<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>scaling: everyone on the record — The Minutes</title>
    <link>https://minutesof.com/on/scaling/</link>
    <description>Everyone on the record about scaling: 22 verbatim quotes from 6 people, December 2012 to May 2026, in one chronology, each with a timestamp and a link to…</description>
    <language>en</language>
    <lastBuildDate>Sun, 30 Aug 2026 22:33:14 +0000</lastBuildDate>
    <atom:link href="https://minutesof.com/on/scaling/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>David Friedberg: Friedberg describes Koch Industries&#x27; 9,000X scaling as algorithmic compounding through contin</title>
      <link>https://minutesof.com/q/4d5e6c49-6cb6-4dbc-ba69-76e40fb3d3ad/</link>
      <guid isPermaLink="true">https://minutesof.com/q/4d5e6c49-6cb6-4dbc-ba69-76e40fb3d3ad/</guid>
      <description>“The success at Coke Industries is built on an algorithm, your principles, and you&#x27;ve been able to adapt that algorithm and as a result scale your business, generate cash, reinvest that cash, generate more cash, reinvest that cash and so on. And you&#x27;ve scaled 9,000 X. You&#x27;ve built a compounding advantage in your business.” — David Friedberg, All-In Podcast</description>
      <pubDate>Tue, 12 May 2026 21:17:30 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel estimates Anthropic needs to reach well above five gigawatts of compute capacity by year-en</title>
      <link>https://minutesof.com/q/dde6295c-09fa-47d0-9fb4-f6266575281d/</link>
      <guid isPermaLink="true">https://minutesof.com/q/dde6295c-09fa-47d0-9fb4-f6266575281d/</guid>
      <description>“Anthropic needs to get to well above five gigawatts by the end of this year, and it&#x27;s gonna be really tough for them to get there, but it&#x27;s possible.” — Dylan Patel, Dwarkesh Patel</description>
      <pubDate>Fri, 13 Mar 2026 16:26:57 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Jensen Huang: Huang contrasts Moore&#x27;s law (100x per decade) with current AI acceleration (1 million times per</title>
      <link>https://minutesof.com/q/31d24717-9675-403b-a23d-dcc1924334ea/</link>
      <guid isPermaLink="true">https://minutesof.com/q/31d24717-9675-403b-a23d-dcc1924334ea/</guid>
      <description>“Remember, Moore&#x27;s law was two times every 18, 10 times every five years, a 100 times every ten. Okay? But where are we now? A million times every ten years.” — Jensen Huang, Cisco</description>
      <pubDate>Wed, 04 Feb 2026 04:34:47 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel reports Microsoft aims to 10x training capacity every 18-24 months, representing a 10x incr</title>
      <link>https://minutesof.com/q/fa0ed250-859f-4800-a571-4fb9a5440edb/</link>
      <guid isPermaLink="true">https://minutesof.com/q/fa0ed250-859f-4800-a571-4fb9a5440edb/</guid>
      <description>“We try to 10x the training capacity every eighteen to twenty four months. And so this would be effectively a 10x increase. 10x from what GPD five was trained with.” — Dylan Patel, Dwarkesh Patel</description>
      <pubDate>Wed, 12 Nov 2025 17:02:10 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner argues AI is scaling faster than the Internet did, citing his investments in Google an</title>
      <link>https://minutesof.com/q/ac7d417d-7577-44b2-9028-5284d3cf66b7/</link>
      <guid isPermaLink="true">https://minutesof.com/q/ac7d417d-7577-44b2-9028-5284d3cf66b7/</guid>
      <description>“ChatGPT became the verb and I would argue, maybe just So, you know, if we go to this slide right here, You know, AI is scaling way faster than the Internet did.” — Brad Gerstner, Khosla Ventures</description>
      <pubDate>Wed, 10 Sep 2025 22:51:48 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel states $10 billion data centers target automated software engineering, not chat models.</title>
      <link>https://minutesof.com/q/119203f3-2761-4bf3-a619-aaad40c2f71a/</link>
      <guid isPermaLink="true">https://minutesof.com/q/119203f3-2761-4bf3-a619-aaad40c2f71a/</guid>
      <description>“So no one is trying to make with these $10,000,000,000 data centers, they&#x27;re not trying to make chat models. Right?” — Dylan Patel, Alex Kantrowitz</description>
      <pubDate>Wed, 23 Apr 2025 16:30:06 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel argues that efficiency gains alone without capability increases cannot justify massive AI i</title>
      <link>https://minutesof.com/q/675ee285-ed22-4889-b6cf-5f6b6892d5c7/</link>
      <guid isPermaLink="true">https://minutesof.com/q/675ee285-ed22-4889-b6cf-5f6b6892d5c7/</guid>
      <description>“That would not pay for all of these build outs. Right? AI is useful today, but it&#x27;s not capable of doing a lot of things.” — Dylan Patel, Alex Kantrowitz</description>
      <pubDate>Wed, 23 Apr 2025 16:30:06 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel says current models use 100,000 GPUs while next generation will require hundreds of thousan</title>
      <link>https://minutesof.com/q/561bd369-2c88-4ea1-8fa4-3d02389be50d/</link>
      <guid isPermaLink="true">https://minutesof.com/q/561bd369-2c88-4ea1-8fa4-3d02389be50d/</guid>
      <description>“And next generation models that are trained on hundreds of thousands or even millions GPUs, right?” — Dylan Patel, Special Competitive Studies Project</description>
      <pubDate>Thu, 13 Mar 2025 16:42:15 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Or or that&#x27;s what it looks like based on or, you know, one gigawatt in twenty twenty six ish. And</title>
      <link>https://minutesof.com/q/eb6b2094-e7b7-4ac7-9761-c8cc9c5e0736/</link>
      <guid isPermaLink="true">https://minutesof.com/q/eb6b2094-e7b7-4ac7-9761-c8cc9c5e0736/</guid>
      <description>“Or or that&#x27;s what it looks like based on or, you know, one gigawatt in twenty twenty six ish. And Meta&#x27;s Meta&#x27;s, like, trying to do, like, two gigawatts by early to mid twenty seven.” — Dylan Patel, Unsupervised Learning: With Jacob Effron</description>
      <pubDate>Tue, 21 Jan 2025 14:00:12 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel calculates next-generation clusters deliver 15x more compute through five times more GPUs a</title>
      <link>https://minutesof.com/q/badb57b8-ea1b-45cb-887a-c8b1bba9f5d8/</link>
      <guid isPermaLink="true">https://minutesof.com/q/badb57b8-ea1b-45cb-887a-c8b1bba9f5d8/</guid>
      <description>“So you got you have five x the GPUs, and you have three x the performance per GPU roughly. So then you&#x27;re at, like, 15 x more compute.” — Dylan Patel, Unsupervised Learning: With Jacob Effron</description>
      <pubDate>Tue, 21 Jan 2025 14:00:12 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Bill Gurley: You know, I can go from 10,000,000 to 100,000,000 to billion to $10,000,000,000 on reasoning in s</title>
      <link>https://minutesof.com/q/62ff4fe1-86c3-4195-b4f5-3a55efd51a21/</link>
      <guid isPermaLink="true">https://minutesof.com/q/62ff4fe1-86c3-4195-b4f5-3a55efd51a21/</guid>
      <description>“You know, I can go from 10,000,000 to 100,000,000 to billion to $10,000,000,000 on reasoning in such a quick succession.” — Bill Gurley, Bg2 Pod</description>
      <pubDate>Mon, 23 Dec 2024 20:25:20 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley notes AI leaders previously implied linear scaling gains but current reality differs from</title>
      <link>https://minutesof.com/q/9e8745ec-5c54-4385-accb-630672e0cc36/</link>
      <guid isPermaLink="true">https://minutesof.com/q/9e8745ec-5c54-4385-accb-630672e0cc36/</guid>
      <description>“They had left that impression. And so we get to this place as you described it, it&#x27;s not quite like that.” — Bill Gurley, BG2 Pod</description>
      <pubDate>Mon, 23 Dec 2024 20:12:56 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel calculates that multi-trillion parameter models require transmitting 40 terabytes of data e</title>
      <link>https://minutesof.com/q/2b1dd77d-4b29-490c-a1f9-fd32caff427b/</link>
      <guid isPermaLink="true">https://minutesof.com/q/2b1dd77d-4b29-490c-a1f9-fd32caff427b/</guid>
      <description>“They&#x27;re doing it for like multi trillion. Right? So let&#x27;s call it 10,000,000,000,000 parameters, four bytes parameter, that&#x27;s 40 terabytes of data you need to transmit in two seconds.” — Dylan Patel, Scaling Intelligence</description>
      <pubDate>Tue, 12 Nov 2024 04:17:00 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel says GPT-4 used 24,000 GPUs, GPT-5 uses 100,000, and Microsoft is building toward a million</title>
      <link>https://minutesof.com/q/49107836-7a8a-4175-851d-c6bc483f45a1/</link>
      <guid isPermaLink="true">https://minutesof.com/q/49107836-7a8a-4175-851d-c6bc483f45a1/</guid>
      <description>“GPT four was trained with 24,000 GPUs roughly, and GPT five is on the order of a 100,000. And then they&#x27;re trying to build this data center over the next few years. That&#x27;s a million.” — Dylan Patel, Scaling Intelligence</description>
      <pubDate>Tue, 12 Nov 2024 04:17:00 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel identifies distributed training across data centers with lower bandwidth as a key unsolved</title>
      <link>https://minutesof.com/q/8bdd8e52-1eff-46b4-82af-03f48966482c/</link>
      <guid isPermaLink="true">https://minutesof.com/q/8bdd8e52-1eff-46b4-82af-03f48966482c/</guid>
      <description>“Everything that we&#x27;ve seen so far is that large scale training has to happen in an individual data center with very high speed networking.” — Dylan Patel, Latent Space</description>
      <pubDate>Tue, 05 Dec 2023 18:16:28 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Dylan Patel: Patel predicts open source will match GPT-4 but frontier models will continue advancing beyond it</title>
      <link>https://minutesof.com/q/17074139-7de1-4dd4-9c7f-5a1a0f39e4f8/</link>
      <guid isPermaLink="true">https://minutesof.com/q/17074139-7de1-4dd4-9c7f-5a1a0f39e4f8/</guid>
      <description>“Open source will match GPT-four, but then it&#x27;s like, what about GPT-four Vision? Or what about five and six and all these kind of stuff?” — Dylan Patel, Latent Space</description>
      <pubDate>Tue, 05 Dec 2023 18:16:28 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>David Friedberg: Friedberg says Climate Corp&#x27;s average data use increased 40x every year for seven consecutive</title>
      <link>https://minutesof.com/q/20708ae0-c5ee-4bdc-9dd6-a71474cb6aa5/</link>
      <guid isPermaLink="true">https://minutesof.com/q/20708ae0-c5ee-4bdc-9dd6-a71474cb6aa5/</guid>
      <description>“In the seven years we ran Climate Corp, by the way, we did a study that our average data use per year increased by 40x every year for seven years.” — David Friedberg, AgFunder</description>
      <pubDate>Mon, 02 Oct 2023 13:03:03 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner argues AWS, GCP and Azure enable software companies to scale faster and more capital e</title>
      <link>https://minutesof.com/q/44e49c41-c433-42d3-b373-5aa3215c29a8/</link>
      <guid isPermaLink="true">https://minutesof.com/q/44e49c41-c433-42d3-b373-5aa3215c29a8/</guid>
      <description>“But the fact of the matter is because of AWS, GCP and Azure, software companies are scaling faster and in a more capital efficient way” — Brad Gerstner, Invest Like the Best</description>
      <pubDate>Tue, 23 Jun 2020 09:30:00 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Gavin Baker: Baker cites research showing AI quality doubles with every 10x increase in training data quantity</title>
      <link>https://minutesof.com/q/585e22b4-452b-4998-98ed-1f84f746f495/</link>
      <guid isPermaLink="true">https://minutesof.com/q/585e22b4-452b-4998-98ed-1f84f746f495/</guid>
      <description>“All that matters is the quantity of data, and it&#x27;s been very well established in multiple papers from both Google and Microsoft research that for every order of magnitude increase in the data you use to train an algorithm, the quality of the AI doubles.” — Gavin Baker, Invest Like the Best</description>
      <pubDate>Tue, 26 Nov 2019 10:30:00 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley says marketplace success requires doing unscalable things early, which frustrates most bus</title>
      <link>https://minutesof.com/q/07becb42-0cc5-4f1f-8a17-9e9b93b04269/</link>
      <guid isPermaLink="true">https://minutesof.com/q/07becb42-0cc5-4f1f-8a17-9e9b93b04269/</guid>
      <description>“And if you took 90% of the entrepreneurs that have been to business school and understand scaling, it drives them nuts.” — Bill Gurley, Invest Like the Best</description>
      <pubDate>Tue, 02 Jul 2019 09:30:00 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley argues scaling requires capital and expertise even if starting a company is cheap.</title>
      <link>https://minutesof.com/q/6646e3b5-8838-46c7-bdfd-1edf67786248/</link>
      <guid isPermaLink="true">https://minutesof.com/q/6646e3b5-8838-46c7-bdfd-1edf67786248/</guid>
      <description>“I think has come out of it is that you may be able to start a company for nothing but if you want to scale a company you will need capital and you&#x27;ll need expertise to scale it to the next level. Not every single player, Benioff who got it up without venture” — Bill Gurley, GigaOm</description>
      <pubDate>Tue, 11 Dec 2012 17:49:11 +0000</pubDate>
      <category>scaling</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley argues that while starting is cheap, scaling a company requires both capital and expertise</title>
      <link>https://minutesof.com/q/c7010802-6090-473f-ba44-eca7f2f1e71a/</link>
      <guid isPermaLink="true">https://minutesof.com/q/c7010802-6090-473f-ba44-eca7f2f1e71a/</guid>
      <description>“And what I think has come out of it is that you may be able to start a company for nothing but if you want to scale a company you will need capital and you&#x27;ll need expertise to scale it to the next level.” — Bill Gurley, GigaOm</description>
      <pubDate>Tue, 11 Dec 2012 17:49:11 +0000</pubDate>
      <category>scaling</category>
    </item>
  </channel>
</rss>
