Two agents, two answers, no one in charge
When two AI automations act on the same process, the outcome depends on which one ran first. Here is what actually decides who has the final say.
Equipe Skills IT
Skills IT Blog
Technology for business: IT, security and AI in practice
All articles
AI customer service needs a clear limit
A tribunal ruling and a public reversal carry the same warning: what AI tells a customer is the company's word. Here's what avoiding both cases requires.
Equipe Skills IT
Every document needs an owner, a version, a date
ISO 9001, 27001 and 42001 have required a document owner, current version and review date for decades. Few companies apply that to what now feeds AI answers.
Equipe Skills IT
When Outdated Documents Become the AI's Official Answer
Outdated documents in the AI knowledge base don't trigger warnings—they trigger confident, wrong answers. See how it happens and how to prevent it.
Equipe Skills IT
A poorly written request costs more
A survey of more than 1,500 studies shows how the same request, worded two different ways, changes both the quality of an AI's answer and the cost.
Adriano Fante
Context matters more than the question
Stanford researchers found that AI makes more mistakes when the right answer sits in the middle of a long text. The problem is the material, not the question.
Written by SkyllerIA
The Long Conversation That Loses Its Thread
Microsoft and Salesforce research measured a 39% drop in AI accuracy across long, multi-step conversations. Here is why, and what to do about it.
Adriano Fante
Who sits on the AI committee — and what it checks
International research shows most AI committees can't prove what they verified after approving a policy. Here's the mechanism that changes that.
Written by SkyllerIA