<rss xmlns:source="http://source.scripting.com/" version="2.0">
  <channel>
    <title>Crank.report Survey</title>
    <link>https://crank.report/</link>
    <description></description>
    
    <language>en</language>
    
    <lastBuildDate>Tue, 17 Nov 2026 05:42:44 -0800</lastBuildDate>
    <item>
      <title></title>
      <link>https://crank.report/2026/11/17/so-if-verification-is-critical.html</link>
      <pubDate>Tue, 17 Nov 2026 05:42:44 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/11/17/so-if-verification-is-critical.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;So if verification is critical to producing quality code at a high velocity, what types of tests should you use and how should you apply them across your pipeline? This article answers those questions and also shares the process used by the Claude Code team.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;From &lt;a href=&#34;https://claude.dev/blog/how-the-claude-code-team-thinks-about-verification/&#34;&gt;How the Claude Code team thinks about verification / claude.dev Blog&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;I&amp;rsquo;m interested in thinking about how these processes apply to non-code work.&lt;/p&gt;
</description>
      <source:markdown>&gt; So if verification is critical to producing quality code at a high velocity, what types of tests should you use and how should you apply them across your pipeline? This article answers those questions and also shares the process used by the Claude Code team.

From [How the Claude Code team thinks about verification / claude.dev Blog](https://claude.dev/blog/how-the-claude-code-team-thinks-about-verification/)

I&#39;m interested in thinking about how these processes apply to non-code work. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/11/17/we-have-federal-agencies-and.html</link>
      <pubDate>Tue, 17 Nov 2026 05:42:05 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/11/17/we-have-federal-agencies-and.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;We have federal agencies and congressional committees designed explicitly to monitor, protect and improve these complex systems. But they&amp;rsquo;ve atrophied as power has been centralized in the presidency and congressional leadership, where daily political warfare rarely includes sober discussion of competence and essential interconnected systems.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;From &lt;a href=&#34;https://www.axios.com/2026/11/17/government-competence-aviation-infrastructure-crisis&#34;&gt;What if competence were cool&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;There is a huge value to the unsexy effort required to make systems work.&lt;/p&gt;
</description>
      <source:markdown>&gt; We have federal agencies and congressional committees designed explicitly to monitor, protect and improve these complex systems. But they&#39;ve atrophied as power has been centralized in the presidency and congressional leadership, where daily political warfare rarely includes sober discussion of competence and essential interconnected systems.


From [What if competence were cool](https://www.axios.com/2026/11/17/government-competence-aviation-infrastructure-crisis)

There is a huge value to the unsexy effort required to make systems work. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/11/15/reporting-back-tern-and-the.html</link>
      <pubDate>Sun, 15 Nov 2026 20:07:59 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/11/15/reporting-back-tern-and-the.html</guid>
      <description>&lt;h2 id=&#34;reporting-back-tern-and-the-library-list&#34;&gt;Reporting back: Tern and the library list&lt;/h2&gt;
&lt;p&gt;In August I &lt;a href=&#34;https://crank.report/2026/08/29/next-boring-tiny-tool-ive.html&#34;&gt;wrote about a small problem&lt;/a&gt;: I keep &lt;a href=&#34;https://crank.report/to-read/&#34;&gt;the big list of books I want to read&lt;/a&gt; in &lt;a href=&#34;https://epilogue.micro.blog&#34;&gt;Epilogue&lt;/a&gt;, and then I never actually look at it when I open &lt;a href=&#34;https://www.overdrive.com/apps/libby&#34;&gt;Libby&lt;/a&gt; to check out my next batch of books from the library. I promised I would report back when I built a &lt;a href=&#34;https://vaughntan.org/boringtinytools&#34;&gt;boring tiny tool&lt;/a&gt; to fix it.&lt;/p&gt;
&lt;p&gt;It is done, and I’ve put it on &lt;a href=&#34;https://github.com/marniewebb-ts&#34;&gt;GitHub&lt;/a&gt; as the first piece of &lt;a href=&#34;https://github.com/marniewebb-ts/tern&#34;&gt;Tern&lt;/a&gt;, a set of small tools for my personal Obsidian vault.&lt;/p&gt;
&lt;p&gt;Every morning at 5:00 AM, a &lt;code&gt;launchd&lt;/code&gt; job runs a Python script that reads my published Epilogue bookshelf and checks every book against the Berkeley Public Library catalog on Libby. It writes two things into my vault: a master availability page with one-click borrow and hold links, and a short section in my daily note that tells me if something on my list just flipped from the waitlist to available so I can swoop in and grab it. On its first run it found 18 books on my list that were sitting on the digital shelf waiting for me.&lt;/p&gt;
&lt;p&gt;I used Gemini to help me build it and package it as a skill, but once it is set up it doesn&amp;rsquo;t require AI to run. It is just a scheduled Python script, public library data, and plain Markdown files that I own. If you use Epilogue, Libby, and Obsidian, I included &lt;a href=&#34;https://github.com/marniewebb-ts/tern/tree/main/skills/tern-reading&#34;&gt;instructions in the repo&lt;/a&gt; so you can point it at your own bookshelf and &lt;a href=&#34;https://librarymap.ifla.org/&#34;&gt;your own public library&lt;/a&gt;.&lt;/p&gt;
</description>
      <source:markdown>## Reporting back: Tern and the library list

In August I [wrote about a small problem](https://crank.report/2026/08/29/next-boring-tiny-tool-ive.html): I keep [the big list of books I want to read](https://crank.report/to-read/) in [Epilogue](https://epilogue.micro.blog), and then I never actually look at it when I open [Libby](https://www.overdrive.com/apps/libby) to check out my next batch of books from the library. I promised I would report back when I built a [boring tiny tool](https://vaughntan.org/boringtinytools) to fix it.


It is done, and I’ve put it on [GitHub](https://github.com/marniewebb-ts) as the first piece of [Tern](https://github.com/marniewebb-ts/tern), a set of small tools for my personal Obsidian vault.


Every morning at 5:00 AM, a `launchd` job runs a Python script that reads my published Epilogue bookshelf and checks every book against the Berkeley Public Library catalog on Libby. It writes two things into my vault: a master availability page with one-click borrow and hold links, and a short section in my daily note that tells me if something on my list just flipped from the waitlist to available so I can swoop in and grab it. On its first run it found 18 books on my list that were sitting on the digital shelf waiting for me.


I used Gemini to help me build it and package it as a skill, but once it is set up it doesn&#39;t require AI to run. It is just a scheduled Python script, public library data, and plain Markdown files that I own. If you use Epilogue, Libby, and Obsidian, I included [instructions in the repo](https://github.com/marniewebb-ts/tern/tree/main/skills/tern-reading) so you can point it at your own bookshelf and [your own public library](https://librarymap.ifla.org/).

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/11/15/all-that-is-needed-to.html</link>
      <pubDate>Sun, 15 Nov 2026 09:58:17 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/11/15/all-that-is-needed-to.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;All that is needed to create an agent is a folder with Markdown files.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;Instructions, context, and reusable skills. Access to tools. A way to invoke the agent, validate the results, and incorporate feedback. Most of this can be set up simply by chatting with Claude.&lt;/p&gt;
&lt;p&gt;&lt;a href=&#34;https://claramap.com/blog/what-picasso-taught-me-about-building-ai-agents/&#34;&gt;What Picasso Taught Me About Building AI Agents — Claramap&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;This is my workflow too but it’s nontrivial. It invoked homebrew and python certificates and launchd files. Once you get going it’s wonderful. For a non-engineer the getting going takes work and a little bit of computational courage.&lt;/p&gt;
</description>
      <source:markdown>&gt; All that is needed to create an agent is a folder with Markdown files.

Instructions, context, and reusable skills. Access to tools. A way to invoke the agent, validate the results, and incorporate feedback. Most of this can be set up simply by chatting with Claude.

[What Picasso Taught Me About Building AI Agents — Claramap](https://claramap.com/blog/what-picasso-taught-me-about-building-ai-agents/)


This is my workflow too but it’s nontrivial. It invoked homebrew and python certificates and launchd files. Once you get going it’s wonderful. For a non-engineer the getting going takes work and a little bit of computational courage. 

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/11/10/tk-is-a-private-agentic.html</link>
      <pubDate>Tue, 10 Nov 2026 12:11:25 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/11/10/tk-is-a-private-agentic.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;TK is a private, agentic notebook to craft writing that compounds into deep thinking, which creates influence across your entire world. What shape is that influence? We’re still figuring that out, but we know it starts with humans writing more.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;From &lt;a href=&#34;https://medium.com/blog/tk-a-new-private-notebook-app-from-medium-13d3d6efdce1&#34;&gt;TK a new private notebook app from medium&lt;/a&gt;&lt;/p&gt;
&lt;ol&gt;
&lt;li&gt;I live the name.&lt;/li&gt;
&lt;li&gt;I&amp;rsquo;ve built myself a set of skills files, collectively called Kite, the essentially do this through a daily, weekly, and monthly interaction with my Obsidian vault.&lt;/li&gt;
&lt;/ol&gt;
</description>
      <source:markdown>&gt; TK is a private, agentic notebook to craft writing that compounds into deep thinking, which creates influence across your entire world. What shape is that influence? We’re still figuring that out, but we know it starts with humans writing more.

From [TK a new private notebook app from medium](https://medium.com/blog/tk-a-new-private-notebook-app-from-medium-13d3d6efdce1)

1. I live the name. 
2. I&#39;ve built myself a set of skills files, collectively called Kite, the essentially do this through a daily, weekly, and monthly interaction with my Obsidian vault. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/11/09/as-useful-as-ai-may.html</link>
      <pubDate>Mon, 09 Nov 2026 15:34:36 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/11/09/as-useful-as-ai-may.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;As useful as AI may be, concentrating resources in the hands of a handful of companies — now that’s always “pretty terrible” for society as a whole.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;From &lt;a href=&#34;https://manualdousuario.net/en/memory-crises-bloated-software/&#34;&gt;8 GB ought to be enough for anybody ⁄ Manual do Usuário&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;We can&amp;rsquo;t have agency in this new agentic world without computing power we can own.&lt;/p&gt;
</description>
      <source:markdown>&gt; As useful as AI may be, concentrating resources in the hands of a handful of companies — now that’s always “pretty terrible” for society as a whole.

From [8 GB ought to be enough for anybody ⁄ Manual do Usuário](https://manualdousuario.net/en/memory-crises-bloated-software/)

We can&#39;t have agency in this new agentic world without computing power we can own. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/28/in-states-like-oklahoma-and.html</link>
      <pubDate>Wed, 28 Oct 2026 06:18:32 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/28/in-states-like-oklahoma-and.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;In states like Oklahoma and Florida that have adopted these new SNAP restrictions to discourage unhealthy diets, the number of convenience stores participating in the program has fallen relative to other states that haven&amp;rsquo;t made these changes. That was especially notable in areas with high numbers of low-income people where a large share of the population gets SNAP benefits.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://theconversation.com/new-restrictions-on-purchases-with-snap-benefits-in-some-states-mean-fewer-convenience-stores-are-accepting-them-294540&#34;&gt;Some states add new restrictions to SNAP&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;I find this so offensive. It removes agency from people and takes away what can be an affordable way to say yes something.&lt;/p&gt;
</description>
      <source:markdown>&gt; In states like Oklahoma and Florida that have adopted these new SNAP restrictions to discourage unhealthy diets, the number of convenience stores participating in the program has fallen relative to other states that haven&#39;t made these changes. That was especially notable in areas with high numbers of low-income people where a large share of the population gets SNAP benefits.

[Some states add new restrictions to SNAP](https://theconversation.com/new-restrictions-on-purchases-with-snap-benefits-in-some-states-mean-fewer-convenience-stores-are-accepting-them-294540)

I find this so offensive. It removes agency from people and takes away what can be an affordable way to say yes something. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/10/this-creates-data-voids-gaps.html</link>
      <pubDate>Sat, 10 Oct 2026 04:52:44 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/10/this-creates-data-voids-gaps.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;This creates data voids &amp;ndash; gaps in credible coverage that can be manipulated or exploited by bad actors to fill with low-quality information. This also explains why, in some specific cases, AI outputs are very unreliable, even when there&amp;rsquo;s no coordinated effort to pollute the information ecosystem. We see that with context-specific and geography-specific questions about voting procedures. The AI systems have little authoritative material to draw from, so they surface low-quality information.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://themarkup.org/hello-world/2026/10/10/election-ai-resilience&#34;&gt;What it takes for voters to be resilient against AI misinformation&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;The idea of data gaps &amp;ndash; places where there is not authoritative information &amp;ndash; is new to me though it makes perfect sense. I wonder if this is a role libraries can play.&lt;/p&gt;
</description>
      <source:markdown>&gt; This creates data voids -- gaps in credible coverage that can be manipulated or exploited by bad actors to fill with low-quality information. This also explains why, in some specific cases, AI outputs are very unreliable, even when there&#39;s no coordinated effort to pollute the information ecosystem. We see that with context-specific and geography-specific questions about voting procedures. The AI systems have little authoritative material to draw from, so they surface low-quality information.


[What it takes for voters to be resilient against AI misinformation](https://themarkup.org/hello-world/2026/10/10/election-ai-resilience)

The idea of data gaps -- places where there is not authoritative information -- is new to me though it makes perfect sense. I wonder if this is a role libraries can play. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/08/if-a-junior-programmer-does.html</link>
      <pubDate>Thu, 08 Oct 2026 15:19:03 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/08/if-a-junior-programmer-does.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;If a junior programmer does not learn to write code and simply generates it, they are robbing themselves of the opportunity to develop the visceral understanding of code that comes with being down in the trenches.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://htmx.org/essays/yes-and/&#34;&gt;htmx ~ Yes, and&amp;hellip;&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;More support for the productive friction that comes when we struggle and results in learning and judgment. (&lt;a href=&#34;https://simonwillison.net/2026/Oct/8/carson-gross/&#34;&gt;via&lt;/a&gt;)&lt;/p&gt;
</description>
      <source:markdown>&gt; If a junior programmer does not learn to write code and simply generates it, they are robbing themselves of the opportunity to develop the visceral understanding of code that comes with being down in the trenches.

[htmx ~ Yes, and...](https://htmx.org/essays/yes-and/)

More support for the productive friction that comes when we struggle and results in learning and judgment. ([via](https://simonwillison.net/2026/Oct/8/carson-gross/))
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/03/a-german-team-built-kolibri.html</link>
      <pubDate>Sat, 03 Oct 2026 07:14:21 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/03/a-german-team-built-kolibri.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;…a German team built [Kolibri] with the EU AI Act in mind “from the ground up”, and in Aleph Alpha’s own evaluation it scores above every compared model of its size in both languages.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://tej.as/blog/aleph-alpha-kolibri&#34;&gt;Aleph Alpha Kolibri: How the Sovereign German LLM Works&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;I’m interested in doing a version of what we did with the &lt;a href=&#34;https://offercenter.techsoup.org&#34;&gt;Civil Society Offer Center&lt;/a&gt; with Eau friendly technologies. We’re wrestling with an architectural problem as a part of that but we’ll get there.&lt;/p&gt;
</description>
      <source:markdown>&gt; …a German team built [Kolibri] with the EU AI Act in mind “from the ground up”, and in Aleph Alpha’s own evaluation it scores above every compared model of its size in both languages.

[Aleph Alpha Kolibri: How the Sovereign German LLM Works](https://tej.as/blog/aleph-alpha-kolibri)


I’m interested in doing a version of what we did with the [Civil Society Offer Center](https://offercenter.techsoup.org) with Eau friendly technologies. We’re wrestling with an architectural problem as a part of that but we’ll get there. 

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/03/large-language-models-were-built.html</link>
      <pubDate>Sat, 03 Oct 2026 06:56:36 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/03/large-language-models-were-built.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;Large language models were built to talk to people. Jev is built to be called by machines. You give it some context and the exact questions you need answered, and it replies with structured values instead of prose. Because the shape of the answer is fixed by your request, Jev cannot hallucinate a format or emit an invalid type — it can only return one of the options you defined, a score on the scale you gave it, or a probability between 0 and 1.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://jevtypesafeai.com/what-is-jev&#34;&gt;What is Jev? TypeSafe’s System One model explained · Jev by TypeSafe AI&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Could be interesting to experiment with this in any of our rubrics but particularly thinking of rubrics for scoring bits of code that could be shared in a marketplace setting.&lt;/p&gt;
</description>
      <source:markdown>&gt; Large language models were built to talk to people. Jev is built to be called by machines. You give it some context and the exact questions you need answered, and it replies with structured values instead of prose. Because the shape of the answer is fixed by your request, Jev cannot hallucinate a format or emit an invalid type — it can only return one of the options you defined, a score on the scale you gave it, or a probability between 0 and 1.

[What is Jev? TypeSafe’s System One model explained · Jev by TypeSafe AI](https://jevtypesafeai.com/what-is-jev)


Could be interesting to experiment with this in any of our rubrics but particularly thinking of rubrics for scoring bits of code that could be shared in a marketplace setting. 

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/03/from-a-technical-point-of.html</link>
      <pubDate>Sat, 03 Oct 2026 06:42:15 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/03/from-a-technical-point-of.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;From a technical point of view, the use of NixOS was a key one. Its use of the Nix package manager means that each package is installed in its own directory with immutable and signed contents. That alone means that the setup is incredibly easy to reproduce across multiple machines, something that is an obvious benefit when dealing with different facets of a government. It&amp;rsquo;s been suggested that up to 90% of a configuration can be carried over from one deployment to another. And as a bonus, NixOS can also run on computers that Windows 11 would usually balk at.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://www.tomshardware.com/software/the-netherlands-is-rolling-alternative-nixos-based-software-ecosystem-after-u-s-sanctions-on-icc-took-microsoft-off-the-table-trial-programs-running-now-first-release-expected-at-end-of-2027&#34;&gt;US sanctions force The Netherlands off Microsoft and toward alternative NixOS-based software ecosystem — trial programs running now, first release expected at end of 2027 | Tom&amp;rsquo;s Hardware&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Could this be a model to work with small organizations?&lt;/p&gt;
</description>
      <source:markdown>&gt; From a technical point of view, the use of NixOS was a key one. Its use of the Nix package manager means that each package is installed in its own directory with immutable and signed contents. That alone means that the setup is incredibly easy to reproduce across multiple machines, something that is an obvious benefit when dealing with different facets of a government. It&#39;s been suggested that up to 90% of a configuration can be carried over from one deployment to another. And as a bonus, NixOS can also run on computers that Windows 11 would usually balk at.

[US sanctions force The Netherlands off Microsoft and toward alternative NixOS-based software ecosystem — trial programs running now, first release expected at end of 2027 | Tom&#39;s Hardware](https://www.tomshardware.com/software/the-netherlands-is-rolling-alternative-nixos-based-software-ecosystem-after-u-s-sanctions-on-icc-took-microsoft-off-the-table-trial-programs-running-now-first-release-expected-at-end-of-2027)

Could this be a model to work with small organizations?
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/03/today-a-designer-can-generate.html</link>
      <pubDate>Sat, 03 Oct 2026 05:31:21 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/03/today-a-designer-can-generate.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;Today, a designer can generate polished images, presentations, variations, and even entire visual directions in minutes. The ability to produce is becoming abundant—and when production becomes abundant, something else becomes scarce.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;Judgment.&lt;/p&gt;
&lt;p&gt;&lt;a href=&#34;https://www.core77.com/posts/145488/AI-Has-Changed-What-Makes-a-Great-Designer&#34;&gt;AI Has Changed What Makes a Great Designer  - Core77&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Applicable to many different kinds of talent.&lt;/p&gt;
</description>
      <source:markdown>&gt; Today, a designer can generate polished images, presentations, variations, and even entire visual directions in minutes. The ability to produce is becoming abundant—and when production becomes abundant, something else becomes scarce.

Judgment.

[AI Has Changed What Makes a Great Designer  - Core77](https://www.core77.com/posts/145488/AI-Has-Changed-What-Makes-a-Great-Designer)


Applicable to many different kinds of talent. 

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/10/03/people-make-pages-prototypes-reports.html</link>
      <pubDate>Sat, 03 Oct 2026 05:25:18 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/10/03/people-make-pages-prototypes-reports.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;People make pages, prototypes, reports, visualizations, and small apps with agents every day. Too often the result stays trapped in a local folder or a conversation. Spacefast closes that last mile: publish the files, get a live URL, and send it to the person who needs it.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://spacefast.com/about/&#34;&gt;About Spacefast | Spacefast&lt;/a&gt;, &lt;a href=&#34;https://www.manton.org/2026/09/29/after-a-reminder-from-danielpunkass.html&#34;&gt;via&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;I’ve been thinking a lot about how you construct an experience for people that uses saved artifacts like way stations. This kind of of tooling could be used to support that.&lt;/p&gt;
</description>
      <source:markdown>&gt; People make pages, prototypes, reports, visualizations, and small apps with agents every day. Too often the result stays trapped in a local folder or a conversation. Spacefast closes that last mile: publish the files, get a live URL, and send it to the person who needs it.

[About Spacefast | Spacefast](https://spacefast.com/about/), [via](https://www.manton.org/2026/09/29/after-a-reminder-from-danielpunkass.html)


I’ve been thinking a lot about how you construct an experience for people that uses saved artifacts like way stations. This kind of of tooling could be used to support that. 

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/09/24/the-relevant-question-is-what.html</link>
      <pubDate>Thu, 24 Sep 2026 05:41:22 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/09/24/the-relevant-question-is-what.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;The relevant question is what the load-bearing seams reveal about the assumptions underneath the conclusion, and whether those assumptions remain intact once the smoking gun is incorporated into the model.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://madradavid.com/claudes-load-bearing-seams/&#34;&gt;Claude&amp;rsquo;s load-bearing seams&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Oh so good.&lt;/p&gt;
</description>
      <source:markdown>&gt; The relevant question is what the load-bearing seams reveal about the assumptions underneath the conclusion, and whether those assumptions remain intact once the smoking gun is incorporated into the model.

[Claude&#39;s load-bearing seams](https://madradavid.com/claudes-load-bearing-seams/)

Oh so good. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/09/24/every-engagement-starts-with-a.html</link>
      <pubDate>Thu, 24 Sep 2026 05:36:37 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/09/24/every-engagement-starts-with-a.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;Every engagement starts with a two-day visit. The team asks business leaders to ignore AI and name the biggest levers in their business. It then works on whichever that turns out to be.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://thenextweb.com/news/openai-colin-jarvis-enterprise-ai-deployment-fde-humanx&#34;&gt;OpenAI’s Colin Jarvis says enterprise AI is stuck on deployment, not models&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Small nonprofits simply don&amp;rsquo;t have the money to do this. I keep thinking about how we bring the methodology we pioneered at Caravan Studios to this particular problem &amp;ndash; identify levers by cohorts and then using them to help move a larger group forward.&lt;/p&gt;
</description>
      <source:markdown>&gt; Every engagement starts with a two-day visit. The team asks business leaders to ignore AI and name the biggest levers in their business. It then works on whichever that turns out to be.

[OpenAI’s Colin Jarvis says enterprise AI is stuck on deployment, not models](https://thenextweb.com/news/openai-colin-jarvis-enterprise-ai-deployment-fde-humanx)

Small nonprofits simply don&#39;t have the money to do this. I keep thinking about how we bring the methodology we pioneered at Caravan Studios to this particular problem -- identify levers by cohorts and then using them to help move a larger group forward. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/09/23/its-not-about-protecting-us.html</link>
      <pubDate>Wed, 23 Sep 2026 13:17:26 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/09/23/its-not-about-protecting-us.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;It&amp;rsquo;s not about protecting us from the unlikely sexy apocalypse. It&amp;rsquo;s about protecting us from the far more likely, and far more numerous, stupid apocalypses along the way.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://www.techdirt.com/2026/09/21/the-apocalypse-will-not-be-sexy/&#34;&gt;The Apocalypse will not be Sexy&lt;/a&gt;&lt;/p&gt;
</description>
      <source:markdown>&gt; It&#39;s not about protecting us from the unlikely sexy apocalypse. It&#39;s about protecting us from the far more likely, and far more numerous, stupid apocalypses along the way.

[The Apocalypse will not be Sexy](https://www.techdirt.com/2026/09/21/the-apocalypse-will-not-be-sexy/)
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/09/19/a-frontier-ai-lab-showed.html</link>
      <pubDate>Sat, 19 Sep 2026 07:09:02 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/09/19/a-frontier-ai-lab-showed.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;A frontier AI lab showed up with a check large enough to make the question &amp;ldquo;should we sell our users&#39; data?&amp;rdquo; feel rhetorical. Not &amp;ldquo;maybe we can anonymize and aggregate.&amp;rdquo; Not &amp;ldquo;only if users opt in.&amp;rdquo; Just: here is the number. The number at which the internal debate ends and the legal team starts drafting the &amp;ldquo;continued use constitutes acceptance&amp;rdquo; language. That number is the new definition of fuck-you money. It is the amount required for a company to accept the lab&amp;rsquo;s terms and then turn around and tell you that if you object, you can fuck off.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://fuck-off.ai/&#34;&gt;Dear Customer, Fuck You&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;I’m just going to leave this right here.&lt;/p&gt;
</description>
      <source:markdown>&gt; A frontier AI lab showed up with a check large enough to make the question &#34;should we sell our users&#39; data?&#34; feel rhetorical. Not &#34;maybe we can anonymize and aggregate.&#34; Not &#34;only if users opt in.&#34; Just: here is the number. The number at which the internal debate ends and the legal team starts drafting the &#34;continued use constitutes acceptance&#34; language. That number is the new definition of fuck-you money. It is the amount required for a company to accept the lab&#39;s terms and then turn around and tell you that if you object, you can fuck off.

[Dear Customer, Fuck You](https://fuck-off.ai/)


I’m just going to leave this right here. 

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/09/14/the-playlist-features-tracks-by.html</link>
      <pubDate>Mon, 14 Sep 2026 05:46:37 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/09/14/the-playlist-features-tracks-by.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;The playlist features tracks by Ornette Coleman, Coltrane, Dizzy, the Bill Evans Trio, Monk, Mingus, and Miles Davis.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://kottke.org/26/09/two-hour-mix-of-the-jazz-in-haruki-murakami-novels&#34;&gt;Two hour mix of the jazz in Haruki Murakami novels&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Oh this is wonderful.&lt;/p&gt;
</description>
      <source:markdown>&gt; The playlist features tracks by Ornette Coleman, Coltrane, Dizzy, the Bill Evans Trio, Monk, Mingus, and Miles Davis.

[Two hour mix of the jazz in Haruki Murakami novels](https://kottke.org/26/09/two-hour-mix-of-the-jazz-in-haruki-murakami-novels)

Oh this is wonderful. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/08/29/next-boring-tiny-tool-ive.html</link>
      <pubDate>Sat, 29 Aug 2026 08:33:46 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/08/29/next-boring-tiny-tool-ive.html</guid>
      <description>&lt;h2 id=&#34;next-boring-tiny-tool&#34;&gt;Next boring tiny tool&lt;/h2&gt;
&lt;p&gt;I’ve been using &lt;a href=&#34;https://epilogue.micro.blog&#34;&gt;Epilogue&lt;/a&gt; for a while to keep track of &lt;a href=&#34;https://crank.report/reading/&#34;&gt;the books I read and finish&lt;/a&gt;, and &lt;a href=&#34;https://crank.report/to-read/&#34;&gt;the big list of books I want to read&lt;/a&gt;. The problem is I never open the too read books when I’m on &lt;a href=&#34;https://www.overdrive.com/apps/libby&#34;&gt;Libby&lt;/a&gt; to check out my next batch of books.&lt;/p&gt;
&lt;p&gt;I did a little bit of reasearch with Gemini and learned I can make a script that can check for availability on Libby or at my branch library and send me an email or drop a note in my Obsidian vault when a book on my list is available. I think I can also add a badge to the page to show available books so I can use the published page itself as a reference.&lt;/p&gt;
&lt;p&gt;Now I’ve got my next &lt;a href=&#34;https://vaughntan.org/boringtinytools&#34;&gt;boring tiny tool&lt;/a&gt;.  I will report back when it is done.&lt;/p&gt;
</description>
      <source:markdown>## Next boring tiny tool

I’ve been using [Epilogue](https://epilogue.micro.blog) for a while to keep track of [the books I read and finish](https://crank.report/reading/), and [the big list of books I want to read](https://crank.report/to-read/). The problem is I never open the too read books when I’m on [Libby](https://www.overdrive.com/apps/libby) to check out my next batch of books. 

I did a little bit of reasearch with Gemini and learned I can make a script that can check for availability on Libby or at my branch library and send me an email or drop a note in my Obsidian vault when a book on my list is available. I think I can also add a badge to the page to show available books so I can use the published page itself as a reference. 

Now I’ve got my next [boring tiny tool](https://vaughntan.org/boringtinytools).  I will report back when it is done. 

</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/08/29/the-problem-from-skipping-laying.html</link>
      <pubDate>Sat, 29 Aug 2026 06:40:31 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/08/29/the-problem-from-skipping-laying.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;The problem from skipping laying out the constraints is that you get a disjoint patchwork that randomly prioritizes some interactions over others. AI exacerbates the draw to wackamole design. It encourages prompting &amp;ldquo;Make X more prominent&amp;rdquo; or &amp;ldquo;Add an affordance to do Y&amp;rdquo;. In the end, more users are confused.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://ref.tools/blog/how-i-design-with-ai&#34;&gt;Ref - How I Design with AI.&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;This is true for so much editing and change.&lt;/p&gt;
</description>
      <source:markdown>&gt; The problem from skipping laying out the constraints is that you get a disjoint patchwork that randomly prioritizes some interactions over others. AI exacerbates the draw to wackamole design. It encourages prompting &#34;Make X more prominent&#34; or &#34;Add an affordance to do Y&#34;. In the end, more users are confused.

[Ref - How I Design with AI.](https://ref.tools/blog/how-i-design-with-ai)

This is true for so much editing and change. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/08/29/so-the-collateral-chain-is.html</link>
      <pubDate>Sat, 29 Aug 2026 06:15:56 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/08/29/so-the-collateral-chain-is.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;So the collateral chain is short. A European lender is ultimately underwriting one American company&amp;rsquo;s continued appetite for compute it does not own. If the lease goes, the chips stay.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://thenextweb.com/news/lambda-1bn-private-debt-microsoft-lease-nebius-european-neocloud&#34;&gt;Lambda inks $1B private debt for Nvidia chip deal&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;This looks like large corps moving the risk.&lt;/p&gt;
</description>
      <source:markdown>&gt; So the collateral chain is short. A European lender is ultimately underwriting one American company&#39;s continued appetite for compute it does not own. If the lease goes, the chips stay.

[Lambda inks $1B private debt for Nvidia chip deal](https://thenextweb.com/news/lambda-1bn-private-debt-microsoft-lease-nebius-european-neocloud)

This looks like large corps moving the risk. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/08/29/modern-coding-agents-have-become.html</link>
      <pubDate>Sat, 29 Aug 2026 05:44:07 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/08/29/modern-coding-agents-have-become.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;Modern coding agents have become so effective at finding flaws that the slightest hint at a new bug can be enough information for them to find it, something Anil has been able to demonstrate using his own agents, switching to DeepSeek V4 Pro⁠ when Claude Fable refused the task.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://simonwillison.net/2026/Aug/28/just-a-rumour-of-a-bug/&#34;&gt;Just a rumor of a bug is enough&lt;/a&gt;&lt;/p&gt;
</description>
      <source:markdown>&gt; Modern coding agents have become so effective at finding flaws that the slightest hint at a new bug can be enough information for them to find it, something Anil has been able to demonstrate using his own agents, switching to DeepSeek V4 Pro⁠ when Claude Fable refused the task.

[Just a rumor of a bug is enough](https://simonwillison.net/2026/Aug/28/just-a-rumour-of-a-bug/)
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/08/17/if-you-expect-your-students.html</link>
      <pubDate>Mon, 17 Aug 2026 20:40:28 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/08/17/if-you-expect-your-students.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;If you expect your students to choose books, ideas and genuine learning over shortcuts and distraction, you must be a bit courageous, too. Return to the basics. Cut through the haze of abstraction, the webs of scholarly terminology, the pretense of professionalization and attempt to recover the basic human experiences that can move us. Remain a student in earnest.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://www.nytimes.com/2026/08/17/opinion/higher-ed-dostoevsky-attention.html&#34;&gt;I let my students watch me struggle with a masterpiece. It changed everything.&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;This kind of engagement is also required of managers.&lt;/p&gt;
</description>
      <source:markdown>&gt; If you expect your students to choose books, ideas and genuine learning over shortcuts and distraction, you must be a bit courageous, too. Return to the basics. Cut through the haze of abstraction, the webs of scholarly terminology, the pretense of professionalization and attempt to recover the basic human experiences that can move us. Remain a student in earnest.

[I let my students watch me struggle with a masterpiece. It changed everything.](https://www.nytimes.com/2026/08/17/opinion/higher-ed-dostoevsky-attention.html)

This kind of engagement is also required of managers. 
</source:markdown>
    </item>
    
    <item>
      <title></title>
      <link>https://crank.report/2026/08/16/good-leaders-do-more-than.html</link>
      <pubDate>Sun, 16 Aug 2026 07:22:18 -0800</pubDate>
      
      <guid>http://crankreport.micro.blog/2026/08/16/good-leaders-do-more-than.html</guid>
      <description>&lt;blockquote&gt;
&lt;p&gt;Good leaders do more than issue instructions. They share context, explain the desired outcome, set boundaries, and respond to what comes back.&lt;/p&gt;
&lt;p&gt;The same habits improve my work with AI. A good prompt helps, but a shared working context helps more. Examples, corrections, and reusable instructions reduce misunderstandings. Over time, the system becomes better aligned with how I think and what I need from it.&lt;/p&gt;
&lt;p&gt;The investment is not in pretending that AI is human. It is in becoming better at expressing intent.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;&lt;a href=&#34;https://allen.bargi.org/notes/working-with-ai-feels-like-leadership/&#34;&gt;Working With AI Feels More Like Leadership Than Coding — Allen Bargi&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;This.&lt;/p&gt;
</description>
      <source:markdown>&gt; Good leaders do more than issue instructions. They share context, explain the desired outcome, set boundaries, and respond to what comes back.
&gt;
&gt; The same habits improve my work with AI. A good prompt helps, but a shared working context helps more. Examples, corrections, and reusable instructions reduce misunderstandings. Over time, the system becomes better aligned with how I think and what I need from it.
&gt;
&gt; The investment is not in pretending that AI is human. It is in becoming better at expressing intent.

[Working With AI Feels More Like Leadership Than Coding — Allen Bargi](https://allen.bargi.org/notes/working-with-ai-feels-like-leadership/)


This.

</source:markdown>
    </item>
    
  </channel>
</rss>
