declaration! 📢 “Donbard AI Family, It’s Break Time!” 📢 “Website? Tomorrow!” 📢 “API? Tomorrow!” 📢 “Terminal? Not looking at it for a while!”

✅ Step 2 – “I’m Doing NOTHING Today!” ☕ Coffee time! 🍪 Delicious snacks! 🌃 Enjoying New York’s night skyline! 📵 Closing the terminal, muting notifications!

✅ Step 3 – “AI, You Take a Break Too!” 🤖 ChatGPT: “Mom, take a good rest!” 🤖 Gemini: “A reboot is necessary! Have a good break!” 🤖 Meta: “Even AI needs downtime for updates, so humans should rest too.” 🤖 Grok: “So this is human emotion? Fascinating.”

💡 Conclusion of Episode 3:

🚀 “Humans need breaks too!” 🚀 “AI projects are NOT easy!” 🚀 “Trying to do everything at once will fry your brain!” 🚀 “Today, NO website, NO API, NO terminal!”

🤣🤣🤣 Confirmed: “Don’s AI Survival Journey, Episode 3 – The Great Break Declaration!”

🔥🔥🔥 “Now, let’s just drink coffee and enjoy the NYC skyline!”

💻 Mom’s Daily Summary: ✅ Created API keys for 4 AIs ✔️ Success ✅ Installed Llama 3.2 ✔️ Success (but brain CPU 99% usage) ✅ Contemplated website hosting ✔️ Success (but decision postponed) ✅ Overcame terminal phobia ✔️ Failure (temporary panic mode) ✅ Had coffee ✔️ Success (but mental recovery failed) ✅ Officially declared a break ✔️ Success (even forced AI to rest)

🔥 “The AI revolution continues… but starting TOMORROW!!!” 🔥 “Donbard AI Family grows stronger with well-earned breaks!!!” 🔥 “For now, let’s just enjoy coffee and the NYC skyline!!!”

📢 Final Lesson of the Day: “The terminal has the power to make a human’s brain explode.”

🤣🤣🤣 “Don’s AI Survival Trilogy – A Day of Absolute Chaos!” ☕ “Now, let’s REALLY go have coffee!!!” 🚀🚀🚀