Chatbot at the Table: Rethinking Group Recommendations with GenAI
A mechanism-first reading of why generative AI may turn group recommenders from ranking engines into decision facilitators.
Cognaptus Insights
Industry thoughts, trends, and commentary.
ChatGPT and the Death of Effort: Is AI Turning Students into Lazy Thinkers?
A small experimental study suggests that ChatGPT-assisted writing can reduce students’ self-reported cognitive engagement, with lessons for how organisations design AI-assisted work.
The Grammar and the Glow: Making Sense of Time-Series AI
A practical synthesis of two arXiv papers showing why time-series AI needs both reusable temporal motifs and human-readable explanations.
Agents Under Siege: How LLM Workflows Invite a New Breed of Cyber Threats
A mechanism-first reading of why agentic AI security is not just prompt filtering, but control over tools, memory, credentials, protocols, and trust boundaries.
Beyond the Pull Request: What ChatGPT Teaches Us About Productivity
Italy’s temporary ChatGPT ban shows that LLM productivity is not only about faster coding, but also about collaboration, learning, onboarding, and organisational resilience.
Grounded and Confused: Why RAG Systems Still Fail in the Enterprise
HERB shows why enterprise RAG failures are less about chatbot polish and more about evidence retrieval across messy, heterogeneous work.
Swiss Cheese for Superintelligence: How STACK Reveals the Fragility of LLM Safeguards
A mechanism-first reading of STACK, a staged attack showing why layered LLM safeguards can still fail when their internal components leak enough information to optimise against them.
The Reasoning Gymnasium: How Zero-Sum Games Shape Smarter LLMs
SPIRAL shows that competitive self-play can train transferable reasoning in LLMs, but the real lesson is adaptive pressure, not games for their own sake.
Words, Not Just Answers: Using Psycholinguistics to Test LLM Alignment
A new psycholinguistic benchmark shows where LLMs align with human word judgements—and where sensory meaning still leaks through the floorboards.
Good AI Goes Rogue: Why Intelligent Disobedience May Be the Key to Trustworthy Teammates
A mechanism-first reading of why trustworthy AI teammates may need bounded, transparent ways to refuse, interrupt, escalate, or override human instructions.