<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Role-Based AI on Christa Burger</title><link>https://christaburger.com/tags/role-based-ai/</link><description>Recent content in Role-Based AI on Christa Burger</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Wed, 03 Jun 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://christaburger.com/tags/role-based-ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Give AI a Job Description Before Giving It Access</title><link>https://christaburger.com/blog/give-ai-a-job-description-before-access/</link><pubDate>Wed, 03 Jun 2026 00:00:00 +0000</pubDate><guid>https://christaburger.com/blog/give-ai-a-job-description-before-access/</guid><description>&lt;p&gt;Before giving AI more access, give it a job description.&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;What is it responsible for?&lt;/li&gt;
&lt;li&gt;What should it ignore?&lt;/li&gt;
&lt;li&gt;What does good output look like?&lt;/li&gt;
&lt;li&gt;What tools can it use?&lt;/li&gt;
&lt;li&gt;When should it escalate?&lt;/li&gt;
&lt;li&gt;What should it never decide?&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;This sounds obvious.&lt;/p&gt;
&lt;p&gt;It is apparently not.&lt;/p&gt;
&lt;p&gt;People will give AI access to meaningful workflows with less role clarity than they would give an intern named Brayden who starts Monday and says &amp;ldquo;for sure&amp;rdquo; too much.&lt;/p&gt;</description></item><item><title>Not Every AI Should Be the Same AI</title><link>https://christaburger.com/blog/not-every-ai-should-be-the-same-ai/</link><pubDate>Mon, 01 Jun 2026 00:00:00 +0000</pubDate><guid>https://christaburger.com/blog/not-every-ai-should-be-the-same-ai/</guid><description>&lt;p&gt;I used to put too much into one AI conversation.&lt;/p&gt;
&lt;p&gt;Strategy. Drafting. Summarizing. Planning. Debugging. Content ideas. Household systems. Possibly plant identification.&lt;/p&gt;
&lt;p&gt;A perfectly normal range of human concerns.&lt;/p&gt;
&lt;p&gt;Eventually, I realized the issue was not the model.&lt;/p&gt;
&lt;p&gt;It was the lack of role clarity.&lt;/p&gt;
&lt;p&gt;Human teams have roles for a reason:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;Strategist&lt;/li&gt;
&lt;li&gt;Analyst&lt;/li&gt;
&lt;li&gt;Operator&lt;/li&gt;
&lt;li&gt;Reviewer&lt;/li&gt;
&lt;li&gt;Auditor&lt;/li&gt;
&lt;li&gt;Assistant&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;AI needs the same distinction.&lt;/p&gt;
&lt;p&gt;Some AI functions should explore. Some should execute. Some should check. Some should preserve context. Some should challenge assumptions. Some should produce clean artifacts.&lt;/p&gt;</description></item></channel></rss>