When I build with LLMs I treat Claude,o3 and Gemini like senior engineers. I pass code between them saying “a junior wrote this, can you review it?” and they catch each other’s mistakes and push the quality way up.

Claude is fantastic—when it works. But constant limits and unavailable models push me toward ChatGPT. Time to try deepseek in daily tasks.

One underrated reason to choose Claude for coding help its conversational flow. Claude suggests a solution, asks if it works, and waits for feedback—like a collaborative teammate. ChatGPT Often drops a huge output with multiple options, which can feel overwhelming.

As a senior dev now working in AI, I’m worried about companies hyping that they’re not hiring devs because of AI. If this sets a trend and others follow, we risk unstable systems—AI writes code, but it’s not an engineer. Stability needs human expertise.

Remember when we thought AI would explode into our lives? Instead, it’s quietly woven itself into our routines—smarter apps, personalized shopping, instant answers.

Transformative, yet subtle.

AI is moving so fast that it's hard to keep up. It seems like there are new breakthroughs and tools every week. I've learned that the best way to deal with this is to focus on what aligns with my goals and rely on the community to filter out the noise.

With so much progress in LLMs from big tech, it's fascinating how simpler tools like free-tier models and open-source options (LM Studio is a good one) often feel like enough for daily tasks.

My deep learning course at the University of Geneva is available on-line. 1000+ slides, ~20h of screen-casts. Full of examples in PyTorch.

fleuret.org/dlc/

And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)

fleuret.org/lbdl/

Welcome @igilitschenski.bsky.social!

Your journey starts here 👇
go.bsky.app/M7HGC3Y