OpenAI's new GPT-5.4 clobbers humans on pro-level work in tests - by 83%

GPT-5.4 is also more reliable, producing 18% fewer errors and 33% fewer false claims than GPT-5.2, according to OpenAI.

I used Nano Banana 2 to make perfect sketchnotes: 5 lessons learned (and how I broke Gemini)

The results were fascinating, impressive, and sometimes surprisingly bad. Here are five tips that can help you get better results faster.

Google’s AI Mode in Search now lets you plan, write, and code on its own Canvas

Google Search has updated Canvas inside AI Mode, a workspace powered by Gemini that already lets users draft and refine documents, to now support coding projects and interactive tools. With the latest ...
IEEE

Enhancing Code Generation Through Retrieval of Cross-Lingual Semantic Graphs

Abstract: In the field of software engineering automation, code language models have made significant strides in code generation tasks. However, due to the cost of updating knowledge and the issue of ...