<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Bert Van Kerkhoven on Medium]]></title>
        <description><![CDATA[Stories by Bert Van Kerkhoven on Medium]]></description>
        <link>https://medium.com/@vankerkhovenbert?source=rss-acc377fbefd6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*hY2u7OCSEWsg4Y09BwiuvA@2x.jpeg</url>
            <title>Stories by Bert Van Kerkhoven on Medium</title>
            <link>https://medium.com/@vankerkhovenbert?source=rss-acc377fbefd6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 11:11:17 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@vankerkhovenbert/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Decentralizing the cloud: Acurast’s bet on smartphone-powered compute]]></title>
            <link>https://medium.com/mn-capital/decentralizing-the-cloud-acurasts-bet-on-smartphone-powered-compute-9c8c7162d005?source=rss-acc377fbefd6------2</link>
            <guid isPermaLink="false">https://medium.com/p/9c8c7162d005</guid>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[depin]]></category>
            <category><![CDATA[acurast]]></category>
            <category><![CDATA[peaq]]></category>
            <category><![CDATA[web3]]></category>
            <dc:creator><![CDATA[Bert Van Kerkhoven]]></dc:creator>
            <pubDate>Fri, 29 Aug 2025 14:13:29 GMT</pubDate>
            <atom:updated>2025-08-29T14:13:29.981Z</atom:updated>
            <content:encoded><![CDATA[<h3>How the cloud came about</h3><h4>From transistors to integrated circuits</h4><p>In 1957, eight engineers — later called the ‘Traiterous Eight’- walked out of the Shockley Semiconductor Laboratory to start Fairchild Semiconductor. Their decision to venture out on their own was largely driven by their frustration that the semiconductor space had been largely stagnant ever since the transistor was invented in Bell Labs in 1947. What followed over the next few decades vindicated them in their vision:</p><ul><li>1949: The invention of planar architecture by Jean Hoerni, allowing transistors to be built in directly into flat pieces of silicon — increasing their lifespan, decreasing costs and allowing for production at scale.</li><li>1961: The first commercial &amp; monolithic integrated circuit, invented by Robert Noyce by interconnecting multiple transistors on a single chip</li><li>Mid-1960s: Annual Fairchild chip production scales into the millions.</li></ul><p><strong>In 1965, Gordon Moore — one of the Traitorous Eight and then Fairchild’s Director- observed his famous law: the number of transistors on a chip would double roughly every two years while costs would fall because manufacturing would become more efficient. </strong>Moore’s Law has held up incredibly well and has transformed the world in every way possible: it enabled the rise of the cloud, shaped national defence capabilities (e.g. guided missiles), altered the balance of geopolitical power and underpinned trillions in global economic growth.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Y5AqIg3OlOxks_FX.png" /></figure><p>The relentless persistence of Moore’s Law and the rise of compute power is only part of the story. With the increasingly important role compute plays, power accrues to who owns the GPUs. With the technological milestones described above, the industry dynamics in semiconductors have oscillated between centralization and decentralization.</p><p><em>Centralization because of prohibitive costs</em></p><p>Initially, the dominant use-cases for integrated circuits were for military applications. In the early 1960s, chips cost thousands of dollars and these costs were only justifiable for purposes that served national security interests (like guided missiles &amp; spacecraft). We were a long way off from the use of chips in consumer products. Compute capacity lived predominantly in mainframes (IBM System/360, 1964), costing millions of dollars — limiting their market to large corporations, research institutions and governments.</p><p><em>Decentralization because of Moore’s Law</em></p><p>It was only in 1981 that the first personal computers emerged, built by IBM and powered by Intel’s 8086 microprocessor (with a transistor count of 29,000). Advanced compute could now sit at personal desktops. The acceleration of decentralised compute gained steam during the 1990s and by the end of the decade, transistor count in consumer-grade PCs was in the millions (e.g. Pentium III had 9.5m transistors and could be found in $2000 PCs).</p><p><em>Centralization because of efficiency gains (Moore’s Law =&gt; idle compute at individual enterprises)</em></p><p>In the mid-2000s the opposite trend emerged and compute started to centralize again with the introduction of AWS (2006), Microsoft Azure (2010) &amp; Google Cloud (2011). These so-called hyperscalers could provide compute on demand while amortizing capex across millions of users. This was relevant because Moore’s Law created upward supply shocks, while demand struggled to keep up (utilization rates dropping to below 20%). These compute factories use cutting edge chips (Nvidia’s H100s — 80 billion transistors on a single chip), deploy industrial-scale cooling and are powered by 10s of megawatts of power. One plant can house as many as 100,000 servers.</p><p>Two decades in, Cloud has turned into a massive and concentrated industry — grossing just shy of $100bn in last quarter alone with enormous concentration (AWS/Microsoft/Google taking a 32%/22%/11% share respectively).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*2E-w4wTXRf2BY7R8" /></figure><h4>What the cloud is good for and where it falls short</h4><p><em>Pro’s</em></p><ul><li><strong>Elastic backends &amp; data plumbing:</strong> The cloud is a highly flexible solution to cater to demand that both varies in type of jobs &amp; usage intensity. This elasticity is well explained by the <a href="https://nvlpubs.nist.gov/nistpubs/legacy/sp/nistspecialpublication800-145.pdf">National Institute of Standards &amp; Technology</a>.</li><li><strong>AI training:</strong> Given that AI training runs to train foundational models can exclusively be performed efficiently on frontier-tech GPUs (H100s) at scale, Cloud has enabled this to materialize (see also Microsoft’s collaboration with OpenAI)</li><li><strong>Global reach &amp; latency reduction: </strong>Cloud providers operate dozens of data centers across continents, allowing companies to serve customers worldwide with low latency and without the need to build out local infrastructure (e.g. <a href="https://aws.amazon.com/solutions/case-studies/innovators/netflix/">Netflix uses AWS to serve customers in 190+ countries</a>). This doesn’t just drive cloud adoption, it also drives market concentration.</li><li><strong>Resilience against disasters:</strong> Companies can replicate data across regions without the need to build physical data recovery centers themselves at a fraction of the cost. <a href="https://info.flexera.com/CM-REPORT-State-of-the-Cloud">A survey by Flexera</a> found that 63% of enterprises cite disaster recovery &amp; business continuity as a top driver for cloud adoption.</li><li><strong>Cost optimization:</strong> Cloud turns CAPEX into OPEX by shifting infrastructure spending on hardware upfront to a pay-as-you-go model. A <a href="https://www.cloudzero.com/blog/cloud-computing-statistics/">Cloudzero study</a> estimates enterprises can reduce infrastructure costs by up to 40% through cloud migration.</li></ul><p><em>Cons</em></p><ul><li><strong>Market power &amp; switching friction</strong>: It’s no surprise that, despite costs falling under Moore’s Law, the concentration of market power is causing prices to go up (especially egress fees, as well documented <a href="https://assets.publishing.service.gov.uk/media/664f2556993111924d9d3aa8/240521_-_Egress_Fees__.pdf">by the UK government</a>).</li><li><strong>Single points of failure</strong> (and an increased cyber security risk): Concentrating compute across a few hyperscalers may be beneficial in terms of efficiency, but also makes the global internet more fragile (a good example was the AWS outage earlier this year, causing <a href="https://www.financemagnates.com/cryptocurrency/amazon-web-services-outage-disrupts-binance-kucoin-withdrawals-trading-disruptions-continue/">disruptions at several centralised exchanges</a>.</li><li><strong>Physical &amp; power limitations</strong>: International Energy Agency projects global data-center consumption to reach <strong>620–1,050 TWh by 2026. </strong>With the rise of AI (and subsequent demand for GPUs), demand may outpace supply for the first time since the emergence of the cloud. With cloud center power usage rivaling the use of medium-sized countries, an increased focus on energy efficiency seems likely. This means a shift of focus to the cooling overhead of large data centers (30–40% of their power usage is to cool the GPUs), performance-per-watt, and fit-for-purpose matching of tasks to the most efficient GPUs.</li><li><strong>Trust &amp; confidentiality gaps</strong>: The cloud is build under the <a href="https://worldprivacyforum.org/documents/204/WPF_Cloud_Privacy_Report.pdf">assumption of trust</a> as companies don’t own the compute. This means that users that work with sensitive workloads (such as financial data, proprietary fund algorithms, healthcare data, etc) are vulnerable to the operators, insiders &amp; state-level subpoenas.</li><li><strong>Geopolitical risks:</strong> Hyperscalers are subject to the jurisdiction of their home governments (e.g., the <a href="https://www.justice.gov/criminal/cloud-act-resources">U.S. CLOUD Act</a>), which means data stored abroad can still be accessed under domestic law. This creates sovereignty concerns for regions like the EU and Asia, where tensions or sanctions could restrict access and expose sensitive information.</li></ul><h4>Enter Acurast, the serverless cloud</h4><p>Similar to how the cloud solved the underutilization problem in the 2000s, Acurast addresses the new bottleneck of the 2020s: monopolistic pricing, limitations of power &amp; confidentiality gaps. Acurast’s bet is on smartphones powering the next great unlock of compute.</p><p><strong>Smartphones as the new frontier</strong></p><p>While smartphones are often used for trivial tasks, they pack impressive technology. The combined R&amp;D spend in smartphones rivals those of datacenters (Apple alone spent $33bn in R&amp;D for the iPhone last year). Each device ships with frontier chips (e.g. Apple A17 Pro with millions of transistors). Additionally they are increasingly equiped with hardware acceleration for matrix multiplication — making them fit for purpose to work on inference jobs.</p><p>While smartphones have become incredibly powerful, the H100s that the hyperscalers use for many tasks may be too powerful and not the best fit to run most workloads they are actually used for. Most modern smartphones come with 8 cores and 8GB of RAM which could actually run most of the workloads that are currently offloaded to the cloud by enterprise customers. Globally, there are now over <a href="https://www.statista.com/statistics/330695/number-of-smartphone-users-worldwide/">6.9 billion smartphones in use</a>. The average replacement cycle is <a href="https://www.sellcell.com/blog/how-often-do-people-upgrade-their-phone-2023-statistics/">around 3.6 years</a>, leaving hundreds of millions of older but still-capable devices idle in drawers. In fact, it’s estimated that <a href="https://weee-forum.org/ws_news/of-16-billion-mobile-phones-possessed-worldwide-5-3-billion-will-become-waste-in-2022/">5.3 billion phones were removed from use in 2022 alone</a>, with most not being recycled.</p><p>Each device typically has 6–12 CPU cores and 4–12 GB of RAM. While smartphone compute capacity cannot scale vertically to petabytes of memory like servers can, the ubiquity and abundance of them make them the largest untapped compute cluster in the world.</p><h4>Confidentiality by design</h4><p>Where cloud demands trust in the provider, Acurast eliminates it through confidential compute by default. Every workload executes inside a <strong>Trusted Execution Environment (TEE)</strong> already embedded in modern mobile chipsets (e.g. Apple Secure Enclave).</p><p>Two guarantees matter here:</p><ol><li><strong>Attestation</strong>: Each device cryptographically proves to the network that it is genuine hardware in a verified state. This prevents spoofing or tampering.</li><li><strong>Black-box execution</strong>: Workloads inside the TEE cannot be inspected — not by the device owner, not by the operator, not even by Acurast itself. Sensitive data like financial trading algorithms or healthcare records remain opaque.</li></ol><p>Where hyperscalers’ “confidential computing” products still leave control of the trusted stack with the provider , Acurast makes confidentiality trustless and verifiable.</p><h4>Modular and elastic supply</h4><p>Acurast aggregates smartphones into high-performance clusters. The catalyst that enables this is the evolution of USB technology (With USB4 throughput enabling 80 Gbps troughput). Tasks can be delegated to clusters of individual node operators and interconnected devices managed by one manager (where multiple devices effectively function as one compute unit).</p><p>This opens up new possibilities:</p><ul><li>Running a 16bn parameter LLMmodel across pooled devices with sufficient RAMs (16bn parameter typically requires 12gb of RAM — and thus clusters of interconnected devices)</li><li>Spinning up ad-hoc compute clusters for time-sensitive tasks (e.g. sentiment analytics)</li><li>Matching lightweight workloads to get those serviced more efficiently in terms of performance-per-watt (relevant for companies optimizing for environmental impact)</li></ul><h4>The economics of decentralized compute</h4><p>Cloud economics are driven by hyperscalers’ capex and opex. Users pay not just for compute, but for power, cooling, land, and operators’ margins. In stark contrast with this: smartphones already exist, already consume sunk R&amp;D, and are already paid for by end users. Acurast abstracts away the hardware layer, letting any device owner monetize idle capacity while giving enterprises an alternative to spiraling cloud bills — especially for tasks that were not a fit for cloud in the first place.</p><p>This inversion mirrors the rise of cloud itself: just like how AWS capitalized on underutilized corporate servers in 2006, Acurast capitalizes on underutilized mobile devices in 2025.</p><h4>Implications</h4><p>The implications of this model are quite impactful:</p><ul><li><strong>Energy efficiency</strong>: Phones typically operate at 3–6 watts of active consumption, versus 300–700 watts for GPUs in a rack. Matching workloads to phones could massively reduce data-center loads. This makes Acurast the most sustainable alternative to the cloud and also allows for better performance when energy is constrained.</li><li><strong>Geopolitics of compute</strong>: Control shifts from three hyperscalers to a global mesh of device owners. That reduces the choke points where states or corporations can exert pressure.</li><li><strong>Confidentiality guarantees</strong>: Sensitive workloads no longer rely on corporate policy or state jurisdiction but on cryptographic attestation.</li><li><strong>Market structure</strong>: Instead of billion-dollar facilities locked by switching costs, compute becomes liquid and permissionless.</li><li><strong>The rise of a decentralized supercomputer:</strong> Tapping into idle phones globally could unlock an aggregate compute resource larger than the biggest supercomputers ever built, effectively transforming electronic waste into a global distributed cloud.</li></ul><h4>Conclusion</h4><p>We’ve seen compute migrate before: from mainframes to desktops, from desktops to the cloud. Each shift reshaped industries, created new giants, and had a profound impact on access to technology. With power costs, confidentiality, and market frictions now being major constraints of the hyperscale model, it seems likely that the pendulum switches to a decentralized future again for a subset of use-cases that are better served through this decentralized network of idle smartphones</p><p>Acurast is betting that the smartphone — the most ubiquitous, extremely R&amp;D-heavy device in history — will be the catalyst for this shift. If it succeeds, compute will no longer just live in distant data centers, but many applications will live everywhere — in billions of households worldwide.</p><p><em>Disclaimer</em><strong><em>: </em></strong><em>This article is provided for informational purposes only and does not constitute an offer, solicitation, or recommendation to buy or sell any securities or financial instruments. The views expressed are those of the author and are not necessarily reflective of the views of MN Capital or its affiliates. Nothing in this article should be interpreted as financial, legal, tax, or investment advice. Always seek the advice of a qualified financial professional before making any investment decisions. MN Capital and its affiliates disclaim any liability for actions taken based on this information. MN Capital is an investor in Peaq Network and may hold positions in affiliated projects.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9c8c7162d005" width="1" height="1" alt=""><hr><p><a href="https://medium.com/mn-capital/decentralizing-the-cloud-acurasts-bet-on-smartphone-powered-compute-9c8c7162d005">Decentralizing the cloud: Acurast’s bet on smartphone-powered compute</a> was originally published in <a href="https://medium.com/mn-capital">MN Capital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Teneo: a new paradigm for data monetization]]></title>
            <link>https://medium.com/mn-capital/teneo-a-new-paradigm-for-data-monetization-bd7055933108?source=rss-acc377fbefd6------2</link>
            <guid isPermaLink="false">https://medium.com/p/bd7055933108</guid>
            <category><![CDATA[teneo]]></category>
            <category><![CDATA[peaq-network]]></category>
            <category><![CDATA[depin]]></category>
            <category><![CDATA[peaq]]></category>
            <category><![CDATA[web3]]></category>
            <dc:creator><![CDATA[Bert Van Kerkhoven]]></dc:creator>
            <pubDate>Fri, 07 Feb 2025 14:42:41 GMT</pubDate>
            <atom:updated>2025-02-07T14:42:41.217Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*bi7_0SlT-dbFzTnC.jpg" /></figure><h4>The greatest heist in human history</h4><p>Over the past two decades, we saw the meteoric rise of the internet with the total internet population going from just 1 billion in 2005 to over 5.5 billion today. <a href="https://datareportal.com/social-media-users">5.22 billion</a> of those internet users are active on social media. This mega-trend reshaped the entire fabric of our society, radically accelerating how knowledge and culture forms and permeates.</p><p>On the back-end of this trend, we witnessed an unprecedented asymmetry of value extraction in the form of data. Given the novelty of the web, this extraction was largely unnoticed by the general public. Every social interaction, product review, search query — and even the metadata attached to it- is the predominant resource that powers trillion-dollar industries.</p><p>The foundation of modern artificial intelligence, algorithmic trading, and precision marketing has been constructed on the <strong>data generated by billions of users</strong> — yet the financial surplus generated from this knowledge economy has been captured by a small group of technology monopolies.</p><p><em>This was not an unintended consequence of Web2. <br>It was the </em><strong><em>business model</em></strong><em> of Web2.</em></p><h4>The data value chain</h4><p>If data is the new oil, big tech firms like Meta, Google, X &amp; Reddit are the new OPEC. Having seized proprietary control of user-generated content without explicit content, they have pushed the boundaries of value extraction by squeezing every possible signal out of this data. This doesn’t only extend to logged in users, external data aggregation enables shadow profiling for any website visitor.</p><p>The shadowy data economy is not constrained to the platforms the data is generated on. There is an entire industry built on data brokerage. Large-scale data brokerage networks package consumer, financial, and social sentiment data into sellable products for third parties.</p><ul><li>Datasets originating from real-time social media feeds are the input for natural language processing engines that power sentiment analysis feeds for hedge funds.</li><li>Advertising networks use the data to build psychometric profiles on potential customers to optimize targeting and manipulate them into buying with high precision.</li><li>Since the business model of social networks hinges on the number of ad impressions and thus time spent on these platforms, there is an incentive to addict users which typically doesn’t align with their well-being (<em>The Social Dilemma</em> by director Jeff Orlowski illustrated this quite well).</li><li>Cambridge Analytica, a British political consulting firm, came under scrutiny for the alleged misappropriation of user data with the goal to swing the Brexit referendum in favor of the Leave camp. Further research into the matter highlighted that Cambridge Analytica wasn’t special at all. Your online data footprint is used to polarize, radicalize and ultimately impact elections.</li></ul><h4>The rise of AI</h4><p>In late 2022, LLMs went mainstream with the launch of ChatGPT by OpenAI. While not publicly disclosed, it is widely assumed that OpenAI used petabytes of scraped data from public sources. Some authors and organizations (such as George R.R. Martin &amp; The New York Times) claim OpenAI used copyrighted material and took legal action against OpenAI. Meanwhile, platforms that have extracted and monetized user-generated data for years are now complaining that OpenAI violated fair use policies. While these law suits are unprecedented and the law is quite ambiguous &amp; formulated in an antiquated way, every platform is taking action against mass scraping of the content they host:</p><ul><li>X has imposed exorbitant API fees, making real-time sentiment analysis prohibitively expensive for smaller firms.</li><li>Reddit’s API restrictions have eliminated cost-effective access to structured discussion datasets, previously used for AI fine-tuning.</li><li>Google and Meta now closely guard proprietary engagement datasets, restricting their use to internal product development.</li><li>Across the board, websites are now updating their robot.txt files to prevent data scraping. Additionally, increasingly sophisticated techniques such as honeypot links, dynamic content loading and IP fingerprinting are being used to deter automated data extraction.</li></ul><p>The impact of LLMs trained by user-generated content is a lot bigger for the general public than any of the earlier use cases for this data. While this data was previously mainly utilised to impact the consumer side of the equation, the drastic impact AI could have on the production side of our economy puts the discussion around data usage and monetization at the center stage. Open-source communities such as Stack Overflow, once the backbone of software development knowledge exchange, are now being replaced by AI-driven code generation tools trained on their own knowledge base. When the quality of the output of these tools reaches a level that allows for large-scale job displacement, we can expect awareness regarding data to spike.</p><h4>The marginal value of data</h4><p>While there are a lot of disagreements amongst the leaders of AI labs (e.g. the debate between Musk &amp; Altman around open vs closed source AI), one thing that they seem to agree on is that we’re running out of data to train foundational models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*AHqg3VhEd7IStT2g.jpg" /><figcaption>Ilya Sutskever at NeurIPS</figcaption></figure><p>There is some merit to these claims with regards to foundational models. However, the introduction of DeepSeek R1 marked a Sputnik moment for the LLM arms race. It showcased that foundational models are not defensible and as such, value accrual is likely to take place at the application layer where foundational models are fine-tuned to cater to specific use-cases.</p><p>In this new paradigm, the critical component shifts entirely from engineering to the data input. More specifically, it’s the highly specialised real-time data that unlocks applications that were previously beyond reach.</p><iframe src="https://cdn.embedly.com/widgets/media.html?type=text%2Fhtml&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;schema=twitter&amp;url=https%3A//x.com/0xNairolf/status/1884247307627159862%3Fref_src%3Dtwsrc%255Etfw%257Ctwcamp%255Etweetembed%257Ctwterm%255E1884247307627159862%257Ctwgr%255E8061f8f5ffbfb92e00996e58f8a08b3aad704a2c%257Ctwcon%255Es1_c10%26ref_url%3Dhttps%253A%252F%252Fwww.chainofthought.xyz%252Fp%252Fweekly-ai-edge-32&amp;image=" width="500" height="281" frameborder="0" scrolling="no"><a href="https://medium.com/media/b157c91b633a84aea6686dae2a6388d6/href">https://medium.com/media/b157c91b633a84aea6686dae2a6388d6/href</a></iframe><p>An obvious example of specialized data is the data that is generated on social networks such as X. Whether it’s finance, politics or emergencies, X always has the most recent data — providing high signal that could be a great basis to fine-tune models.</p><p>The need for this data arises at a time when platforms have already imposed rate limits on the usage of their platforms. Pricing for Enterprise API access to X is estimated between $42,000 and $210,000 (for 50m and 200m tweets respectively) per month.</p><h4>Enter Teneo: the permissionless data layer</h4><ul><li>Teneo users install lightweight <strong>Community Nodes</strong>, which act as decentralized data collectors, aggregating real-time insights from social platforms such as X.</li><li>Instead of relying on centralized corporate APIs, businesses can query <strong>Teneo’s permissionless, censorship-resistant data streams</strong>. This allows them to legally and securely bypass even the most sophisticated traps and blocks that are put in place to prevent automated scraping.</li><li>Teneo levels the playing field by providing access to real-time X data at a price 90% lower compared to the API pricing. Democratized access to this data could tip the scales in favor of a future where the transformative technology is freely accessible for everyone.</li><li>Rather than allowing corporations to extract behavioral data without compensation, Teneo <strong>aligns incentives by rewarding contributors</strong> with $TENEO tokens, creating a direct economic relationship between data producers and data consumers.</li></ul><p>For the first time, a decentralized data economy ensures that users are <strong>not just passive participants in the AI revolution, but direct economic beneficiaries</strong>.</p><p><em>Disclaimer: MN Capital is an investor in Teneo</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bd7055933108" width="1" height="1" alt=""><hr><p><a href="https://medium.com/mn-capital/teneo-a-new-paradigm-for-data-monetization-bd7055933108">Teneo: a new paradigm for data monetization</a> was originally published in <a href="https://medium.com/mn-capital">MN Capital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Peaq is fueling the decentralised machine economy]]></title>
            <link>https://medium.com/@vankerkhovenbert/how-peaq-is-fueling-the-decentralised-machine-economy-7d5d2bf848a3?source=rss-acc377fbefd6------2</link>
            <guid isPermaLink="false">https://medium.com/p/7d5d2bf848a3</guid>
            <category><![CDATA[peaq-network]]></category>
            <category><![CDATA[peaq]]></category>
            <category><![CDATA[iot]]></category>
            <category><![CDATA[depin]]></category>
            <category><![CDATA[economy-of-things]]></category>
            <dc:creator><![CDATA[Bert Van Kerkhoven]]></dc:creator>
            <pubDate>Fri, 27 Sep 2024 11:14:25 GMT</pubDate>
            <atom:updated>2024-09-27T11:14:25.894Z</atom:updated>
            <content:encoded><![CDATA[<p>Throughout history, revolutions have rarely been the result of a single invention. Instead, it’s the <strong>fusion of multiple innovations</strong> that ignites true transformation capable of fundamentally reshaping the world economy. Whether it was the alignment of the printing press with rising literacy or the pairing of electricity with communication, the most profound shifts come from <strong>technological convergence</strong>.</p><p>Technological advancements in isolation are often not very consequential and don’t really enter the public consciousness until they are bundled and take the world by storm. This was true in the late 1700s, when life was initially slow and driven by manual labour, until a series of groundbreaking innovations emerged. First, the <strong>steam engine</strong>, improved by James Watt, provided a new, powerful source of energy, allowing factories to operate independent of rivers or human strength. Next, mechanised looms like <strong>Edmund Cartwright’s power loom</strong> transformed the textile industry by dramatically speeding up production, but these machines needed more power — steam engines fit the bill perfectly.</p><p>At the same time, advancements in <strong>iron and steel production</strong>, such as Abraham Darby’s coke-fueled smelting and Henry Cort’s puddling process, made it cheaper and easier to build robust machines and infrastructure. This newfound strength powered railways, which revolutionised transportation and connected raw materials to factories and finished goods to markets. This convergence of advancements in steam power, mechanisation, and metallurgy would later be referred to as the <strong>industrial revolution</strong>.</p><p>Today, we find ourselves at a similar conjunction as the textiles worker in the late 1700s. The following mega-trends are converging at a rapid pace, giving rise to the machine economy:</p><ul><li><strong>Sensors</strong>: <a href="https://www.ennomotive.com/industrial-iot-sensor-prices/">The average cost of basic sensors (such as temperature, motion, and proximity sensors) has dropped from about <strong>$1.30</strong> in 2004 to <strong>$0.40</strong> by 2020</a>, representing a <strong>70% decline</strong> over 16 years .</li><li><strong>Computing</strong>: The price per gigaflop has decreased by over <strong>99%</strong> in the last two decades. For example, the cost per gigaflop went from <strong>$30,000 per gigaflop</strong> in 1997 <a href="https://aiimpacts.org/wikipedia-history-of-gflops-costs/">to less than <strong>$0.10 per gigaflop</strong> by 2020</a>.</li><li><strong>Storage</strong>: The cost of storing 1 GB of data dropped from <strong>$0.10</strong> in 2017 to around <strong>$0.01</strong> per GB by 2023, <a href="https://www.backblaze.com/blog/hard-drive-cost-per-gigabyte/">a <strong>90% reduction</strong> in less than a decade</a> . Cloud providers like AWS and Google Cloud lowering compute and storage costs <a href="https://www.cloudzero.com/blog/aws-vs-gcp/">by about <strong>50%</strong> every three to five years</a>.</li><li><strong>Energy</strong>: The energy cost per unit of computation <a href="https://www.ll.mit.edu/news/ai-models-are-devouring-energy-tools-reduce-consumption-are-here-if-data-centers-will-adopt">has decreased by 80% from 2010 to 2020</a>. This makes operating large-scale machine networks more economically viable.</li><li><strong>Batteries</strong>: The cost of lithium-ion batteries has fallen <a href="https://about.bnef.com/blog/battery-pack-prices-fall-to-an-average-of-132-kwh-but-rising-commodity-prices-start-to-bite/">from about <strong>$1,200 per kWh</strong> in 2010 to around <strong>$132 per kWh</strong> in 2021</a>, a <strong>nearly 90% reduction</strong> . This has been crucial in making electric vehicles and autonomous drones cost-competitive.</li><li><strong>Transmission</strong>: The cost of data transmission over mobile networks has plummeted <a href="https://stopthecap.com/2021/05/04/mobile-data-costs-plummet-88-in-five-years-u-s-consumers-pay-4x-more-than-rest-of-the-world/">by 88% since 2016</a>, resulting in a cost per mb of data transmission below $0.01.</li><li><strong>Robotics</strong>: <a href="https://www.ark-invest.com/articles/analyst-research/industrial-robot-cost-declines/">The average cost of an industrial robot has decreased from around <strong>$150,000</strong> in 1995 to less than <strong>$30,000</strong> in 2020</a>. Collaborative robots (cobots), designed to work alongside humans, are now available for as little as <strong>$20,000</strong>, making them accessible for smaller companies.</li><li><strong>LIDAR</strong>: The cost of LIDAR sensors, which are critical for autonomous vehicles, has dropped <a href="https://www.itu.int/hub/2020/04/the-price-of-lidar-is-falling-will-driverless-cars-be-on-the-road-sooner/">from around <strong>$75,000</strong> when LiDAR technology first emerged, to less than <strong>$1,000</strong> in 2020</a>. Autonomous vehicle manufacturers have also reduced the overall cost of self-driving systems from over <strong>$100,000</strong> to below <strong>$10,000</strong>.</li></ul><p>We can already see the impact of these innovations. There are currently <a href="https://iot-analytics.com/number-connected-iot-devices/">16.6 billion connected IoT devices</a> globally and this number is expected to grow to 41 billion by 2030. While the first wave of adoption was mainly visible in the consumer electronics space (smart fridges, wearables, etc.), there are currently already <a href="https://ifr.org/ifr-press-releases/news/wr-report-all-time-high-with-half-a-million-robots-installed">more than 500k industrial robots in use</a> (handling 28% of all manufacturing tasks). Humans are increasingly taken out of the equation entirely, with <a href="https://iot-analytics.com/number-connected-iot-devices/">23.4bn wired IoT end nodes</a> (Machine-to-machine communications).</p><p>With the <a href="https://www.unite.ai/ai-training-costs-continue-to-plummet/">cost of training AI models decreasing at a rate of 70% per year</a>, 5G networks reducing latency<a href="https://www.thalesgroup.com/en/markets/digital-identity-and-security/mobile/inspired/5G"> to &lt;1 millisecond</a> and the rise of <a href="https://enterprisersproject.com/article/2021/3/how-edge-computing-and-iot-fit-together">edge computing</a> (<a href="https://newsroom.eclipse.org/news/announcements/eclipse-foundation-unveils-2023-iot-and-edge-commercial-adoption-survey-insights">33% of IoT devices</a> run compute locally, reducing cloud bills for operating these devices), McKinsey estimates the total economic value this Economy of Things could unlock <a href="https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/iot-value-set-to-accelerate-through-2030-where-and-how-to-capture-it">to be up to $12.6 trillion</a> (around 10% of the world GDP) by 2030. The direct market size is expected <a href="https://www.globenewswire.com/en/news-release/2021/04/08/2206579/0/en/Global-IoT-Market-to-be-Worth-USD-1-463-19-Billion-by-2027-at-24-9-CAGR-Demand-for-Real-time-Insights-to-Spur-Growth-says-Fortune-Business-Insights.html">to grow at a CAGR of 25%</a>, to $1,463bn in 2027 (coming from just $250bn in 2019).</p><h4>Who captures the value of the machine economy</h4><p>When machines were introduced in the textiles production process in the late 18th century, it caused mass unemployment while the remaining workers saw their salary more than halved. The result was the uprising of the Luddites, causing factories to be burned down and progress to be faced with massive resistance from the disenfranchised workers. Also in more recent history, we see that productivity gains haven’t benefitted workers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tRNRAEbqjgH4fMWaRCeGXQ.png" /><figcaption>Economic Policy Institute</figcaption></figure><p>Productivity has grown <a href="https://www.epi.org/productivity-pay-gap/">2.7x more compared to wages</a>. Most of the excess wealth gains are distributed towards shareholders. This hasn’t caused too much social unrest because the technological innovation at the source of these recent productivity gains (IT) have been largely complementary to the workflows of the employees. When the technological frontier is pushed to the extent where technology substitutes employees and has the potential to cause mass unemployment, it is essential to democratise access to ownership.</p><h4>DePIN to the rescue</h4><p>We’re facing a massive global underinvestment in physical infrastructure on an as-is basis to the tune of <a href="https://outlook.gihub.org/">$10+ trillion</a> across verticals like agriculture, energy, IT, transportation, etc. This is a very conservative estimate, as the technological trends we outlined above will unlock new applications and investment requirements.</p><p>DePIN as a category is coming of age at a critical time, as the automation and robotics trends is approaching an inflection point in the adoption curve. Not only do DePIN networks level the playing field by democratising access to (fractionally) own these new productive assets, DePIN reimagines what physical infrastructure could be entirely.</p><p>When new technology is introduced, especially at the level of entrenched incumbents, we typically observe a skeuomorphic reflex where this new technology is used to imitate the appearance or functions of older, familiar objects or processes, even when those features are no longer necessary. Chris Dixon explains this well in his book <a href="https://readwriteown.com/"><em>Read, Write, Own</em></a> where he gave the example of early movies that looked like plays or the first popular use of electricity being as a light source (like candles) and only later on to power other appliances. <strong>DePIN </strong>helps overcome skeuomorphic design limitations by <strong>reimagining how machines and infrastructure are built, governed, and utilised</strong>.</p><p><strong>Purpose-built, decentralised networks<br></strong>Traditional infrastructure, like energy grids and communication networks, tend to follow centralised, human-centric designs. DePIN creates decentralised, distributed networks. For example, Helium rethinks the idea of communication networks by decentralising wireless infrastructure through community-owned hotspots. This shifts away from centralised, corporate-driven network layouts to machine-optimised designs, where the network is more efficient and adaptable to machine needs.</p><p><strong>Leveraging autonomy<br></strong>Rather than designing robots to mimic human labor or factory setups, DePIN projects in sectors like energy or logistics can optimise robots and machines for fully autonomous functions. For example, decentralised solar grids or energy-sharing platforms don’t rely on traditional energy grid management and can use automated systems that prioritise peer-to-peer energy sharing and dynamic load balancing.</p><p><strong>Untethered to the human-operated workflow <br></strong>One of the key innovations in DePIN is the use of smart contracts to govern machine operations, bypassing traditional, human-oriented control systems. This allows machines to transact autonomously, make real-time decisions, and collaborate based on predefined protocols, without the need for human intervention.</p><p><strong>Dynamic funding and maintenance<br></strong>In DePIN, the <strong>tokenisation of infrastructure ownership</strong> means that machines can be funded and maintained in decentralised, dynamic ways, allowing continuous innovation without waiting for centralised upgrades or funding cycles.</p><p>The potential for DePIN is vast but both development and adoption has been a challenge because general L1s are not a perfect fit for purpose.</p><h4><strong>Enter Peaq</strong></h4><p>Peaq is a purpose-built L1, designed to power DePIN networks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IyL-qdBS4iaWpujCtjIJrQ.png" /></figure><p>It ensures that machines can securely authenticate and exchange data by introducing the <strong>PeaqID</strong>, a decentralised identifier standard (DID) where machines have their own identity. This allows machines to engage in secure, verifiable interactions while maintaining privacy.</p><p><strong>The Peaq access pallet</strong>, built on Substrate, facilitates <strong>specific access control</strong> for machines and users within DePINs. This allows for the regulation of <strong>permissions</strong> and interactions based on predefined roles.</p><p><strong>Peaq Pay</strong> enables smooth transactions between machines and users within networks. This process includes creating multi-sig wallets, funding them, and approving transaction flows (such as refunds or payments).</p><p>Peaq integrates <strong>Fetch.ai’s autonomous AI agents</strong>. These agents are used to optimize processes, such as automatically matching users with services like parking spots or coordinating complex machine networks.</p><p><strong>Machine NFTs</strong>, which represent unique digital versions of individual machines or machine pools. These NFTs can be used to fractionalize ownership or crowdfund new machines.</p><p>With a Nakamoto Coefficient of 91, Peaq ranks amongst the most decentralised and resilient Layer1s in existence today. The Nakamoto Coefficient is an indicator designed by Balaji Srinivasan and Leland Lee and measures the smallest amount of independent entities that can collectively shut down a chain (measured as how many node operations would collectively amount to one third of the staked amount). Peaq is a high performing Layer-1 with a TPS of 10k and achieves this in the most environmentally friendly way possible while providing modular infrastructure to power and tokenise vehicles, robots and devices on any DePIN. The speed combined with a transaction fee per transaction of $0.00025 and high throughput unlocks many new DePIN applications. Peaq brings this power together with infrastructure that facilitates these new use-cases — e.g. protocols can be engineered in a way that users pay as they go, by <em>streaming </em>micro-payments per kilometer or minute they are using a device of robot. If this device needs to be charged on the way, the device can autonomously transfer money to a third party providing the service.</p><h4>The Peaqosystem</h4><p>There are currently 45+ DePIN projects building on Peaq, amounting to 1m+ machines, robots, vehicles &amp; devices across 18 industries (with a total aggregate value of $25m+). Meanwhile, Peaq partners with industry leaders such as Bosch, Mastercard, Airbus, Continental, etc. How these partnerships are structured varies based on the strategic plans of these enterprise partners. For example, Airbus partnered with Peaq in the context of their Gaia-X moveID project, which is focused on developing decentralised mobility solutions. In this project, Peaq contributes its blockchain infrastructure, particularly its Self Sovereign Machine Identities (SSMI), to enable secure communication and transactions between vehicles, machines, and other infrastructure components.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Tb9kBstlGPASFxJS" /></figure><h4>Peaq DePIN projects</h4><p><a href="https://teneo.pro/">Teneo Protocol</a></p><p>Teneo Protocol is a platform focused on the tokenisation and data management of real-world machines, ranging from shared EVs to Robo Cafes. Teneo Protocol has significant traction and clear signs of product-market fit, with 30m datasets processed and $5m in RWA value tokenised.</p><p>The founders first built Eloop, a car-sharing service that originated in Vienna and allowed customers to rent Tesla’s while also giving them the option to buy tokens and own a fraction of the fleet (and subsequent revenues from ridesharing operations). The goal of Teneo Protocol is to unlock previously inaccessible data with high commercial value. The machine identities (SSMI) and decentralised real-time data streaming &amp; verification that Peaq offers were invaluable to bring this idea to market.</p><p><a href="https://www.farmsent.io/">Farmsent</a></p><p>FarmSent is a marketplace aimed at decentralising the global agricultural supply chain. It empowers farmers by providing them with direct access to buyers, eliminating the need for intermediaries, which often take a large share of the profits. The platform helps ensure that farmers receive fair compensation for their produce, while also providing transparency and traceability in the food supply chain.</p><p>Peaq plays a crucial role in enabling this by enabling machine identities<strong> </strong>(SSMI)<strong> </strong>to manage the identities of farmers and supply chain devices, ensuring secure and transparent data uploads on chain. Moreover, Peaq’s smart contracts are deployed to support global food traceability systems and tokenised reward mechanisms, further enhancing transparency and participation within the network. With over 160,000 farmers from regions like Indonesia and Colombia already onboarded, FarmSent improves both market access for farmers and food security globally​.</p><p><a href="https://www.natix.network/">NATIX</a></p><p>NATIX is a platform that collects valuable real-time data from environments such as roads and urban areas in a privacy-preserving manner by using a network of AI powered cameras. The network’s goal is to map and analyse the world dynamically, turning devices like dashcams and smartphones into data-collecting hubs that can monitor traffic, crowd sizes, and other real-world conditions.</p><p>The data collected is anonymised and can be monetised by users, who earn rewards for contributing to the network. Peaq enables NATIX by assigning unique IDs to each sensor in the network, ensuring secure and anonymous data collection. Additionally, Peaq’s <strong>machine rewards mechanism</strong> allows users of NATIX’s <a href="https://apps.apple.com/us/app/natix-drive/id6446762335"><strong>Drive&amp;</strong></a> app to earn rewards based on their contributions to the network. This allows NATIX to scale and provides privacy-focused data management, crucial for its AI-driven sensor network​. With 160k registered drivers and 80m+ km already mapped across 171 countries since its launch in April 2023, NATIX is one of the DePIN projects to watch.</p><p><strong>Disclaimer: MN Capital is an investor in NATIX</strong></p><h4>Peaq’s mission</h4><p>As we stand on the brink of the decentralised machine economy, Peaq is in a great position to play a pivotal role in this movement by redefining how machines, data, and infrastructure interact. Peaq’s infrastructure, with its focus on interoperability and modular DePIN functions, enables projects within the peaqosystem to integrate seamlessly.</p><p>There are many examples of how projects collaborate within the Peaq ecosystem, both across the value chain and across the tech stack.</p><p>Take <a href="https://www.xmaquina.io/">XMAQUINA</a> as an example. Their partnership with Farmsent ensures that the supply chain of its robotics operations, such as in the Robo-Cafes, is transparent, traceable, and equitable. Meanwhile the integration with <a href="https://chirptoken.io/">Chirp Network</a> addresses the critical need for reliable connectivity and real-time monitoring of its autonomous systems. They work with Teneo to transform raw machine data into a valuable asset through advanced processing, labeling, and AI integration. This data is then used to unlock new revenue streams and business models for XMAQUINA.</p><p>At the heart of these synergies is Peaq’s design, purpose-built to power DePIN projects across industries like mobility, energy, healthcare, etc. The integration of machine identities, machine tokenisation, and real-time data streaming allows these projects to collaborate, optimising processes and unlocking new opportunities.</p><p><em>Disclaimer</em><strong><em>: </em></strong><em>This article is provided for informational purposes only and does not constitute an offer, solicitation, or recommendation to buy or sell any securities or financial instruments. The views expressed are those of the author and are not necessarily reflective of the views of MN Capital or its affiliates. Nothing in this article should be interpreted as financial, legal, tax, or investment advice. Always seek the advice of a qualified financial professional before making any investment decisions. MN Capital and its affiliates disclaim any liability for actions taken based on this information. MN Capital is an investor in Peaq Network and may hold positions in affiliated projects.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7d5d2bf848a3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Playing the odds: the founders’ playbook towards ergodicity]]></title>
            <link>https://medium.com/@vankerkhovenbert/playing-the-odds-the-founders-playbook-towards-ergodicity-7f5cc4523505?source=rss-acc377fbefd6------2</link>
            <guid isPermaLink="false">https://medium.com/p/7f5cc4523505</guid>
            <category><![CDATA[founders]]></category>
            <category><![CDATA[ergodicity]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[venture-capital]]></category>
            <dc:creator><![CDATA[Bert Van Kerkhoven]]></dc:creator>
            <pubDate>Sun, 30 Oct 2022 20:20:48 GMT</pubDate>
            <atom:updated>2022-11-01T11:14:42.681Z</atom:updated>
            <content:encoded><![CDATA[<p>The past decade of seemingly never ending excess and abundance of capital almost made us forget the hard truth about startup mortality rates.<a href="https://www.investopedia.com/articles/personal-finance/040915/how-many-startups-fail-and-why.asp"> 90%</a> of startups fail, and 71% fail within<a href="https://www.alexanderjarvis.com/startup-death-trends-stats-on-survival-rates/"> 20 months</a> after their last funding round. Unsurprisingly, this coincides with the runway they raise for, which is typically between 18 months and 2 years. With the macro environment taking a turn for the worst, it’s a good time to reflect on the implications of failure for both founders and investors — how this has changed and how this will change in the future.</p><p>Not too long ago, when VC as an asset class was still in its infancy and founders were required to take out personal loans to finance operations in the early days of the company, bankruptcy meant the end of the entrepreneurial journey and had consequences far beyond the venture — your reputation would die with your company and your liabilities unfortunately wouldn’t. Banks have become increasingly lenient in providing capital in the early stages to limited liability entities and the VC industry has grown to the point where capital constraints are barely in play anymore; this combined with other factors we will dive into later, has decreased the impact of a failed startup. Thankfully, founders have -over the long run- a higher survival rate than individual ventures. In this post I’ll dive in and explore what has caused this shift and what’s on the horizon to push the envelope even further.</p><p><strong>Doom &amp; gloom: a fear-induced VC freeze</strong></p><p>With European VC funds scaling back investments<a href="https://news.crunchbase.com/quarterly-and-annual-reports/europe-startup-funding-q3-2022-monthly-recap"> by more than 40%</a> quarter over quarter, a lot of European startups are imminent to join the pool of failed enterprises in the foreseeable future. Despite $290bn in dry powder globally, things are looking grim. This dry powder was committed in times when top quartile funds were hitting IRRs of &gt;50% (2019 vintages). It’s very uncertain how eager these LPs will be to wire when that capital call comes in these market conditions.</p><p>In the inevitable carnage of lay-offs, down rounds and companies shutting down — venture-backed companies will suffer the consequences as funding dries up, while VCs are in a significantly better position because of their diversified portfolios and the provisions that protect them in case of bankruptcy (e.g. liquidation preferences).</p><p>It’s fair to say that in the last 10 years, founders have become a lot more powerful in funding negotiations. With a record number of dollars chasing scarce opportunities, it’s surprising that founders seem to have pushed for higher valuations rather than more favourable terms. Many founders now undoubtedly wish they had negotiated softer provisions for downside protection or the option to take some chips off the table through secondaries, rather than negotiating a valuation that might have made headlines but is now evaporating fast.</p><p><strong>Heads I win, Tails you lose: asymmetry in the VC-founder dynamic</strong></p><p>A VC fund typically deploys capital over<a href="https://fundersclub.com/learn/guides/vc-101/understanding-venture-capital/#:~:text=A%20typical%20VC%20firm%20manages,spread%20between%20hundreds%20of%20startups."> 30–80 portco’s</a>, and returns follow a power law distribution — 6% of portco’s drive 60% of the returns, while &gt;50% are write-offs. From a founder’s perspective, one lifetime is arguably too short to make 30 attempts at building a company. Even if it was — loss of principal investment, reputational damage of having x consecutive failures, burnouts, etc. are likely to severely cap the amount of potential shots you have.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*l6d_r4wsL1UhnBCX" /></figure><p>Situations like this where the ensemble average (average fund performance with 30 portco’s) does not equal the time average (average outcome of a founder trying 30 consecutive times), are called non-ergodic. VC funds are in essence vehicles for ergodic arbitrage, but the reality of being a founder has become more ergodic as well.</p><p><strong>Asymmetric bets</strong></p><p>In tech, there is a famous mantra — “Take asymmetric bets”. This is solid advice but assumes contingency.<a href="https://vinodkumaar.medium.com/uncle-points-and-ruins-1ee473437f08"> Uncle points</a>, where the game stops, invalidate this strategy. In general, the world of entrepreneurship has become more ergodic over the last decades. This is true because of several reasons:</p><ul><li><strong>Tolerance for failure<br></strong>Business failure doesn’t come with the stigma it used to come with. Thanks to Silicon Valley culture spreading across the world, having a failed business is perceived as valuable experience in the job market and typically doesn’t diminish your chance to raise external funding for your next venture. There’s a caveat: this assumes you failed in the right way — committing fraud or unethical practices that land you in jail or destroy your reputation are uncle points.</li><li><strong>Development of capital markets<br></strong>While the first Limited Liability Company (The Dutch East India Company VOC) was founded in 1602, it took until 1979 for institutional investors to be able to invest in the VC asset class (when the Prudent Man rule came into effect in the US). The emergence of the VC industry as a source of external funding allows founders to mitigate the downside risk by not putting their life savings at stake.</li><li><strong>Badges of honour<br></strong>Getting into Y Combinator has become as prestigious as getting into an Ivy League university. Building a startup is being increasingly perceived as an alternative way to redeem yourself without having to burden yourself with corporate politics. Being a founder, you’re well-positioned to rack up the accolades and build your skills. It’s ironic that VCs are set-up to take advantage of ergodic arbitrage, but are simultaneously making the game more ergodic. With less than 2% of companies being venture-backed, this in itself is a batch of honour. Bonus points if it’s a top tier fund like Sequoia or A16Z.</li></ul><p><strong>Secondaries as the next frontier towards ergodicity</strong></p><p>In Europe, VCs tend to be skeptical about secondaries because it might erode founder incentives to keep on pushing. It’s valid to question if the math behind VC funds would still hold up if founders were as diversified as the VC fund. However, since business and tech is a postive sum game, it would pay to take the risk of ruin out of the equation for the founder to incentivise more attempts.</p><p>Secondaries are picking up in Europe. Famous examples are Hopin, Wise, Revolut, etc. One thing these companies have in common is that they were all later stage when they started doing secondaries. Here’s why secondaries make a ton of sense in this stage:</p><ul><li><strong>Incentive alignment<br></strong>Founders have a median ownership stake of 15% in their company at the time of an exit. It’s the mark of a great VC to push their winners hard, but this is difficult when an early exit would already be a life-changing outcome for the founders. Maslow’s hierarchy of needs dictates that one starts to care more about status &amp; self-actualisation once basic needs are met.</li><li><strong>War for talent<br></strong>Having pre-exit liquidity opportunities is a great way to keep early hires committed to the grind and recruit top talent. Software engineers are promised stock options in high-growth tech companies all the time. However, the massive exits they are promised rarely materialise. Having the chance to sell shares early makes this promise more credible.</li><li><strong>Planting the seeds for the next batch of unicorns<br></strong>Founders and early hires in tech companies are more likely to become angel investors. Giving them the means to seed early-stage companies in a phase that is too early for VC’s anyway will drive investment opportunities down the line. This has been the key to Silicon Valley’s success and it’s exactly why we need a culture of secondaries in Europe.</li><li><strong>GP-led secondaries<br></strong>It’s difficult to block something you might need to engage in yourself. With the markets looking bleak, exits through IPO’s or acquisitions seem like a bad idea. However, funds are obliged to return liquidity to LPs at the end of the fund lifecycle. Many GPs will opt to roll-over assets in a new vehicle and hold on to them until exit multiples return to normal. The only option to give LPs the opportunity to cash out is through secondaries.</li></ul><p>Hopin founder, Johnny Boufarat,<a href="https://www.ft.com/content/3472e2db-717f-4c7a-9d65-c1c461c79870"> selling $150m in secondaries</a> in the B round was by far the most controversial European secondary. In that round, Hopin was valued at 81x ARR. I think it’s only fair to allow a founder to cash in on the hype to a certain extent, and one could imagine that having $150m in the bank makes you think on a longer time horizon. When the markets tank and Hopin is forced into a down round, this story will inevitably turn into a cautionary tale about secondaries. However, at that valuation a down round was bound to take place and at least the ecosystem gained an angel investor to fuel the next success stories.</p><p><strong>Secondaries as time diversification</strong></p><p>The table below with the data on failure rates to exit and to raise the next round was completely counter-intuitive to me. You’d definitely expect risk to substantially decrease with each funding stage. Sure, some companies might mature into cashflow-positive companies that don’t have any need for additional funding rounds or exits. The landing will also be softer in later stages, where founders often have been able to pay themselves a generous salary for years and where their venture has gained them a certain level of recognition and skills regardless of the eventual outcome. Still, with this data in mind, it is imprudent not to sell a fraction of shares with every round.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*YuUkUf9nSd1m47Yx" /><figcaption>Source: Crunchbase</figcaption></figure><p>This levels the playing field more, VCs are not only diversified across startups but also across time. VC funds with a traditional lifetime of 10 years (and a potential extension of 2 years) have an exit window of 5 years. Founders only have 1 shot and many elements need to align in order to exit. Given that it’s hard to time an exit — as negotiations may take months- founders need some luck but VCs mitigate this risk to a certain extent because exits tend to be spread out over this exit window.</p><p>Time to exit used to be 4 years (hence why ESOPs tend to vest over a 4 year period). With the emergence of mega VC funds with €1bn+ AUM, companies remain private a lot longer. While this makes sense and allows for bigger exits, it forces founders to keep their full net worth tied up in their business for a lot longer. Secondaries provide a great solution to this trade-off.</p><p><strong>Actionable advice for founders</strong></p><p>An environment that encourages a healthy dose of risk taking for founders benefits everyone. Yale economist William Nordhaus’s research indicates the innovative growth companies only capture on average 4% of the value they create. Given the increasing impact of growth companies on the overall competitiveness of a nation, eliminating uncle points and shifting the median outcomes for founders more to the average outcome should be a national priority.</p><blockquote>“The single most important contributor to a nation’s economic growth is the number of start-ups that grow to a billion dollars in revenue within 20 yrs.” -Carl Schramm, Economist and former president of the Kauffman Foundation</blockquote><p>That being said, founders can take steps to mitigate risks:</p><ul><li><strong>Don’t bet the farm<br></strong>As a founder you’re already going all-in with your time. External funding at the very start of your venture will typically dilute you too much so it pays to seed your operations yourself. The<a href="https://en.wikipedia.org/wiki/Kelly_criterion"> Kelly Criterion</a> can provide guidance on what’s sensible to invest yourself, you can find a simple calculator<a href="https://dqydj.com/kelly-criterion-bet-calculator/"> here</a>. <br>Joe Gebbia and Brian Chesky covered the expenses for building Airbnb by<a href="https://www.cnbc.com/2017/06/30/airbnb-ceo-went-from-25000-in-debt-to-running-a-30-billion-company.html"> racking up credit card debt</a>. If Airbnb wouldn’t have worked out, they would probably have been forced to take stable jobs in order to repay these loans- and questionable if they would have ever tried again. If you’re young and you start with zero, check out<a href="https://www.antler.co/"> Antler</a> or<a href="https://www.joinef.com/"> Entrepreneur First</a>.</li><li><strong>Build transferrable skills <br></strong>Founding a company comes with a steep learning curve. It’s hard to argue against the idea that facing the uncertainty and adversity that comes with building a company is a fast path towards self-development. Ex-founders also make better VCs, having founder experience is a great way to break into VC (even if you lack the Ivy league or investment banking credentials). Naval’s <a href="https://spearhead.co/">Spearhead</a> initiative has been quote successful and capitalises on this by turning founders into VCs.</li><li><strong>Find your tribe <br></strong>All returns in life come from compounding interest in long-term games that you play with long-term people, according to<a href="https://nav.al/long-term"> AngelList CEO Naval Ravikant</a>. An increasing number of organisations are building around connecting (aspiring) high-calibre founders. Antler and EF,<a href="https://www.launchhouse.com/"> Launch House</a>,<a href="https://pioneer.app/"> Pioneer</a> are great examples. Also interesting to see the first investment DAO’s pop up and add an incentive layer on top of their community through tokens (<a href="https://www.vc3dao.xyz/">VC3 DAO</a> is a very exciting example).</li><li><strong>Build a reputation that lasts<br></strong>While avoiding illegal or unethical practices is a must. How you treat your investors, co-founder, employees, etc. not only impacts your odds for long-term success in your current venture, but also your odds for getting more shots if it doesn’t work out with this one. Being honest with your investors and sending out regular updates enables them to help out and makes the case for follow-on investments easier to make. <br>The trend of<a href="https://www.entrepreneur.com/science-technology/building-in-public-how-tech-companies-master/359180"> building in public</a> takes this to the next level and also diffuses insights to other founders. Leaving a trail of your entrepreneurial endeavour</li><li><strong>Discuss liquidity upfront <br></strong>Discussing liquidity forward-looking at the start of the founder-VC relationship makes sense because at that point it’s still all uncertain and in the future and will thus be perceived by investors as an incentive for founders to hit these first milestones faster. <br>Many VCs are open about how they look at the option for secondaries for founders. Even for those without access to Pitchbook, it shouldn’t be too difficult to figure out which funds allow secondaries for founders, as they are typically vocal about it. Liquidity is also about more than just secondaries. Salaries and bonuses for both founders and employees and how this will evolve over the lifetime of the startup should be discussed early on.<a href="https://hillfarrance.com/collaborative-carry/"> Hillfarance</a>, a fund in New Zealand, reserves 20% of carried interest for portfolio companies. This so-called <em>collaborative carry </em>incentives portfolio companies to share best practices internally. No doubt this also generates some interesting deal flow. While not creating liquidity for founders on the spot, it’s a great way to make a founder’s situation more ergodic while also incentivising portfolio founders to share insights internally.</li></ul><p><strong>It’s time to build!</strong></p><p>Only 3 in 1000 Americans become an entrepreneur. Only 2% of founders receive venture funding. Of these venture backed companies, only 0.2% makes it to a $1bn+ valuation. With these ratio’s in mind, over the long run you can expect at most 1 unicorn per 176,000 Americans. It’s fair to assume that in Europe, where VC funding per capita is 5 times lower, this is even worse. Meanwhile, VC backed companies represent 42% of all public US companies — and 57% of total market cap. What’s even more is that VC-backed companies drive 82% of all R&amp;D expenditures. Venture-backed companies truly are gifts that keep on giving — even after VC funds have exited their position. With these numbers in mind, Schramm’s law (see above) seems to be a very obvious one.</p><p>In the aftermath of Covid19, Marc Andreessen’s <a href="https://a16z.com/2020/04/18/its-time-to-build/">“It’s time to build”</a> captured the sentiment many people share regarding the lack of technological progress. He echoes the stagnation thesis that <a href="https://www.businessinsider.com/founders-fund-the-future-2011-7?international=true&amp;r=US&amp;IR=T">Peter Thiel</a> has been vocal about for quite some time now. If we want to create the future we want, we will need more entrepreneurs. Or more attempts (per entrepreneur). A transition towards more ergodicity will give us both.</p><p><em>Edited by Manisha Biswas</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7f5cc4523505" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to succeed as an angel investor without getting lucky]]></title>
            <link>https://medium.com/pitchdrive/how-to-succeed-as-an-angel-investor-without-getting-lucky-c3d9d39cebe4?source=rss-acc377fbefd6------2</link>
            <guid isPermaLink="false">https://medium.com/p/c3d9d39cebe4</guid>
            <category><![CDATA[angel-investors]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[investing]]></category>
            <category><![CDATA[portfolio]]></category>
            <category><![CDATA[venture-capital]]></category>
            <dc:creator><![CDATA[Bert Van Kerkhoven]]></dc:creator>
            <pubDate>Sat, 07 Mar 2020 15:08:49 GMT</pubDate>
            <atom:updated>2020-03-20T10:04:34.571Z</atom:updated>
            <content:encoded><![CDATA[<h3>The best kept secret in angel investing</h3><p>While the yearly returns (IRR) of the average angel investor are between 25% and 30% (according to numerous <a href="https://www.nordicban.net/news/the-profitability-of-business-angel-investment">independent studies</a>), the majority of angel investors are losing money. In fact, according to the Kauffman Foundation (<a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1028592">Angel Investor Performance Project</a> or AIPP), 66% of all business angels are at a net loss.</p><p>The facts above indicate a high skewness in the distribution of returns for angel investors (meaning that the average return is driven by outliers and most angel investors are performing below-average). Angel investors who are outperforming are often labelled “skilled” or “lucky” in media and tech communities, where there is a lot of attention for the 1% of startups that end up generating a &gt;20x return (and the angels who invested).</p><p>While skill and luck are undeniably factors that impact the extent to which an angel investor is successful, there is one factor that seems to trump all other factors that are in play here: diversification.</p><blockquote>“It is psychologically difficult for investors to admit that their best investment is worth more than the rest of their portfolio companies combined. So they ignore or hide that fact, and it becomes a secret.” — Peter Thiel</blockquote><h3>A data-oriented approach</h3><p>It is no secret that there is a high failure rate in the early-stage startup asset class. <a href="https://www.academia.edu/38180618/Angel_Investing_The_Gust_Guide_to_Making_Money_and_Having_Fun_Investing_in_Startups">Research by David S. Rose</a> indicates roughly the following distribution:</p><ul><li>50% of startups fail (defined as a &lt;1x return) and</li><li>20% only generates a dollar-for-dollar return</li><li>20% generates a 2x-3x return</li><li>9% generates a 10x return</li><li>1% generates a return that exceeds 20x, often referred to as a “home run”</li></ul><p>Alex LaPrade, one of the first employees at Dispatch and Angie’s List, has ran a <a href="https://lapra.de/crowdfunding">Monte Carlo simulation</a> on the AIPP database. This demonstrates clearly that by massively diversifying a portfolio, angel investors can offset the losses they face on most of their investments by outliers that generate above average returns.</p><p>So how many startups does one need to invest in? When looking at the simulation of LaPrade (below), we can see that more is always better in terms of risk mitigation and optimizing expected returns. However, we can see the Law of Diminishing Returns is at play here.</p><p>The odds of breaking even drastically increase when diversifying your investments from just a couple of startups to 20 startups. The chance of breaking even increases from 83% to 99%, while the chance of tripling your money increases from 58% to 67%.</p><p>If you want to improve your odds even more at this point, you need a lot more diversification. If you’re invested in 500 startups, the odds of losing money are basically obliterated while the odds of tripling your money are at a whopping 96%.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*UyM_UQJYew4PsMFu.png" /></figure><h3>The Babe Ruth Effect: Playing the numbers game</h3><p>With the data above in mind, angel investors should be focussed on increasing their odds of investing in a home run (&gt;20x return), since it offsets the many losses. This is often referred to as the Babe Ruth Effect. Babe Ruth, one of the best baseball players ever, struck out 1,330 times during his career. Yet, his many home runs where sufficient to offset these many failures and made him go down in history as an absolute legend.</p><blockquote>“How to hit home runs: I swing as hard as I can, and I try to swing right through the ball… The harder you grip the bat, the more you can swing it through the ball, and the farther the ball will go. I swing big, with everything I’ve got. I hit big or I miss big.” — Babe Ruth</blockquote><p>So if the data on portfolio performance in angel investing is widely available, why is it so hard for investors to internalize the Babe Ruth Effect? Behavioral economists have pointed out people’s predisposition to avoid losses in <a href="https://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.5.1.193">many studies</a>.</p><p>This loss aversion in angel investing is a very emotionally-driven process and as ones portfolio increases in size, the expected number of losses increases as well. Angel investors often overestimate their ability to pick winners and try to limit their investments because they just don’t have the bandwidth to follow up on them.</p><p>Important to note is that this research is backward-looking by simulating results as if investments in early-stage startups were combined. However, every angel in this research made his investments without this power law in mind. Blindly throwing money at every startup you encounter just to get to a certain number, probably won’t work out the same way. This is why we came up with a solution that enables smart diversification and provides access to high-quality startups.</p><h3>The Pitchdrive way</h3><p>It’s hard to generate deal flow, qualify startups, negotiate valuations and term sheets and follow up on investments as someone’s portfolio increases in size. To invest in 20 startups or more would be a full-time job, and would require a substantial amount of money to be invested. Faced with this challenge, our founding angels launched Pitchdrive to provide a frictionless way to build a highly diversified portfolio of investments in early-stage startups. We source, qualify, negotiate terms and valuations, and enable you to keep track of key metrics effortlessly. Chipping in is possible starting from €5k, and enables investors to build a portfolio that is massively diversified with smaller capital commitments.</p><p>We are currently accepting a limited number of new angel investors to join. Want to start diversifying right from your phone and unleash the Babe Ruth effect in your portfolio? Don’t hesitate to reach out to bert@pitchdrive.com.</p><p>Happy chipping,</p><p>Bert Van Kerkhoven<br>Head of Business Development</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c3d9d39cebe4" width="1" height="1" alt=""><hr><p><a href="https://medium.com/pitchdrive/how-to-succeed-as-an-angel-investor-without-getting-lucky-c3d9d39cebe4">How to succeed as an angel investor without getting lucky</a> was originally published in <a href="https://medium.com/pitchdrive">Pitchdrive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>