Best of kanfa.macbudkowski.com on Kiwi News
Most upvoted by Kiwi News readers
20 stories from kanfa.macbudkowski.com have been shared on Kiwi News since 31 July 2023, with 139 upvotes and 16 comments in total. The latest was shared on 6 March 2026. They are listed below, most upvoted first.
Text under a title is an AI summary of the linked article.
- My short crypto writing story
This first-person essay traces the author's path from frustration with the financial system and Big Tech to crypto writing, arguing that platforms like Twitter and Discord are poorly suited to writers. The author describes a 2013 Master's thesis on Facebook overstimulation, then moved to Twitter in 2021 and to Farcaster in July 2022 after Dan Romero followed them. The author co-founded Kiwi, a crypto Hacker News-style site with about 300+ curators and 2,000+ monthly readers, and promotes its Feedbot feature, which automatically posts links from a curated list of top crypto blogs.
- 10 lessons about community building from 7 months of running Kiwi
The author argues that Kiwi News built a strong community by letting members co-create the product, not by promotion alone. The article says Kiwi has about 1,300 monthly readers and 200 curators after seven months, with curators' submissions shaping the feed and the community's quality directly influencing the product. It credits shared activities such as a leaderboard that rewards "kiwi karma," offline meetups at ETH conferences in Paris, Berlin, Warsaw and Istanbul, and a Retro PGF round where members voted using their karma.
- Rise, Marginalization & Return of the Niche Internet
Niche communities built around deep interest in one subject have been marginalized by ad-driven platforms, though the text claims they can return stronger thanks to recent changes in technology and business models. It traces how Yahoo shifted from a curated list toward mainstream traffic, and how MySpace and Facebook kept users on their sites by limiting external links to show more ads. The text argues that ad-based models lean toward mainstream audiences and will likely penalize external links, and it notes that Reddit's revenue is 150 times smaller than Meta's.
- What mom hasn't told you about building in consumer crypto
The author, a former Kiwi founder, argues that consumer crypto products are harder to validate than normal consumer startups because speculation and non-user supporters distort the usage signal. Farcaster Pro made over $1M in revenue within 24 hours of launch, but almost half of its buyers had fewer than 100 followers, and the first 10,000 buyers who paid $120 received $600 in airdrops the next day. At Kiwi, over 100 people bought the $10 Kiwi Pass in its first three months, but only some of them wanted to use the product. The text also points to a small market of about 45M monthly active Ethereum addresses and to falling Kiwi usage during the 2024 Base memecoin mania.
- Turning Nouns into an infinite culture machine
The article argues that Nouns, the CC0 NFT project, creates a lot of brand value but captures little of it, and proposes ways to bring more capital back to the project. It contrasts traditional IP licensing, where The Pokemon Company earns over 75% of its revenue from licensing, with Nouns, which lets anyone use its brand for free and captures only the most expensive final step of sales. Suggested routes include voluntary royalties from funded projects, profit-sharing deals with brands, and making Nouns-funded projects more visible. The article cites a treasury of 3,873 ETH (about $10,000,000) as of writing.
- Why niches matter
Niches matter, the author argues, because bringing people with a shared specialty together can drive major progress and give them a sense of belonging. The text cites Renaissance Florence, the Manhattan Project and Silicon Valley as projects that gathered niche-focused people to cooperate, compete and build on each other's work. It also points to open source, where 99% of new software projects reportedly rely on open source components, and to the crypto community, which it says led to a $2T+ industry. The text adds that niche communities offer belonging, and that a niche makes it easier to stand out than the crowded mainstream Internet.
- The 100 hour gap between a vibecoded prototype and a working product
The author argues that the "vibecoded an app in 30 minutes" claims are misleading, based on spending about 100 hours building Cryptosaurus, a simple app where users get a dino styled as their profile picture. The initial prototype took about an hour, but getting the UI, edge-case image outputs, AWS infrastructure, smart contract minting and Farcaster Mini App integration into production took far longer. After launch, a missing nonce check caused some paid requests to fail, and the author refunded affected users $1 extra. The app reached 1,000+ downloads and 180+ paid $2 for a dino, and the author estimates vibecoding gave a 10-100X speed improvement.
- Building castles without kings
The article argues that most people in Ethereum are there for its ideology rather than its technology, and that the community's shared "why and how" (open source, decentralization, freedom to build as one wants) defines its roadmap and products. It says that decentralization makes Ethereum's smart contracts, stablecoins and other apps more secure than alternatives like Tron's, and it urges the community to shift toward consumer adoption and speak about benefits such as stablecoins, Aave lending and prediction markets, not only defensive features like censorship resistance.
- Don’t try to onboard the next billion users to Ethereum, here's a better way
The author argues that instead of targeting the "next billion users" for Ethereum, projects should build hundreds of narrow apps aimed at 100,000 to 10 million users. The argument draws on history: apps like Microsoft Office, Google and Instagram reached a billion users only after the underlying technology was already widely adopted, such as IBM PCs, home internet and smartphones. The text notes 136 million global ETH owners per crypto.com research, but only about 500k daily active Ethereum addresses per Etherscan. It also says narrow apps such as NBA Top Shots, Sorare and Polymarket can lead users to explore other parts of Ethereum.
- The trap of Speculation-Market-Fit
The author argues that crypto has been dominated by speculation because the market rewards quick financial gains, and that this discourages non-financial products. Crypto products have become "online casinos on steroids," and the author notes that most people find it hard to name five big non-speculative successes. The author sees hope in the diverse newcomers, in less competition among non-financial products, and in users who spend more freely after earning crypto gains, which could let builders create products that generate revenue quickly.
- Meditations on onchain media
The author's central point is that onchain media has not yet found its own format, and that current efforts still resemble putting existing media online as-is. The essay compares this to 1997, when online media's value came from new formats like longer articles, images, video and hyperlinks, plus new distribution that let anyone reach large audiences, not just from storing content online. It lists obvious onchain paths, including token-gated access, onchain edit history, seamless payments, and support for citizen journalism, but calls these an early "put this PDF online" stage. It encourages experimentation, such as using tokens to incentivize journalistic objectivity and ledgers to track the provenance of ideas.
- Distribution Advantage: how to set up GTM for your crypto product (without bs)
The article argues that crypto products win through go-to-market (GTM) execution, not only product quality, and that founders should first choose the right market, segment and channel. It cites Uniswap, where Hayden ran one-to-one promotion at conferences, offices and via DMs for about a year before its launch at Devcon 4 in Prague. Coinbase's early focus on Bitcoin and Blur's focus on pro NFT traders, who generated over 90% of OpenSea volume, are presented as examples of picking the right market. The article also recommends starting with a beachhead market, citing Satoshi's early launch of Bitcoin to a cypherpunk mailing list.
- Farcaster: Hard to build, easy to break, and the path to resilience
Farcaster has faced heavy criticism over the past few months, with Defiant saying it has only 10% of its reported DAUs, Peter Szilagyi saying its users are primarily bots, and Liron saying its valuation is overblown. The author argues the network's problems stem from permissionless growth, memecoin-driven spam, and inflated DAU numbers, and that the team's response has been to create channels that separate the user base so bad actors cannot degrade the experience for everyone. The article says Farcaster went permissionless in October 2023, that a $150M raise drew further scrutiny, and that the team has since expanded its channel system.
- Let's chat
Mac, the author of this blog, is offering free thirty-minute... wait, no duration is stated. Let me restate accurately. Mac is offering free Office Hours calls to readers who like his writing, run their own projects, or want his perspective on something. He says he wants to hear about builders' projects, challenges and visions, especially from long-time readers, and that a call could be a first step toward working together, since the crypto industry is small. He currently does only part-time projects, may consider full-time work or a business development partnership if the chemistry is right, and says the calls are also meant to be fun after eight years that included a co-founder's death, 80-hour CEO weeks and the Kiwi project with little funding.
- Escaping the Memeplex
- Privacy, ninjas and digital houses
Privacy advocates need to make the case that digital privacy deserves the same acceptance as privacy in one's home, and the author argues that people should be shown why, not just told. The author compares online tracking to ninjas following a reader who buys a newspaper, noting that digital tracking is hard to notice while privacy tools are easy to use unnoticed. The author argues that people should accept crypto privacy as a trade-off, as they accepted banks, the Internet and https despite their use in crime, and that wider adoption would help build support. Apple's privacy ads are cited as an example of communicating three key truths.
- Thoughts on more meritocratic social feeds
- Most people don’t care about decentralization. And it’s our fault.
The author argues that most people don't care about decentralization, and that the way it is explained is a large part of the reason. Phrases like censorship-resistant, trustless and no single point of failure trigger unhelpful associations for most people, and decentralization lacks the everyday shortcuts that features like horsepower or megapixels have. The proposed fix is to explain it through concrete examples: contributing to a DAO or open-source project on Discord or GitHub instead of trying to influence FAANG, voting on changes to a platform rather than losing an audience to algorithm changes, and moving friends between web3 social products without being locked into one. Note: the article is written in the first person, so the claims above are the author's own.
- Why people hate contemporary art & how NFTs can save it
- IP Protocols Part I: Why the next MARVEL might start as an NFT collection
The author argues that the next MARVEL-style universe could work more like Ethereum, with permissionless creation, rather than relying on a single studio's control. The article describes IP universes such as Star Wars and MARVEL as "protocols" whose rules let fans build apps like movies, games and merchandise, and it claims that the value of these protocols grows with the number of high-quality apps and users. It contrasts Disney's restricted model, which it says generated about 3.6X more money for Toy Story than Avatar when merchandise is counted, with Ethereum, where it says scams and hacks damaged reputation. It cites Nouns as an example of a bottom-up brand, and it argues that an additional curation layer is needed to avoid being overwhelmed by low-quality work.