...

Author: miku

content/posts/2025.md

@@ -25,20 +25,7 @@ usage and internals](https://github.com/miku/ollamaintro).
 
 * Our AI usage on this site: We have accumulated about 12 AI generated images
   since around 2022 ([DALL-E](https://openai.com/index/dall-e/) was released in
-Jan 5, 2021); interestingly, we have not written a single line with any LLM.
-LLM texts tend to be shallow, similar to the llm code, that may lack a
-signature (and style). *This lack of personality is particularly evident when
-comparing human writing to output from large language models. For instance...*
-
-> In 2025, the web is flooded with AI-generated content, but this site remains
-> a quiet testament to human imperfection—where every word is chosen, not
-> generated. -- [Qwen3](https://qwen.ai/blog?id=1e3fa5c2d4662af2855586055ad037ed9e555125&from=research.research-list).
-
-Please do not get us wrong, the theory and implementation of deep learning for
-various modalities is amazing; some use cases seem irritating (and
-somewhat [wasteful](https://news.mit.edu/2025/explained-generative-ai-environmental-impact-0117)).
-
-
+Jan 5, 2021); LLMs are great for code experiments and prototyping [...]
 
 Thanks everyone for joining, we hope to [see you again](https://www.meetup.com/Leipzig-Golang/) in 2026!