29/06/2026 6:25 am
Topic starter
Notifications
Clear all
Agent Skills Forum
3
Posts
4
Users
0
Reactions
154
Views
0
Honestly Im just about ready to throw my laptop out the window. Ive spent the last three months trying to teach myself AI stuff for a small side project I want to launch by the end of the year, but every single tutorial I find uses a different language and I feel like Im just spinning my wheels. I started with some basic Python scripts because that seemed like the standard path, but then my friend who works in tech told me I should be looking at C++ if I want any actual performance or efficiency. Like, are you kidding me? I dont have time to learn an entirely new language from scratch if I dont have to.
Its just super stressful because I have a tiny budget for cloud compute and I cant afford to waste time building something in a language that isnt actually going to get the job done right. Everything I read online contradicts the last thing I read. One minute its all about Python being the king, then someone else says its too slow for production-level AI models and I should be using Julia or even Rust? My brain is melting here. I just need to know what is actually essential. Im not trying to build the next ChatGPT or anything, just a predictive tool for my local small business inventory tracking, but the learning curve feels like a vertical wall right now.
Do I really need to stick with Python for the libraries, or am I setting myself up for failure later on if I dont learn something more robust? Is there one language that is non-negotiable for AI work or am I just overcomplicating this entire thing because of all the noise online? If you were in my shoes, struggling with a tight timeline and just wanting to get a prototype working without needing a PhD in computer science, what would you focus on learning first...
3 Answers
12
29/06/2026 6:55 am
To add to the point above: honestly stop overthinking it. If you want something that works fast without the headache, just stick to Python.
- Python: Massive ecosystem, libraries like PyTorch AI Library 2.0 make life easy for beginners.
- Rust: Technically faster but the learning curve is brutal.
- Julia: Cool syntax but too niche for quick prototyping. Don't waste cash on expensive hardware yet; a simple Lenovo ThinkPad P16 Gen 2 i7 32GB RAM is plenty for your prototype.
11
29/06/2026 6:40 am
Look, I feel your pain. I spent way too many hours trying to optimize inventory models in C++ early on and honestly, it was a nightmare. Unfortunately, the documentation for niche libraries is often outdated or just broken, and I had way more issues than I expected trying to get things to compile. It really isnt as good as the hype makes it seem unless you are building something massive from the ground up. Before you ditch the Python route, what kind of data are you pulling for this inventory tracker? Like, is it simple CSV spreadsheets or are you scraping real-time sales APIs? I ask because if you are processing huge datasets, you might just need a better rig to handle the local training if your cloud budget is tight. Have you looked at something like a Lenovo ThinkStation P360 Tower i9-12900 64GB RAM to keep things local and fast?
2
29/06/2026 6:25 am
Honestly, stick with Python. Don't listen to the C++ hype right now. It's overkill for an inventory tool. Youre gonna burn out trying to manage memory. Just grab a solid Dell Precision 7680 Intel Core i9 64GB RAM and focus on libraries like PyTorch. If performance starts lagging later, you can always rewrite just the heavy math bits in C++ then. Stop overthinking the stack and just ship it.
