<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>ai infrastructure: everyone on the record — The Minutes</title>
    <link>https://minutesof.com/on/ai-infrastructure/</link>
    <description>Everyone on the record about ai infrastructure: 20 verbatim quotes from 4 people, January 2024 to August 2026, in one chronology, each with a timestamp…</description>
    <language>en</language>
    <lastBuildDate>Sun, 30 Aug 2026 22:32:46 +0000</lastBuildDate>
    <atom:link href="https://minutesof.com/on/ai-infrastructure/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Gavin Baker: Baker identifies memory as the single most important factor for increasing token output per compu</title>
      <link>https://minutesof.com/q/498c1547-6142-4dfd-9717-055d939ec2fc/</link>
      <guid isPermaLink="true">https://minutesof.com/q/498c1547-6142-4dfd-9717-055d939ec2fc/</guid>
      <description>“It&#x27;s the single most important thing you could do to increase token output per unit of compute, and then that obviously, definitionally, actually lowers costs,” — Gavin Baker, Invest Like the Best</description>
      <pubDate>Tue, 04 Aug 2026 08:00:00 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner asserts Elon Musk is the best in the world at turning electrons into tokens.</title>
      <link>https://minutesof.com/q/1008cf46-972d-4b15-b3aa-45f932f60fb6/</link>
      <guid isPermaLink="true">https://minutesof.com/q/1008cf46-972d-4b15-b3aa-45f932f60fb6/</guid>
      <description>“There&#x27;s no human being on the planet that is better at turning electrons into tokens than Elon Musk. Right? Let me say that again.” — Brad Gerstner, Altimeter Capital</description>
      <pubDate>Fri, 29 May 2026 17:05:40 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner says Musk is building the largest data centers on Earth and eventually in space.</title>
      <link>https://minutesof.com/q/61e8f8d6-f0ae-42ed-aa4e-c33b32a5b543/</link>
      <guid isPermaLink="true">https://minutesof.com/q/61e8f8d6-f0ae-42ed-aa4e-c33b32a5b543/</guid>
      <description>“Elon is building the biggest data centers on Earth and eventually in space of anybody, and nobody is better at turning those electrons into tokens.” — Brad Gerstner, CNBC Television</description>
      <pubDate>Thu, 28 May 2026 17:00:15 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner asserts Elon Musk is the best at converting electrons into AI tokens globally.</title>
      <link>https://minutesof.com/q/387a95cd-d890-4f6a-b70d-3463af3d5021/</link>
      <guid isPermaLink="true">https://minutesof.com/q/387a95cd-d890-4f6a-b70d-3463af3d5021/</guid>
      <description>“There&#x27;s no human being on the planet that is better at turning electrons into tokens than Elon Musk.” — Brad Gerstner, CNBC Television</description>
      <pubDate>Thu, 28 May 2026 17:00:15 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Dylan Patel: Patel notes AI software stacks update constantly with nightly builds making performance measureme</title>
      <link>https://minutesof.com/q/3aab3b19-37a2-4c04-a1ec-5e5018b0ee10/</link>
      <guid isPermaLink="true">https://minutesof.com/q/3aab3b19-37a2-4c04-a1ec-5e5018b0ee10/</guid>
      <description>“with software changing literally multiple times a week, right, PyTorch has nightlies, VLM has nightlies, Asteeling has nightlies, CUDA drivers update constantly.” — Dylan Patel, TensorWave</description>
      <pubDate>Thu, 30 Apr 2026 18:18:14 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner contrasts today&#x27;s GPU utilization with 2000 when only 7% of fiber was utilized.</title>
      <link>https://minutesof.com/q/548fba38-3666-49ec-a5fa-dfd869fc6448/</link>
      <guid isPermaLink="true">https://minutesof.com/q/548fba38-3666-49ec-a5fa-dfd869fc6448/</guid>
      <description>“Remember back in 2000, we talked a lot about dark fiber. Only 7% of the fiber going into the ground in the year 2000 was being utilized. 7%.” — Brad Gerstner, Altimeter Capital</description>
      <pubDate>Tue, 03 Feb 2026 21:01:23 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner says Databricks and Snowflake are accelerating because they provide the data substrate</title>
      <link>https://minutesof.com/q/28161199-1201-4d20-bcda-aa3078c6c6e7/</link>
      <guid isPermaLink="true">https://minutesof.com/q/28161199-1201-4d20-bcda-aa3078c6c6e7/</guid>
      <description>“But there are other companies that we own and love, like Databricks and Snowflake, that are actually accelerating. Why are they accelerating? They are the data substrate upon which enterprises are building agentic solutions.” — Brad Gerstner, CNBC Television</description>
      <pubDate>Tue, 06 Jan 2026 18:27:14 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Tech companies are investing over the course of the next four or five years, it&#x27;s about 10 time</title>
      <link>https://minutesof.com/q/01ffca9f-8e11-4998-80d2-adc9447c7130/</link>
      <guid isPermaLink="true">https://minutesof.com/q/01ffca9f-8e11-4998-80d2-adc9447c7130/</guid>
      <description>“Tech companies are investing over the course of the next four or five years, it&#x27;s about 10 times the size of the Manhattan Project on an inflation adjusted or GDP adjusted basis.” — Brad Gerstner, BG2 Pod</description>
      <pubDate>Fri, 31 Oct 2025 23:11:44 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley flags NVIDIA&#x27;s promise to buy unsold CoreWeave capacity as highly unusual arrangement.</title>
      <link>https://minutesof.com/q/f01ef787-0218-496c-9b10-9c8217b3118d/</link>
      <guid isPermaLink="true">https://minutesof.com/q/f01ef787-0218-496c-9b10-9c8217b3118d/</guid>
      <description>“One of the more peculiar of all the deals is, and this was disclosed in a CoreWeave filing, was NVIDIA has promised to buy any of CoreWeave&#x27;s service availability that they can&#x27;t sell to anyone else.” — Bill Gurley, Bg2 Pod</description>
      <pubDate>Tue, 14 Oct 2025 23:56:17 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner says the UAE-US AI campus announced is five gigawatts, equivalent to 2.5 million GPUs.</title>
      <link>https://minutesof.com/q/80b86fdf-a612-4f6e-84b0-ea3470e3aaa7/</link>
      <guid isPermaLink="true">https://minutesof.com/q/80b86fdf-a612-4f6e-84b0-ea3470e3aaa7/</guid>
      <description>“So in in Abu Dhabi, they announced a five gigawatt US AI you know, it&#x27;s a UAE, US AI campus. It&#x27;s an incredible architectural campus.” — Brad Gerstner, BG2 Pod</description>
      <pubDate>Thu, 22 May 2025 16:23:05 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley says Stargate will be a property-owner and operating-company structure with one owning and</title>
      <link>https://minutesof.com/q/df4f224d-99fa-44a3-a351-2c11e7f42795/</link>
      <guid isPermaLink="true">https://minutesof.com/q/df4f224d-99fa-44a3-a351-2c11e7f42795/</guid>
      <description>“People are telling me it&#x27;s gonna be a prop co op co structure, which I didn&#x27;t even know it was twenty four hours ago, but it&#x27;s it&#x27;s like a one person owns the property.” — Bill Gurley, CNBC Television</description>
      <pubDate>Fri, 24 Jan 2025 21:30:16 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley notes TSMC announced 20% across-the-board price increase as expected proof point of supply</title>
      <link>https://minutesof.com/q/c6c1ad8c-5c13-4f2b-b47b-f745e6942016/</link>
      <guid isPermaLink="true">https://minutesof.com/q/c6c1ad8c-5c13-4f2b-b47b-f745e6942016/</guid>
      <description>“Another thing that came out this morning that I was kind of waiting on as a proof point, because it didn&#x27;t make sense to me that it hadn&#x27;t happened yet, TSMC said they&#x27;re gonna do across the board 20 price increase.” — Bill Gurley, BG2 Pod</description>
      <pubDate>Thu, 31 Oct 2024 17:46:51 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner says Jensen confirmed GPU clusters are already at 200-300k scale and will reach 500k t</title>
      <link>https://minutesof.com/q/0442e975-6387-46e1-861e-c0e1580fb267/</link>
      <guid isPermaLink="true">https://minutesof.com/q/0442e975-6387-46e1-861e-c0e1580fb267/</guid>
      <description>“I said, are we already at the phase of two and three hundred thousand cluster scale? And he said, yes. And then I said, and will we go to 500,000 a million?” — Brad Gerstner, BG2 Pod</description>
      <pubDate>Sun, 13 Oct 2024 10:35:00 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner proposes dedicating individual nuclear reactors at Diablo to major tech companies for</title>
      <link>https://minutesof.com/q/87f2bdd7-b1c0-4af8-836a-2b8fcf3ab06b/</link>
      <guid isPermaLink="true">https://minutesof.com/q/87f2bdd7-b1c0-4af8-836a-2b8fcf3ab06b/</guid>
      <description>“I could imagine if you had four more reactors sitting here, right, if you could land them at the right price, you could have one reactor for Meta, one for Amazon, one for Microsoft, one for Nvidia, you could have the data center sitting right next to them.” — Brad Gerstner, BG2 Pod</description>
      <pubDate>Mon, 09 Sep 2024 02:12:29 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Gavin Baker: Baker proposes a new metric decomposing MFU that he developed the morning of the interview.</title>
      <link>https://minutesof.com/q/27c177e8-03c2-4627-956e-295dd3a471e4/</link>
      <guid isPermaLink="true">https://minutesof.com/q/27c177e8-03c2-4627-956e-295dd3a471e4/</guid>
      <description>“MFU is the most important metric because it gives you all of these advantages and ways to differentiate yourself amongst five people, six people who&#x27;ve trained these GPT core class models.” — Gavin Baker, Invest Like the Best</description>
      <pubDate>Tue, 27 Aug 2024 08:00:00 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Gavin Baker: Baker explains that 50% MFU versus 40% MFU enables 25% faster time to market for AI models.</title>
      <link>https://minutesof.com/q/b18ea10e-8f3d-44a4-afda-d9363c3a12ee/</link>
      <guid isPermaLink="true">https://minutesof.com/q/b18ea10e-8f3d-44a4-afda-d9363c3a12ee/</guid>
      <description>“You could choose between faster time to market. If you run a 50% MFU and your competitor&#x27;s running 40 for an equivalent amount of trading flops, you could be in market 25% faster.” — Gavin Baker, Invest Like the Best</description>
      <pubDate>Tue, 27 Aug 2024 08:00:00 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Gavin Baker: Baker explains higher MFU allows 25% faster time to market with same GPU and power spending.</title>
      <link>https://minutesof.com/q/2ef0281e-1acf-4b2f-a09a-427746b64fa5/</link>
      <guid isPermaLink="true">https://minutesof.com/q/2ef0281e-1acf-4b2f-a09a-427746b64fa5/</guid>
      <description>“You have the same amount of GPUs and the same amount of power presumably. You could choose between faster time to market.” — Gavin Baker, Invest Like the Best</description>
      <pubDate>Tue, 27 Aug 2024 08:00:00 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Gavin Baker: Baker explains MFU runs at 35-40%, measuring the percentage of theoretical compute actually used</title>
      <link>https://minutesof.com/q/91f8d62c-9179-4722-83b1-a80397265ed3/</link>
      <guid isPermaLink="true">https://minutesof.com/q/91f8d62c-9179-4722-83b1-a80397265ed3/</guid>
      <description>“MFU, model flops utilization, and that generally runs around 35 to 40%. And that&#x27;s literally the percentage of compute, theoretical compute flops that you&#x27;re actually applying to trading.” — Gavin Baker, Invest Like the Best</description>
      <pubDate>Tue, 27 Aug 2024 08:00:00 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Bill Gurley: Gurley notes hyperscalers now spending $200 billion in total CapEx with questions about return on</title>
      <link>https://minutesof.com/q/743c4b3d-d172-40cc-bd69-aa8649d5a14e/</link>
      <guid isPermaLink="true">https://minutesof.com/q/743c4b3d-d172-40cc-bd69-aa8649d5a14e/</guid>
      <description>“In fact, I think if people are gonna be critical of anything, a lot of people are looking at the total CapEx of the hyperscalers now at $200,000,000,000 and saying, when are you gonna get a return on the dollars that you&#x27;re spending?” — Bill Gurley, Bg2 Pod</description>
      <pubDate>Mon, 10 Jun 2024 14:22:07 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
    <item>
      <title>Brad Gerstner: Gerstner highlights Super Micro raised full-year guidance by 40% to $14 billion.</title>
      <link>https://minutesof.com/q/584d2a5d-006f-487e-8ec5-9d97132a3343/</link>
      <guid isPermaLink="true">https://minutesof.com/q/584d2a5d-006f-487e-8ec5-9d97132a3343/</guid>
      <description>“Super Micro took their full year guide up after the bell today by 40%. 40% to $14,000,000,000. So that&#x27;s why you&#x27;re seeing these moves.” — Brad Gerstner, CNBC Television</description>
      <pubDate>Mon, 29 Jan 2024 23:20:37 +0000</pubDate>
      <category>ai infrastructure</category>
    </item>
  </channel>
</rss>
