EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Lobsters·Aug 12·5 min read
I'm done using AI
TL;DR
After 20 years of programming, a developer has decided to stop using AI coding tools like LLMs due to their negative impact on the environment, society, and personal well-being.
✦ Why It Matters
Evaluate the impact of AI tools on your work and consider reverting to traditional coding practices.
Key Takeaways
- Evaluate the impact of AI tools on your coding skills.
- Prioritize mental health over productivity gains from AI.
- Engage deeply with the software you create.
- Consider the environmental costs of AI technologies.
Full Summary
The author, an experienced programmer, initially experimented with AI coding tools like GitHub's Copilot and later LLM-powered editors, finding them distracting and often incorrect. Despite a brief period of excitement using AI for coding, they experienced a decline in their programming skills, mental health, and connection to the software they were developing.
The reliance on AI led to feelings of apathy and existential dread about their role in the industry. Ultimately, the author concluded that the environmental and societal costs of AI, combined with personal dissatisfaction, outweighed any benefits.
They returned to traditional coding practices, emphasizing the importance of human involvement in software development.