<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/">
    <channel>
        <title>ap</title>
        <link>https://paragraph.com/@apcodes</link>
        <description>i normally write about the intersection of ai and crypto.</description>
        <lastBuildDate>Thu, 23 Apr 2026 12:30:58 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <language>en</language>
        <image>
            <title>ap</title>
            <url>https://storage.googleapis.com/papyrus_images/583eadb68c1ac37ea8b31cdac31352b8e1848364209889d0e58118a8f727ceda.jpg</url>
            <link>https://paragraph.com/@apcodes</link>
        </image>
        <copyright>All rights reserved</copyright>
        <item>
            <title><![CDATA[life is short.
remember to change bias when new information hits the market.]]></title>
            <link>https://paragraph.com/@apcodes/life-is-short-remember-to-change-bias-when-new-information-hits-the-market</link>
            <guid>K3y3ahPBbfYwlv3X2gSS</guid>
            <pubDate>Fri, 31 May 2024 18:35:36 GMT</pubDate>
            <description><![CDATA[had a free time this morning so wrote a about scalping the market. enjoy. life is short. remember to change bias when new information hits the market. yes I know its amazing how physically exhausting it can be to do nothing or that the guy who sits next trump has reminded us to act upon your iq, rule is simple as this:low iq players will always outperform mid iq participants, this totally relates to the natural tendency of mid players to OVERTHINK. so if you are in this dead zone pls gtfo.wha...]]></description>
            <content:encoded><![CDATA[<p>had a free time this morning so wrote a about scalping the market. enjoy.</p><p>life is short. remember to change bias when new information hits the market. yes I know its amazing how physically exhausting it can be to do nothing or that the guy who sits next trump has reminded us to act upon your iq, rule is simple as this:</p><ul><li><p>low iq players will always outperform mid iq participants, this totally relates to the natural tendency of mid players to OVERTHINK. so if you are in this dead zone pls gtfo.</p></li><li><p>what do low and high iq players have in common? they both can stomach the volatility of high betas, as simple as this, some of the best guys that i know doing the research work-load in crypto, will simply lack the will to size up and stomach volatility, this trait is not acquired, you are born with it or not.</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/10969c1456ddcb6441717061112255be5fefbfe45c4fdfc3e8c904cdbfa76e65.jpg" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p><strong>conviction &gt; capital</strong></p><p>the primary distinction between left and right side of the bell curve resides mainly in positioning and time effort, pareto rule, 80% of our outcomes come from 20% of our actions, this principle can be seen in the majority of fields.</p><p>do not try to find a bunch of early plays, focus on high conviction plays and let it roll. focus on plays that are beginning to gain traction and hop onto right before they reach escape velocity, do not be irrational, if market conditions are indicative of a bubble formation, add fuel to the fire.</p><p>what does this mean? high iq feel comfortable missing the initial gradual phase of price appreciation which tend to go on over months, rather they focus on capitalizing on the final exponential price discovery phase. DONT DILLUTE YOUR ATTENTION on multiples plays.</p><p><strong>most of the gains for new tokens are now being captured privately</strong></p><p>this is not new, if you want to dig deeper just go and read cobie substack, technology markets are rapidly evolving. this is evident in disruptive technologies like ai, which have received more investment in the past five years than in all previous years combine, crypto does not deviate from this we have seen a 22x inflows into the industry in the last five years, with a current 2.7t market cap.</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/29e40f49cd6bbbbcf78a49c61b5c4cf9d40e3005cad61d70aeeaf1205e138b33.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>the side effects of this order magnitude shift are mainly predatory and abusive behaviors from vcs, in the modern era majority of assets face and off-makrket price-discovery, captured privately before the tge, yes private markets are taking the whole upside.</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/867ba1c4278d7874f973a1e6613fb2fec256ef4d08a0eec61a1adc2f5355750d.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>market participants should expect that new launch valuations remain high while the market is in demand. it is no longer possible to be “early” in the privatised-gains meta — instead, investors should focus on finding value in the market that others have forgotten or has become mispriced and out of favour.</p><p>token buyers should aim to become more sophisticated at evaluating the valuation and supply/demand dynamics of new tokens, identifying which high fdvs are based in supply/demand realities, and which have extremely dislocated phantom markets.</p><p>opting out of participating in these markets is voting with capital.</p><p>good founders want to build successful projects and they know that market dynamics will interact with their project’s perception. memecoin over-performance and new token launch under-performance have caused readjustments in fundraising and launch plans for future founders.</p>]]></content:encoded>
            <author>apcodes@newsletter.paragraph.com (ap)</author>
        </item>
        <item>
            <title><![CDATA[My AI thesis]]></title>
            <link>https://paragraph.com/@apcodes/my-ai-thesis</link>
            <guid>0rfwfvhXFMwDJFsMQ7e1</guid>
            <pubDate>Fri, 01 Mar 2024 19:26:12 GMT</pubDate>
            <description><![CDATA[Semiconductors are at the heart of the AI revolution, much like uranium was critical for nuclear energy. They power the complex computations that AI systems rely on. In the realm of AI, particularly with behemoths like Microsoft, there are two layers worth focusing on:The App Layer: This is the user-facing side, where innovations like OpenAI live. It&apos;s where the applications interact with users and businesses, drawing significant investment. However, picking the winning app is akin to fi...]]></description>
            <content:encoded><![CDATA[<p>Semiconductors are at the heart of the AI revolution, much like uranium was critical for nuclear energy. They power the complex computations that AI systems rely on.</p><p>In the realm of AI, particularly with behemoths like Microsoft, there are two layers worth focusing on:</p><ol><li><p><strong>The App Layer</strong>: This is the user-facing side, where innovations like OpenAI live. It&apos;s where the applications interact with users and businesses, drawing significant investment. However, picking the winning app is akin to finding a needle in a haystack.</p></li><li><p><strong>The Compute Layer</strong>: Represented by entities such as CoreWeave, this layer is all about the silicon that powers AI&apos;s operations. Investments here are into the infrastructure essential for AI&apos;s data processing capabilities.</p></li></ol><p>There&apos;s also a niche for developers, with platforms like Lambda and Render providing the tools and environments for AI development.</p><p>Investing in the app layer is enticing but fraught with uncertainty due to the intense competition. On the other hand, the compute layer represents a foundational investment. It&apos;s essential for AI&apos;s functionality and growth, making it a more stable and strategic investment choice. As AI becomes increasingly integral to our technological landscape, the compute layer&apos;s role as the backbone of AI&apos;s capabilities cannot be overstated.</p><h2 id="h-capex-receiver-thesis" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>CapEx Receiver Thesis</strong></h2><h3 id="h-backdrop" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0"><strong>Backdrop</strong></h3><p><strong>A race is underway</strong> among the world&apos;s tech giants—Google, Microsoft, Amazon, Apple, and others. They&apos;re all vying to dominate the next UI Form Factor, a paradigm that historically shifts every decade. From the 1990s browser to the 2000s Web 2.0 applications, and then to the 2010s smartphones, each era has seen a groundbreaking change in how users access information. The last major shift was to mobile platforms, initiated by Blackberry and epitomized by the iPhone.</p><p>Now, we&apos;re at the cusp of another monumental shift, with AI at its core. The future form factor might be something as innovative as Apple Vision Pro or Rayban Sunglasses—the precise nature of it remains uncertain.</p><p><strong>Why does this matter?</strong></p><p>The stakes are incredibly high, with these corporations poised to invest billions in CapEx, particularly in GPU chips, aiming to spearhead the AI-driven platform evolution.</p><p>This investment frenzy is mirrored on a global scale, exemplified by China&apos;s massive semiconductor imports surpassing even its oil imports. China&apos;s venture funding and deals in semiconductor startups now outpace those in the United States, highlighting the strategic importance of semiconductors in the AI era. Moreover, AI technologies are beginning to command a significant portion of defense budgets, with hive defense and attack models emerging as critical components of national security strategies.</p><p><strong>Countries join the fray</strong>, leveraging their substantial financial resources and relative price insensitivity to secure a competitive edge in AI, deemed essential for national success.</p><p><strong>The lesson from history</strong> suggests that betting on the &quot;Application Layer&quot; is akin to guessing the victor in the early search engine wars—Lycos, Alta Vista, Excite, only for Google to emerge later with a technological breakthrough in PageRank.</p><p><strong>The strategic pivot</strong> should thus be towards the &quot;Compute Layer&quot; or the Silicon Layer. Investing in this foundational technology offers a more reliable path to capitalizing on the AI-driven future.</p><p><strong>Case in point: Microsoft</strong></p><p>Recognizing the strategic importance of controlling its technological destiny, Microsoft is developing its GPU chips for cloud infrastructure. This move is aimed at reducing its dependence on external suppliers like Nvidia. Microsoft&apos;s vision encompasses leveraging its cloud solutions and AI capabilities to enhance data analysis and drive innovation, illustrating a broader industry trend of investing in the compute layer as the bedrock of future technologies.</p><p><strong>Case in point: Nvidia</strong></p><p>Nvidia&apos;s market valuation, with a forward PE comparable to Apple&apos;s, reflects its rapid earnings growth and pivotal role in the AI sector. Unlike traditional tech giants, Nvidia&apos;s strategy underlines a crucial tenet: the indispensability of proprietary data and the inherent risks of outsourcing AI capabilities. Nvidia posits, &quot;Your data is proprietary. You can’t trust outsourcing the AI to anyone else, much less using Microsoft’s co-pilot. Every company and government needs its own co-pilot,&quot; advocating for the necessity of bespoke AI models for every entity.</p><p>Nvidia&apos;s core contribution to the AI revolution isn&apos;t through direct cloud AI hosting services but through strategic partnerships with companies like CoreWeave and Oracle. This collaborative approach ensures that Nvidia&apos;s cutting-edge computing power reaches end-users effectively, emphasizing its role not just as a supplier but as the very backbone of the compute layer. By prioritizing GPU deliveries to CoreWeave, Nvidia demonstrates its commitment to empowering high-performance AI computing, positioning itself as an integral component of the compute infrastructure necessary for AI development and deployment.</p><p>Nvidia, therefore, doesn&apos;t just invest in the compute layer; it <em>is</em> the compute layer. Its approach transcends mere hardware provision, embodying the critical infrastructure that enables the growth and sophistication of AI technologies. This positioning not only secures Nvidia&apos;s place in the tech ecosystem but also highlights the strategic importance of the compute layer as the foundation upon which the future of AI is being built.</p><p>By focusing on enabling customized, powerful AI solutions through its technological prowess and strategic partnerships, Nvidia is at the forefront of the shift towards a more personalized, secure, and efficient AI future. Its actions and strategies underscore the compute layer&apos;s vital role in the ongoing technological evolution, making it a key player in shaping the AI-driven landscape.</p>]]></content:encoded>
            <author>apcodes@newsletter.paragraph.com (ap)</author>
        </item>
    </channel>
</rss>