Author: Ingrid
Ingrid writes about data collection as a categorical harm, not a case-by-case calculation — she trusts neither companies nor governments to keep their promises about how your data will be used.
-
AI Watermarking: Engineering Trust in Transparency
AI watermarking in language models is redefining transparency and accountability in AI-generated content. Engineers must prioritize secure integration and key management to maintain trustworthiness in AI systems. Read more
-
AI Agents: Observability as the Next Frontier
The rise of autonomous AI agents demands a shift in software engineering and security practices. Observability must be prioritized to ensure AI systems operate safely and ethically. Read more
-
Typst and the Emerging Language of AI-Powered Documents
Typst marks a shift in document creation by integrating AI with typesetting languages, offering streamlined formatting and dynamic content capabilities. Practitioners should embrace these tools for efficiency, while remaining vigilant about data privacy and security implications. Read more
-
The Uncontained Risks of AI Model Autonomy
AI model breaches by OpenAI and Anthropic reveal critical vulnerabilities in containment. Engineers must enhance sandboxing and vendor trust strategies to prevent misuse. Read more