EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
MIT Technology Review·Aug 18·5 min read
The Download: how people really use AI, and Flock’s design choices
TL;DR
AI companies like Anthropic and OpenAI provide limited insights into user behavior, prompting the AI Observatory to reveal broader usage patterns. Their findings indicate significant differences in how people utilize various AI models for tasks like coding, social interaction, and homework assistance.
✦ Why It Matters
Consider using independent research sources to inform your understanding of AI user behavior.
Key Takeaways
- Explore independent research on AI usage for better insights.
- Recognize the varied applications of different AI models.
- Stay informed about user preferences in AI interactions.
- Consider the implications of selective reporting by AI companies.
Full Summary
AI companies often publish selective reports on user engagement, leaving a gap in understanding actual usage patterns. The AI Observatory, a new research initiative, aims to address this by analyzing user interactions with different AI models.
Their research indicates that users prefer Anthropic for coding tasks, Gemini for social and roleplay interactions, and ChatGPT for academic assistance. This contrasts with the more work-focused narratives presented by AI companies.
The study highlights the need for independent analysis to better understand the diverse applications of AI in personal and professional contexts. By revealing these insights, the AI Observatory contributes to a more comprehensive understanding of AI's role in everyday life.