If LLMs Can Code, Why Are Apps Still Broken?

This video explores the paradox of modern software development in the age of AI. Despite advances in Large Language Models (LLMs) and AI-assisted coding tools, apps and software still face major challenges. The discussion covers how technical limitations, human factors, and outdated processes contribute to these issues. Key topics include the role of tools like Windsurf and AI-enhanced IDEs, how developers integrate these technologies, and the social dynamics shaping adoption. The video also looks ahead to the future of coding, where integration and automation could redefine software development.

LM Studio: lmstudio.ai/
Windsurf: codeium.com/windsurf
Bolt: bolt.new/
Cursor: www.cursor.com/
VS Code: code.visualstudio.com/
JetBrains (Intelli/J, Rider, etc): www.jetbrains.com/
Obsidian: obsidian.md/
Obsidian Smart Composer: github.com/glowingjade/obsidian-smart-composer
That 64GB Mac Studio Your Boss Should Buy You (affiliate): amzn.to/4gk9GjL

Gear (affiliate)
Studio camera body, Sony Alpha ZV-E10: amzn.to/4gk9GjL
Studio light[s] Elgato Key Light: amzn.to/404pVuH
Studio lens Sigma 16mm f/1.4 DC DN : amzn.to/3VNWWd3
Studio microphone Shure MV7: amzn.to/41G6VoE

00:47 - Technical Challenges of AI for Software Dev
01:40 - From Copy and Paste to Accept Changes
02:24 - Advanced Multifile Editing and Running
03:58 - Too Much To Keep Up
04:41 - Remind AI to be professional
05:56 - Better With Older Technologies
06:34 - Homework Assignment
07:39 - Enterprise Challenges
08:44 - Not The Same as a Year Ago
09:13 - Training Your Replacement?
09:45 - Pair programming?
10:13 - Just a Dev Manager Now?
10:43 - First Rule of AI Club...
11:27 - Human liability shield?
12:30 - Where in the process?
13:07 - Wait for the next one?
13:44 - It's only $6k...
14:02 - What does it mean to be a dev?