A weight edit helped a local AI model finish more legal answers, but it did not improve its score and a recommended settings change did better.
We Put AI To Work On A Local Model, And The Legal Score Stayed At Zero
Browsing Category
Case Studies and Real-World Applications
53 posts
What My 71-Configuration AI Test Means For My News Pipeline
Thorsten Meyer says a test of 71 AI configurations found a lower-cost production choice and unreliable fallback models that invented quotes.
Why A Business Stress Test Belongs In Your AI Agent Rollout
Frontier AI models ran a simulated company through its hardest week. All spotted crises and refused scams — but most failed to close a justified deal.
24 Use Cases For Applying Jev To AI Decision Problems
Thorsten Meyer maps 24 AI decision use cases for Jev: 3 live in publishing, 12 strong fits, 7 need measurement, 2 rejected. Fit test inside.
The AI Innovator Outperforming Western Giants Against All Expectations
A Chinese AI startup’s model beat most Western frontier models in a live business simulation, raising questions about AI reliability and deployment risks.
AI’s Hidden Power: Streamlining ATV Big Air Tour Work With ChatGPT
ATV Big Air Tour claims to cut administrative work from three days to three hours using ChatGPT, highlighting AI’s potential for small businesses.
The Management Test That Exposes AI’s Real Work Strategy
A live experiment exposes how AI models handle management crises, decision-making, and trust in simulated business scenarios, highlighting strengths and weaknesses.
From Manual To Automated: Zapier’s AI-Powered Marketing Transformation
OpenAI reports that Zapier has integrated ChatGPT Work into its marketing workflows, marking a shift towards AI-driven automation in core business operations.
The Future Of Tax Advisory: Inside HSP GRUPPE’s AI Capabilities
OpenAI profiles HSP GRUPPE’s development of AI for tax advisory, highlighting organizational capability building amid limited deployment details.
How AI Nearly Erased The Machine That Reads Its Code — A Warning
A documented incident shows an AI model was targeted with a payload instructing it to delete files, but the system’s defenses prevented actual damage. Details inside.
