<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Business on Robert Bost</title>
    <link>https://bost.dev/tags/business/</link>
    <description>Recent content in Business on Robert Bost</description>
    <generator>Hugo -- 0.147.7</generator>
    <language>en-us</language>
    <copyright>2025 Robert Bost</copyright>
    <lastBuildDate>Mon, 13 Jul 2026 10:42:33 -0600</lastBuildDate>
    <atom:link href="https://bost.dev/tags/business/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>When Answers Have Zero Latency</title>
      <link>https://bost.dev/posts/when-answers-have-zero-latency/</link>
      <pubDate>Mon, 13 Jul 2026 10:42:33 -0600</pubDate>
      <guid>https://bost.dev/posts/when-answers-have-zero-latency/</guid>
      <description>&lt;p&gt;What happens when an AI can give us a good answer with effectively zero latency?&lt;/p&gt;
&lt;p&gt;I do not mean literally zero milliseconds. Physics will continue to be annoying. I mean that the time between asking a question and getting a useful answer becomes too small to matter. The answer arrives quickly enough that waiting for it no longer affects what we do next.&lt;/p&gt;
&lt;p&gt;The obvious consequence is faster generation. Emails, code, plans, product ideas, contracts, and analysis all appear nearly instantly. But generation is only one part of the change. An AI can also evaluate its answer, find weaknesses, produce alternatives, test them, and repeat the process. If each pass approaches zero latency, then the entire loop begins to collapse:&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
