· ai
A non-decision protocol for human–AI systems with explicit stop conditions
Overview I’m sharing a technical note proposing a non-decision protocol for human–AI systems. The core idea is simple: AI systems should not decide. They shoul...
-
- · ai
Will AI Ever Be Good Enough to Not Need Spending Limits?
'markdown “Won’t AI just get better at this?” Short answer No. Understanding why reveals something fundamental about how we should think about AI safety.
- · ai
All AI Videos Are Harmful (2025)
Article URL: https://idiallo.com/blog/all-ai-videos-are-harmful Comments URL: https://news.ycombinator.com/item?id=46498651 Points: 19 Comments: 6...
- · ai
Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
Overview Meet Llama Guard, a simple tool built to make chats with AI safer and clearer for everyone. It looks at what people ask and what the AI answers, and s...
- · ai
AI sycophancy panic
Article URL: https://github.com/firasd/vibesbench/blob/main/docs/ai-sycophancy-panic.md Comments URL: https://news.ycombinator.com/item?id=46488396 Points: 38 C...
- · ai
AI Sycophancy Panic
Article URL: https://github.com/firasd/vibesbench/blob/main/docs/ai-sycophancy-panic.md Comments URL: https://news.ycombinator.com/item?id=46488396 Points: 10 C...
- · ai
Nightshade: Make images unsuitable for model training
Article URL: https://nightshade.cs.uchicago.edu/whatis.html Comments URL: https://news.ycombinator.com/item?id=46487342 Points: 16 Comments: 2...
- · ai
In the next 30 days, I’m talking about the democratisation of AI with one mission: AI should feel practical, affordable, and safe, especially for small businesses and founders.
Cleaned‑up Markdown markdown !Forem Logohttps://media2.dev.to/dynamic/image/width=65,height=,fit=scale-down,gravity=auto,format=auto/https%3A%2F%2Fdev-to-upload...
- · ai
Adversarial Attacks and Defences: A Survey
Overview Today many apps use deep learning to perform complex tasks quickly, from image analysis to voice recognition. However, tiny, almost invisible changes...
- · ai
Instructions Are Not Control
!Cover image for Instructions Are Not Controlhttps://media2.dev.to/dynamic/image/width=1000,height=420,fit=cover,gravity=auto,format=auto/https%3A%2F%2Fdev-to-u...
- · ai
The Loop Changes Everything: Why Embodied AI Breaks Current Alignment Approaches
Stateless vs. Stateful AI ChatGPT and similar chat models are stateless: each API call is independent and the model has no: - Persistent memory – it forgets ev...
- · ai
Stop Begging Your AI to Be Safe: The Case for Constraint Engineering
I am tired of “Prompt Engineering” as a safety strategy. If you are building autonomous agents—AI that can actually do things like query databases, move files,...