This channel is not for vibe-coders. It's for professional devs who want to use AI as powerful assistant, while still keeping the control of their codebase. My name is Povilas Korop, and I'm passionate about coding with AI. So I started this THIRD YouTube channel, in addition to my other ones Laravel Daily and Filament Daily. You will see a lot of my experiments with AI: I will try new things and share my discoveries along the way.
I tested the popular development trend of using frontier LLMs as managers to delegate tasks to cheaper subagents. The experiment compared speed, token cost, and final output against a direct, single-model approach. I found one specific workflow that actually doubled build times and drained resources, proving that delegation is not always the smartest choice.
I ran a rigorous coding benchmark on the new GPT-5.6 family to find out exactly how cheap you can go before your code breaks. One specific model configuration matched the price of older mini models while blowing them away on speed and accuracy. But watch out, because tweaking one single setting to save a few pennies will completely ruin your output.
I put four leading AI models to the test on a messy, real-world bug in an old codebase. While every model managed to patch the issue, only one specific engine approached the task like a disciplined senior engineer. You won't believe which model wrote a failing test first to prove the bug existed before writing a single line of production code.
I've just run the newly released Grok 4.5 through my grueling 25-prompt coding benchmark, and the results completely shocked me. After its disastrous earlier versions, this SpaceX-backed model did something I didn't think was possible on my leaderboard. It isn't perfect, but it might change which tool you use to write code tomorrow.
I've been benchmarking a newly released AI model that is currently completely free, and the results are turning the coding leaderboard upside down. While it completely crashed on a few niche frameworks, it managed to match the performance of one of the industry's most expensive premium models on complex web development tasks. You'll want to see where this zero-cost option lands before the free window closes.