<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Zain Dana Harper</title>
    <description>The latest articles on DEV Community by Zain Dana Harper (@zaindanaharper).</description>
    <link>https://dev.to/zaindanaharper</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4012901%2F96528b1b-4fc2-4151-b96d-d9a66ba6fca8.jpg</url>
      <title>DEV Community: Zain Dana Harper</title>
      <link>https://dev.to/zaindanaharper</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/zaindanaharper"/>
    <language>en</language>
    <item>
      <title>Five short papers on making AI-assisted work re-checkable</title>
      <dc:creator>Zain Dana Harper</dc:creator>
      <pubDate>Tue, 07 Jul 2026 04:56:55 +0000</pubDate>
      <link>https://dev.to/zaindanaharper/five-short-papers-on-making-ai-assisted-work-re-checkable-3n3n</link>
      <guid>https://dev.to/zaindanaharper/five-short-papers-on-making-ai-assisted-work-re-checkable-3n3n</guid>
      <description>&lt;p&gt;I have released five short papers on one idea: making AI-assisted work re-checkable. All five are archived on Zenodo with DOIs, and source and tests are public.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;EMET&lt;/strong&gt; — a byte-level integrity witness whose verdict {MATCH, DRIFT, UNVERIFIABLE} cannot, by construction, express "trusted." Four implementations, 44 conformance vectors. &lt;a href="https://doi.org/10.5281/zenodo.21230267" rel="noopener noreferrer"&gt;10.5281/zenodo.21230267&lt;/a&gt; · &lt;a href="https://github.com/HarperZ9/emet" rel="noopener noreferrer"&gt;source&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;BuildLang&lt;/strong&gt; — a compiler that puts ambient capabilities in the function type and seals re-derivable receipts verified by re-execution. &lt;a href="https://doi.org/10.5281/zenodo.21231253" rel="noopener noreferrer"&gt;10.5281/zenodo.21231253&lt;/a&gt; · &lt;a href="https://github.com/HarperZ9/buildlang" rel="noopener noreferrer"&gt;source&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Witnessed Independence&lt;/strong&gt; — records whether a verifier graded its own work, and refuses to decide when independence is not positively witnessed. &lt;a href="https://doi.org/10.5281/zenodo.21232206" rel="noopener noreferrer"&gt;10.5281/zenodo.21232206&lt;/a&gt; · &lt;a href="https://github.com/HarperZ9/coherence-membrane" rel="noopener noreferrer"&gt;source&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Proof Packets&lt;/strong&gt; — an envelope for one agent action whose verdict is derived from checks, so a claim can never vouch for itself. &lt;a href="https://doi.org/10.5281/zenodo.21231837" rel="noopener noreferrer"&gt;10.5281/zenodo.21231837&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Re-Perceived Effects&lt;/strong&gt; — replaces an actuator's self-report with a re-perceived effect: on-disk hash against intended hash, return code, witnessed bytes. &lt;a href="https://doi.org/10.5281/zenodo.21231747" rel="noopener noreferrer"&gt;10.5281/zenodo.21231747&lt;/a&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The common thread: a verdict should be a re-derivable function of its inputs, should say plainly what it does &lt;em&gt;not&lt;/em&gt; claim, and should fail to UNVERIFIABLE rather than to trust.&lt;/p&gt;

&lt;p&gt;More at &lt;a href="https://harperz9.github.io/" rel="noopener noreferrer"&gt;harperz9.github.io&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>opensource</category>
      <category>science</category>
      <category>softwareengineering</category>
    </item>
    <item>
      <title>Four short papers on making AI-assisted work re-checkable</title>
      <dc:creator>Zain Dana Harper</dc:creator>
      <pubDate>Tue, 07 Jul 2026 01:12:23 +0000</pubDate>
      <link>https://dev.to/zaindanaharper/four-short-papers-on-making-ai-assisted-work-re-checkable-48do</link>
      <guid>https://dev.to/zaindanaharper/four-short-papers-on-making-ai-assisted-work-re-checkable-48do</guid>
      <description>&lt;p&gt;I have released four short papers on one idea: making AI-assisted work re-checkable.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;EMET&lt;/strong&gt; — a byte-level integrity witness whose verdict {MATCH, DRIFT, UNVERIFIABLE} cannot, by construction, express "trusted." Four independent implementations, 44 conformance vectors.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;BuildLang&lt;/strong&gt; — a compiler that puts ambient capabilities in the function type and seals re-derivable receipts verified by re-execution.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Witnessed Independence&lt;/strong&gt; — a mechanism that records whether a verifier graded its own work, and refuses to decide when independence is not witnessed.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Proof Packets&lt;/strong&gt; — an envelope for one agent action whose verdict is derived from checks, so a claim can never vouch for itself.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Source and tests for all four are public.&lt;/p&gt;

&lt;p&gt;Papers: &lt;a href="https://harperz9.github.io/publications.html" rel="noopener noreferrer"&gt;https://harperz9.github.io/publications.html&lt;/a&gt;&lt;/p&gt;

</description>
      <category>ai</category>
      <category>opensource</category>
      <category>security</category>
      <category>testing</category>
    </item>
  </channel>
</rss>
