This is a much needed first attempt at a benchmark to measure how much given AI models will play along with users pushing them in delusional or potentially psychologically dangerous directions. Some early signal that full GPT-5 (not chat) is a less psychologically risky model.
ETHICS
-
Chat Control regulation and AI research critique emerge
By
–
bouarf, il aura moins d'impact que le chatcontrol qui va arriver, et d'autres truc a la con de chercheurs qui ont fait le helloquitx
-
Open House 2025: AI Implementation in Finance and Defense
By
–
[Event Report] Open House 2025: Tackling Tough Challenges in Finance and Defense, the Cutting Edge of AI Implementation in Society https://
sakana.ai/open-house-202
5/
… -
AI Chatbots as Therapy: Building Emotional Resilience Through ChatGPT
By
–
3 AI Prompts That Turn ChatGPT Into Your Friend And Therapist More people than ever are turning to #AI for #therapy and #companionship, using #chatbots to share their thoughts, process #emotions, and build #resilience. This article examines the growing trust in AI as a source
-
AI Capability Test: Personalized Reunion Speech Generation Benchmark
By
–
I accidentally stumbled upon a new AI vibe benchmark. Assuming you’re at least 5 years out of college, ask for an excruciatingly detailed reunion speech for your exact university and year. There is enough out there that a human could do this without attending the specific
-
AI writing style detection and recognition in social media
By
–
Interesting findings from this post: 1. It should be obvious to anyone who has interacted with LLMs before that the writing style of the tweet is a conspicuous caricature of AI slop (e.g. em dashes, the "it's not… it's…" construction, rambling, florid prose, etc.). Yet, many
-
Interns Scaling AI Experiments With Unprecedented Computing Power
By
–
the world of AI is crazy right now because an intern will be like "sorry, let me double-check the numbers on that tonight" and then spin up an experiment with the power of 10,000 washing machines
-
Echo chambers celebrating themselves: critical analysis
By
–
the echo chamber celebrating the echo chamber
-
Trust Over Technology: Human Behavior in Cybersecurity Breaches
By
–
When digital breaches strike with such magnitude, it becomes clear that trust, rather than technology, is often the weakest layer in cybersecurity, especially when human behavior turns into the attack surface. Infographic by @VisualCap via @antgrasso #cybersecurity
-
AGI Predictions Fall Short: AI Still Struggles with Basic Tasks
By
–
2024: In 2025 we will have AGI
2025: “how many Bs in blueberry” “what is 8.11-8.9”