The Occult Gatsby
The Occult Great Gatsby The Great Gatsby Annotated for Esoteric Content: The Fourth Way, Alchemy, Tarot, and Kabbalah Author is my father, Jon Woodson. I'm the web guy who wanted to bring this to a wider audience with a more interactive way to pop-up the notes about esoteric content. Author's blurb from Amazon self-publishing page: F. Scott Fitzgerald’s novel, The Great Gatsby, is a manual for spiritual development. It is not a unique text, for many American modernist novels of canonical stature are similarly grounded in the occult. Fitzgerald was a follower of A. R. Orage who not only taught the Gurdjieff Work (the Fourth Way) but organized a vast literary network to further its goals. The chief reason that this occult literary tendency has remained in the shadows is that the arbiter of the movement has been poorly understood and critically neglected. The Great Gatsby belongs to a genre of the novel invented by Carl Van Vechten. Van Vechten’s novel, The Blind Bow-Boy (1923) was the model for The Great Gatsby and for many other modern masterpieces. My annotated presentation of Fitzgerald’s novel performs three functions. Since emblematic novels are all written using the same phonetic-syllabic code, the alchemical cabala, I have decoded the novel and presented a running account of the coded subtext in footnotes. The actual texts that have decrypted are presented in italics.
A Handwritten Blogging Platform
The idea is a place for beautiful notes like https://insidevoices.handwritten.blog.
Product analytics (and evals) for agent sessions on your MCP
Hi HN! We’re Theodore and Louis, founders of Armature (YC P26). We reconstruct the entire session behind the MCP tool calls you receive, including what the user asked their agent to do and what the agent thought. You wrap your MCP in 3 lines of code (our SDK is available in Typescript, Python and Go) and start seeing in your dashboard: - All sessions reconstructed: it’s like reading the real conversation the user had inside Claude or ChatGPT! - A ranking of your MCP most popular use cases, built from sessions clustering - The most frequent issues your users’ agents encounter so you can fix them. Here is a quick demo: https://youtu.be/ZFlvquhyNMQ The story behind this is that we initially launched Armature as a standalone testing tool (https://www.ycombinator.com/launches/QQc-armature-making-you...) that could naturally be used through an MCP itself. We quickly realized we had no idea how our users were using Armature MCP and if they were satisfied with it or frustrated. It’s something we had also experienced in our previous companies: Louis built MCPs exposed to millions of users and Theo was a Forward Deployed Engineer at Palantir before joining a Datadog spin-off as Founding Engineer. Both testing and product analytics had always been real pains when exposing a product to agents but we always thought there wasn’t much we could do about analytics because the conversation lived in our users’ AI client. Then it struck us: what if we asked the agents why they were making this or that tool call? And what’s the user's intent or potential frustration? So we started experimenting with MCP instrumentation and the use-cases actually surprised us! Many of our first customers had implemented workarounds for their CI to trigger new tests or for their coding agents to fetch the results efficiently. Even though we talked to our first users regularly, they had never shared this feedback with us. We then built automations to automatically cluster use-cases, identify issues frequently encountered and let our own coding agents fix them. When our CTO friends heard about this, they wanted to try it for themselves so we gave them access to a cloned version of our internal product and they started sharing feedback like they never did on our “real” product! That’s when we decided to start working seriously on MCP Analytics as a product. At first we were afraid of degrading MCP performance so we iterated until we reached the exact same success rate as without our instrumentation (89.17 % vs 89.15 % pass rate out of 870 runs). Then privacy was an obvious constraint so we applied the same methods we had learned from working with banking data or building sensitive data scanning in logs. Today, redaction runs client-side before reaching our servers. There are still a lot of things we haven’t fully figured out: not all fields are equally filled by all models, session fingerprinting for serverless / stateless MCPs isn’t perfect, and use-case clustering remains to be optimized. But we are finally launching our analytics product to everyone, self-serve at https://armature.tech with a set-up that takes less than 5 minutes and a generous free tier. And now we are working on fully closing the loop, bringing evals back in our product so we can: identify top workflows and issues -> recommend fixes and improvements -> test fixes at scale on the same workflows run by users, across all harnesses and models -> open PRs to ship fixes directly. The evals can be generated automatically from the session analytics so you can catch every regression and can test every improvement’s real impact across all models and harnesses before shipping it. Here’s an example to make it more concrete: 10 days ago, a marketing automation platform which has had early access to what we built for weeks identified thanks to MCP Analytics that users were frustrated not being able to change their target audience after campaign creation. So they shipped the feature and tested it successfully locally with Claude Code on Fable 5. Then a few days later when preparing their new MCP public release, they ran a suite of evals on Armature and realized that small models could hallucinate audience_ids which would lead their MCP to send the campaign to ALL their contacts by default (which could obviously lead to disasters in prod). This is the kind of story that makes what we are building feel so helpful! Now, the most useful feedback for us would be to know what’s still missing in our product so you can feel you are now in full control of the “Agent Experience”. And if you run an MCP in production we’d also love to know: what do you do today to know if agents succeed and if the users behind them are happy?
What if one fund held tokenized Gold, tech stocks and digital assets?
That's the idea behind Wealtii. I built a platform that lets anyone invest from just $10 into diversified funds spanning tokenized precious metals, tokenized equities, and leading digital assets. Everything is fully automated, so investors can gain diversified exposure without manually managing multiple assets or platforms. We recently launched our new RWA Blended Funds, and to celebrate the launch, all platform fees are 0% for a limited time. I'd love to hear what the HN community thinks about the concept, the product, and where you think it could be improved.
Trykeet
Hi HN! We’re Zack and Tommy the Co-Founders of Keet (https://trykeet.com). We are building a mobile app that generates courses on any topic, with short videos for explanation and games for reinforcement. Courses mirror a real curriculum to help you learn over an extended period of time. Tommy and I met in linear algebra class in college and spent the next 4 years taking classes together. Our friendship was formed around learning new things. In school, someone else designs the curriculum, delivers the content, and writes the assessments. All you have to do is show up. Learning without this structure was frustrating, and there was a lot of friction to get started. Zack has a coffee obsession and struggled to assemble youtube videos, conversations with ChatGPT, and the books he was reading into a coherent understanding of all of the different variables that go into brewing a cup of coffee. His attempts gave him the freedom to follow his curiosity, but the instructional design was difficult. Every piece of content either presumed some prerequisite knowledge or none at all. Keet is our attempt at providing the autonomy to teach yourself anything while adding a structure conducive to learning. It finds a custom starting point and sequences lessons in a logical order. We have found Keet most useful in the following scenarios: - You have a subject matter interest that you enjoy passively learning about. (i.e. you really enjoy learning about medieval history and generate courses on medieval engineering) - You want to explore a niche topic of a subject area you already know a lot about. (i.e. You know a lot about biology but want to explore how migratory animals sense Earth’s magnetic field.) - You see a really niche topic get mentioned somewhere you want to explore more (i.e History of Penny Universities, Double Entry Book keeping or Robert Moses and the creation of the BQE) When you create a course we ask some questions about how difficult it should be, how much depth the course should go into, and a few optional intent questions to understand what content you want to learn about. In the future, we’d like to have a global prerequisite map so that we can understand what each user already knows, so we can generate tailored courses to someone’s prerequisite knowledge. If someone majored in Computer Science, for example, their course should look pretty different to someone who might have no background in STEM. After understanding your goals and intent, we categorize the course based on Biglan categories. Biglan categories have four quadrants: - Hard–Pure (e.g., mathematics, theoretical physics) - Hard-Applied (e.g., engineering, and applied sciences) - Soft–Pure (e.g., history, literature, philosophy) - Soft-Applied (e.g., policy, management, education, social work) We use these categories to adjust instructions to better suit the topic. One instance of this is when searching for examples. A hard-pure course will look for worked problems, proofs, and real world cases, whereas soft-pure courses use primary source narratives and contrasting perspectives. We tried building this as a website and then as a mobile app with just text. The website was bad because we couldn’t engage with the content from anywhere, and when we did engage, the text was boring and couldn’t add clarity to complex ideas. Vox and 3B1B videos made complex ideas accessible, which inspired us to move towards video based lessons. There was never one aha moment that made it work. Over the course of the past year, we’ve been using and iterating on the product until we enjoyed using it. We have two types of explainer videos that are built using Manim (https://www.manim.community/) and Remotion (https://www.remotion.dev/). Manim is used for videos that require math visualizations while the Remotion videos are intended to animate processes and show primary source material to mimic something like a Vox video. We are also building more engaging ways to complete assessments via custom activities. For example, in a genetics course, learners might interact with a Watson-Crick DNA model rather than answering a multiple choice question about base pairs. You can download the Test flight beta (https://testflight.apple.com/join/wkWW2enA) today. We are giving 3 free course generations to every user. You can use the code “KEETHN” on the waitlist screen. When you download the app, try to generate a course on a niche interest/something very specific. The coolest courses that we’ve seen are generated in those categories. [ Notes ] - Personalization is still early. Today, Keet mostly adapts around the topic and course goal; we want it to adapt much more around a learner’s background, pace, and weak spots. - Course generation is slow. We made a deliberate trade off to sacrifice generation time in order to gain higher quality courses, since users are going to be taking these courses for an extended period of time. - We are actively working to improve reinforcement quality. It is easy to generate quizzes, but much harder to generate interactions drive home the concept. [ Pricing ] We plan to have a monthly subscription that gives users credits they can use to generate courses. This will be a model similar to Suno (https://suno.com/).
Speko
Hi HN! I'm Bek, founder of Speko, a platform that finds an optimal combination of speech-to-text, LLM, and text-to-speech models, given your constraints, among all our public benchmarked options, and tells you why. Demo: https://www.youtube.com/watch?v=no2LY2gRh-c Typical production voice agent is an ensemble of three models: STT, an LLM, and TTS. Each of those layers offers a dozen credible vendors, and each month there are new models on the market. Almost everyone evaluates once, picks a stack of their choice, and never rechecks because switching from a vendor to another involves yet another integration and arguments about the numbers. The result is that you use voice agents running last quarter's models while better and cheaper options are available. Before founding Speko, I spent four years as cofounder and CTO building voice agents for enterprises across Asia in 10+ languages. Each time a new speech model would arrive, we repeated the same ritual: hire native-speaking raters, benchmark it against our existing stack, and update production if it improved. Speko turns this process into an API. A team running thousands of calls a day told us: "we can literally go to this dashboard, switch the model, and it will do it for us." How it works: you send a request with your optimization criteria (accuracy, latency, cost or balanced), language and region. The router filters to models which we measured for the given combination of constraints, benchmarks them, selects the winner, and returns a response with headers containing provider, model names, and the scores. The gateway prefetches signed session plans, so a new session dials the provider straight from memory; no control-plane round trip while a caller waits. Failover happens only during connection setup stage: if the provider refuses the connection attempt, we start connecting to the runners-up. Some of the customer stories: one founder came to us not knowing what to pick at all: he gave us his use case and now routes everything through the platform. A property management AI runs LiveKit in Python and had not updated STT or TTS since launch: they did not know their STT had high error rates on their calls, better options existed, and swapping always looked like an R&D project. One team did not know which models to pick for Spanish. A medical team did not know which STT handles medical vocabulary best. In every case we helped find the right stack from the benchmarks, and now they route through us. The measuring part is public: we pass the same inputs to every model in one region in different dated runs and we publish the boards, including those where our selections perform worse than alternatives. A launch demo answers which 30-second clip sounds better; production asks which model survives minute eight, so we test spontaneous speech, money and dates, ten-minute takes, and the rankings change. We trained an automatic scorer for TTS naturalness on our blind head-to-head listening votes; on providers it has never seen a vote for, it picks the same winner our raters do about as often as raters agree with each other. We don't train or sell models ourselves, that's precisely how we keep our rankings impartial. We also open sourced the gateway for teams who want to avoid an extra network hop on the audio path and don't want to share keys with our cloud (https://github.com/SpekoAI/gateway, MIT): one Go binary, which is running as a sidecar in your agent's container, speaks one local protocol over Unix socket, pins provider hosts and attaches your keys. In BYOK mode it doesn't communicate with us at all. Notice that the anonymous, content-free telemetry is enabled by default, and one env var disables it. Cost: the gateway and BYOK setup will be free forever, we charge for the hosted router and managed keys with consolidated billing. Since we started the batch in late June, external usage has grown about 25 percent per week on average, front-loaded toward the launch weeks. I would love feedback from the community: how do you pick speech models now, and what makes you trust the third-party benchmark? https://speko.ai/
Check if any of the $656M in unclaimed royalties at The MLC is yours
Hi HN, I built this. Quick background on why it exists: When music is streamed on digital streaming platforms (think Spotify, Apple Music, Pandora), there are two separate royalty streams: one for the recording, paid through your distributor (DistroKid, TuneCore, CDBaby), and one for the underlying work (generally known as publishing). The work side's mechanical royalties are collected by The MLC, a nonprofit that was created by the 2018 Music Modernization Act (MMA). If you haven't registered your songs with The MLC, there are issues with your metadata, or about half a dozen other reasons, that money will never reach you, it just collects in a big pile we call the "black box". That pile is big. The MLC's own dashboard currently shows over $656M is held (themlc.com/blanket-royalties), and the biggest problem is the MMA dictates that The MLC cannot hold that money indefinitely. At some point, by law, The MLC must distribute this big pile of cash, and since they don't know who it's supposed to go to, they pay it out through a process called "market share" (themlc.com/marketshare). Market share means they pay it out, pro-rata, to the artists, songwriters and publishers that are in the system, which in practice means the largest publishers collect most of the leftovers. This is slated to begin in January 2027 and will pay portions of the pool out monthly (themlc.com/unclaimed-accrued-royalties), as of this morning, the next 12 months of market share sums up to $76.61M, starting with $6.41M in January. Full disclosure: I run Doubly, which is an independent publishing administrator, so I have a commercial interest in this space. That being said, we're a team of two people and we don't have the bandwidth to directly assist the hundreds of thousands of artists and songwriters who are going to start losing this money in January 2027. That's why I built this self-service tool, no signup, no email, no paywall, the reports are completely actionable without needing anything from us. The tool: paste a Spotify artist link (or search for a Spotify artist by name). It pulls every release on that profile, checks each recording against The MLC's public bulk data, and tells you per-song whether it's fully claimed, partially claimed, registered but unmatched, or missing entirely. It then provides a link to the exact MLC tool that fixes each case. It also estimates the dollars stuck, as a rough range. Even if you're not an artist yourself, you probably know someone who is a musician, please share it with them. Thanks for your time and I'm happy to answer any questions about the industry as a whole, The MLC, the tech, anything really.
I built a lite LPU that can do inference on Karpathy's MicroGPT
We had no guide or course that teaches chip design at our university. We had taken a digital logic course, but were disappointed with the fact that the most complex project we did was building a full adder in Quartus using logic blocks, not even in RTL!5 Therefore, we decided to challenge ourselves to dive deep into machine learning (ML) hardware and learn as much as we could on our own. We wanted to prove that basic math (like y = mx + b) and basic logic circuits are enough to help anyone understand how modern AI hardware works. Our goal was to design our own version of the LPU from scratch and run a simple Transformer-style model on it, proving that with minimal Machine Learning and computer design knowledge, it’s totally possible. We were also driven by a simple question: What makes the LPU architecture so compelling that even Nvidia licensed it? Keep in mind, this article is not intended to serve as a tutorial for “how to build an LPU from scratch,” and our architecture is not a 1:1 LPU. It serves as an educational resource for how someone with minimal hardware experience can approach this field, and our journey in building what we think an LPU would look like.
Wirespan, a daily power grid optimization puzzle I made for my son
This was a lot of fun to build! I have an autistic son who is extremely interested in power lines, electrical transmission towers, and how the grid works. I decided to build a little game for him. When I finally showed it to him with bated breath, he rolled his eyes and said, “Is it 3D? I’ll only like it if it’s 3D!” So, I decided to polish it up and release it to the world! You get a 9x9 map with one power plant, three houses, and terrain obstacles, and the goal is to connect the power plant to the houses with an optimal path. Beneath the coat of paint, it’s a graph theory optimization puzzle, very nearly a Steiner-tree puzzle. Any two nodes within Chebyshev distance 2 are wired together automatically. But the wires travel in Euclidian space, so reach is measured on a square and wire on a circle. The skill is balancing that asymmetry to find the shortest route to light up the whole map. Because the original goal of this project was to provide joy to my son, I’ve done my best to make it accessible to everyone. It’s localized into 18 languages so far, has a high-contrast mode meeting AAA standards, reduced motion mode, etc. I do collect usage telemetry to help me improve the game, but I’ve done my very best to take privacy very seriously, collecting anonymous stats about the gameplay itself, with a random id the game generates rather than anything about the player or device. I also generate the privacy policy directly from the analytics code, as a contract to ensure nothing is ever collected without being fully disclosed. I had a really great time with this project, and hope to continue to improve it! I’d love any feedback I can get! Please let me know if you have any questions, and I’ll stick around to answer them! (Reposting because my first attempt landed during a super quiet hour and didn't really get seen)