Google JUST WON the AI Coding Game...

Incredible how google turned it around from being a joke to a leading LLM in the AI race

Google are firmly in the lead with Gemini 2.5 Pro... If Ultra is a step up, the competition is cooked.

You said that last time, and in my tests it was still crap

Please test 2.5 pro PREVIEW and let us know how it compares to this older 2.5 pro experimental model you tested. Please and ty!! 🙏 You rock for sharing and detailing your tests ❤

I was working yesterday on some brittle code with several models - hours of frustration. I heard about the new Gemini Pro 2.5 drop and so I packaged up my small app dir with Repo Prompt (highly recommended) and one shoted a working solution in AI Studio (it wasn't available in the API yet). It really is better, and I was already using the earlier version of Pro2.5 as my daily driver. The speed of coding model improvement is incredible!

Nice job Wes, it was interesting seeing how much fun you had with this one

thia is CRAZY impression. tried my basic text-based game prompt, and it spit out something that beat the prior 2.5 pro rather well. not bad at all for a one shot. your videos are my FAVORITE for Ai news bc you ACTUALLY demo it for us and dont just talk and talk like most others and youre incredibly consistent. thanks Wes!

thanks for the ai coding news, it really helps me to stay up to date

Almost got hit by a full self driving Tesla.

One exponential step closer to a matrix style realistic simulation