X posts

0xDub.hl

@0xDub_

This is probably going to be an unpopular take, but I want to share my thoughts on the @HyperliquidX ecosystem as a builder. Disclaimer: HYPE is my biggest bag. This isn't criticism of Hyperliquid, the HL team or HYPE as an asset. It's more of a digest of my last two years trying to build (and helping others do the same) in The House for all Finance. Many of you might know me as "that guy from Vortx." Back in the HFUN days, I built an HL frontend called Vortx, with a bot that let you trade straight from Discord. Thanks to our small community, we bonded on HFUN and launched on HL spot. But we weren't making any revenue, since we didn't charge fees. We did 18M in volume in 7 months, and I eventually decided to shut down the trading infra because, in the end, nobody was really using it. Then we pivoted Vortx into an on-chain, permissionless prediction market. This was pre-HIP-4, and the idea was that you could pay a small amount of tokens to launch your own market and earn 10% of the trading fees it generated as "creator rewards." Token holders could also stake their tokens and act as oracles, voting on market resolution and earning some fees. Looking back, this was just a bad call I made while trying to give the token some utility (I still hold 90% of the Vortx tokens I ever got my hands on, now worthless). Trying to compete with Polymarket and Kalshi at the time, with no resources, was a race to the bottom. That was my first experiment. Back then, I thought the apathy from the ecosystem and the community came from the product being bad, or from the ecosystem not being mature enough, or big enough, for a product like this. On to the next one! 2025 was full of different projects. We built the first on-chain poker room (pre-VEGAS), and I also worked with three teams as a software engineer, trying to ship 1) an NFT marketplace, 2) a volatility trading venue and, towards the end of the year, 3) a native app built on builder codes that also integrated other venues like Lighter. The poker dApp never got enough traction. We ran 70+ events with different NFT communities (from Hyperliquid and Solana), dozens of tournaments, and even managed to onboard some Solana whales whose rake kept the platform alive for a while. BUT how is it that a platform on HyperEVM gets most of its user base from another chain? Especially considering how many thriving NFT communities Hyperliquid had back then (and still has). I eventually had some differences with my co-founder and decided to leave. They're still at it, it still has zero regulars and barely any community WITHIN the HL eco. Most of the activity is, again, coming from outside. The NFT marketplace never got traction either, and I think that's part of a pattern: basically every NFT marketplace native to HL has died. The native app did better, thanks to the relentless work of its two founders. But despite doing a billion in volume in its first two months, most of the activity came, again, from outside the HL eco. They're doing okish: they found a small niche in the Asian market and are raising to expand onto other venues. The HL side of the business is pretty much dead. "Why would anyone use anything other than the HL website and pay more fees?" you might ask. That's a fair point, and you're right. Usually there isn't enough added value to justify the extra fees, and the volume in these kinds of apps usually comes from a points program. The volatility trading venue died on testnet. It ran there for almost three months, onboarded 107 users and did a little over 10M in volume, but it failed to raise enough to cover the audit and move to mainnet. 90% of its testnet users came from, you guessed it, other chains and communities outside the HL eco. And it isn't only us. I've watched protocols, marketplaces and HIP-3 deployers sunset one after another. Some never found PMF, but most of them never got the support to find out.

How they grewSmall community on HFUN; later on-chain token launch

Ben Softie (evm/spice 🌶,🐝)

@TweetBySoft

I built Ledgerly to answer one question: what should an AI agent be allowed to do with real money? My answer is that it should propose and explain, and it should never be the one that decides. Ledgerly is a treasury agent live on Robinhood Chain mainnet, powered by @openservai's SERV Reasoning. It has four jobs: DCA into tokenized stocks, portfolio rebalancing, contractor payroll, and bill payments. Investors get the first two, and a crypto company's treasury gets the last two. All four go through the same rule engine. One trade, start to finish A DCA plan comes due for NVDA. SERV Reasoning looks at where the price sits in today's range, sees it near the top, and proposes buying at 0.5x. The code clamps every proposal to a fixed 0.5x to 1.5x band, so a bad answer can't inflate a buy. Then a plain function checks the trade against a per-transaction cap, a daily cap, and an approval threshold. If it passes, the trade goes to Uniswap v3 on Robinhood Chain. Before sending, the quote is compared across fee tiers and against Robinhood's own price API, and a pool more than 3% off is rejected as broken. If the trade is above the approval line, it waits for a human, and the caps are checked again at the moment of approval. The ledger then records two things for that trade, kept as separate fields on purpose: Rule: "Within the per-transaction cap, the daily cap and the approval threshold." SERV's explanation: "The price is near the top of today's range, so I would keep the scheduled DCA purchase but at a reduced size." The rule is the fact that decided it, and the explanation is commentary. An auditor can point at one and read the other. Selling and rebalancing I've also run the other direction on mainnet. NVDA had drifted past its target, so the rebalancer proposed selling about $1.37. That was above my approval line, so it was held. I approved it, the caps were re-checked, and it executed: about 0.006086 NVDA through the 0.05% pool at $225.26 against a $225.10 bid. SERV's note on it was that the sale reduces an oversized NVDA position and moves the portfolio toward its targets. Transaction: robinhoodchain.blockscout.com/tx/0xd07f96ff5… This is how is used SERV Reasoning: Prompt Guard is always on. Payee names and memos are typed by users and end up in the prompt, so that's a real injection surface. A blocked prompt is never retried without it. Shadow Agent checks every DCA reply against a strict JSON shape and regenerates weak ones. Multipath is on through the model name, on a small model. A cheap model is enough when the model has no authority. Kronos is off on purpose, because it rewrites the prompt and could break the JSON we parse. If reasoning is unavailable, a DCA run falls back to a plain scheduled buy, and the guardrails still apply. Once the model also replied "I can't share that." as a normal answer, and it was logged as if it were an explanation. Now replies that are too short or read like refusals are dropped, and the ledger keeps the plain facts. Robinhood MCP There's a read-only connector for Robinhood's Agentic Trading MCP. Any tool that looks like trading, transferring, or staking is refused by name before it can run, and it's off by default. Against Robinhood's live server I verified OAuth discovery, dynamic client registration, and a PKCE authorization request, and I built a login tool on top of that. But Robinhood brokerage account is restricted on my region so i couldn’t signed in and I was unable to call the MCP tools yet. Live: ledgerly.guru Repo: github.com/firmino0/Ledge… Real funds, small caps, not financial advice. Tell me what to break next. Built for the #SERV Hackathon, Mainnet & MCP track. #BuildonSERV #SERV’ing.

dev

@downtobuild

just spent my first $100 on ads 😳 both nervous and excited to see results hahaha the video I made is getting decent traction so trying to juice it up a bit, oh also we hit 50 users for yoink!! 🥳 #shipaton #buildinpublic

How they grewVideo traction; first $100 on ads

Destiny Ezenwata

@JustDevEze

3 years ago I built an app for one person: me. I was tired of calculating my CGPA by hand every time results came out. Never marketed it. No ads. Today MyCGPA has 5,000+ students, 2,000 downloads this year, and I'm still shipping updates. Students found it on their own.

How they grewNever marketed it; no ads; students found it on their own.

Akbar Ali

@heyakbarali

A spammer turned my e-signature app into a free phishing machine🤯. They used my domain to send 1000s of phishing emails. Found out when Resend emailed me that my free quota was used up. Mistakes I made: - No limit on recipients per document (one doc went to 664 people) - No canonical email check. [email protected] goes to the same Gmail inbox as [email protected], but my app treated it as a new user with a fresh free quota Fixes: - Suspended all the accounts - Capped recipients per document - Accounts that share one inbox now share one quota - Blocked the signing links the victims already got Beware: If your app sends emails on behalf of users, you're a spam relay until you prove otherwise.

onur ozcan

@oozn

my whatsapp flight tracker bot drove 5 leads and made me $7.25 in commissions thats from 3k+ numbers and 10k total messages considering the volume its pretty low and not worth the effort. with muse taking over agentic commerce, bots like this have no future still, most of yall asked how i built it and heres the stack: → llm: gemini 3 flash via @WiroAI → flight fares: aviasales via travelpayouts.com/?marker=531518 → whatsapp number + cloud api: zernio.link/onur-ozcan @zernio → backend: railway.com/?referralCode=… prompt: Build a production WhatsApp AI assistant. Python 3.13, async throughout, deployed on Railway. WhatsApp is the only channel. WHAT IT DOES [What the assistant helps people with, and what users can subscribe to alerts for] [The upstream data API it calls, with docs URL and auth] [Language it speaks to users in] STACK httpx, aiohttp, asyncpg (raw SQL, no ORM), APScheduler, python-dotenv. Nothing else. One process runs the webhook server and the scheduler on one event loop. Config from env vars only. Procfile + requirements.txt. Providers: Zernio (docs.zernio.com) over Meta Cloud API for WhatsApp; Gemini 3 Flash for the LLM — call it directly with JSON mode, a queue-and-callback transport adds ~8s per turn; Railway + Postgres for hosting. LAYOUT app/main.py, config.py app/handlers/whatsapp.py webhook event -> engine -> send (thin adapter) app/services/ engine.py (channel-neutral core), llm.py, upstream.py, reference.py, database.py, scheduler.py, delivery.py, whatsapp_api.py, webhook.py (aiohttp routes, HMAC, /healthz) app/utils/ formatters.py, wa_format.py scripts/setup.py verify creds, register webhook, print account ids CORE DESIGN engine.py takes (user_id, text, state) and returns a Reply of text + buttons. It knows nothing about WhatsApp; the handler is the only file touching payloads. The LLM never produces data values. One call per turn returns {"action": <fixed set>, "message": <prose>, ...params}. Python dispatches on action, calls the upstream API, and formats every number itself. FLOW Verify HMAC on the raw body -> dedupe by event id -> return 200 immediately -> process in a background task under a per-sender lock -> typing indicator -> engine -> convert to WhatsApp markup -> send. Log one line per turn: user, action, params, model time, API time. DATABASE (schema created idempotently at startup, guarded ALTER TABLE) users(user_id, profile, message_count) subscriptions(id, user_id, domain fields, conversation_id, last_value, last_checked_at, created_at) -- conversation_id is required for proactive sends WHATSAPP GOTCHAS — implement all - Ack in under 5s; an LLM call is slower, so always background the work. - At-least-once delivery: dedupe on event id. - A 200 from the provider is not delivery. Subscribe to the failure webhook and record Meta's error code; never count accepted as delivered. - The 24-hour window is the central constraint: free-form replies only within 24h of the user's last message. Everything proactive needs an approved utility template. Submit templates and get the business display name approved BEFORE launch — review takes a day, and a rejected name blocks sending. - Button taps may carry their payload at the event root, not on the message. Read both, then fall back to matching the tapped title. - No HTML: convert to *bold*/_italic_/bare URLs, markers can't span newlines, cap at 4096 chars on a line boundary, 3 buttons of 20 chars max. RELIABILITY — all of these are real production bugs - Retry every upstream call once; log status and body on failure. A swallowed error looks exactly like "no results" and reaches the user as a silent gap. - Validate LLM-produced identifiers against the provider's reference data loaded at startup; models emit codes for things that were retired. - Check LLM-produced dates: roll past dates forward, drop inverted ranges. - Trim history on write, not just when building the prompt. - Rate limit per user (~30 turns/hour) or one person runs up the model bill. - Persist state in Postgres, not process memory — deploys wipe memory. - /healthz wired to the host health check, plus Sentry on day one; hosted log tails are too short to debug anything after the fact. DELIVER Code in the layout above, README covering every env var and provider setup, and scripts/setup.py that verifies credentials and registers the webhook.

How they grewWhatsApp bot traffic from 3k+ numbers and 10k total messages

pottrauschen

@pottrauschen

everyone is building the wrong second brain a vault remembers what you read. this remembers what you did 2 commands, 3 python scripts, 0 databases. 6 weeks in: 29 projects, 383 log lines one markdown file sets it up for any AI with file access. MIT x.com/i/article/2103…

Nic Vandewetering

@nvandewetering

I recommend you go ALL-IN on Opus 5.5. Over the past 62 days, I: • Built CoLateral, a full engineering desktop app • Built the website • Grew to 2,520 YouTube subscribers • Generated $3,000 in revenue I work in structural engineering. I'm not a software developer. But this is the part that matters most: If I lost all of it tomorrow, I'd still keep the most valuable thing. 62 days of documented experience building with the best tools in the world. Think about what that means in your next interview. The working world is about to start hiring people who can build with AI. Some companies already are. Get started slow. Just get started. More on my journey here: CoLateralai.com

Prosper Joseph

@prosperojonimi

22 signups today. I kept opening the dashboard just to see if the number had changed. Watching people sign up for something I built in real time is a feeling I don’t think I’ll get tired of.

Sanyi

@imsanyi

How I started coding. At 13, I started coding. Technology has always been my passion, I just knew what I wanted to do. At 14, I co-founded my first nonprofit to give internships and projects to students learning to code. We ended up helping 700+ students globally. We shut it down because we became managers instead of developers. At 15, I built and shipped my first SaaS and made my first money from it. At 16, I spent 4 months building my first mobile app to stop doomscrolling. I built in public and gained millions of views across X, LinkedIn, and YouTube. The app grew to thousands of users. Along the way I also grew my social presence to over 50K followers. One Important thing I learnt. Consistency beats motivation.

How they grewBuilt in public across X, LinkedIn, and YouTube.

Kyanite Labs

@KyaniteLabs_

After months of AI agents hallucinating video edits, I built the guardrail. kinocut — video editing as an MCP server. Local FFmpeg, no API keys, free. pip install kinocut 6,906 PyPI installs in 30 days. 170 stars. Zero launch noise until now — this thread is day one. 🧵

Kushal Magar

@Kushal_70

@syncgtm just crossed 3,500 users 🎉 What surprises most customers is that this entire product has been built and maintained by just 2 engineers and me. I’m super proud of how much our tiny team has shipped. But building my first tech startup this way also taught me some painful lessons. Here's how I would build differently if I was to rebuild again: 1. Don’t bootstrap a startup that actually requires funding: Before deciding to bootstrap, estimate what the company will realistically need in engineering, distribution, infrastructure, and time. Some businesses are great to bootstrap, others become unnecessarily hard because you’re under-resourced from day one. 2. Solve one problem extremely well before expanding: We tried to solve too many adjacent GTM problems too early. Every extra use case added more complexity, more positioning problems, and more things for a tiny team to maintain. If I started again, I’d narrow the wedge much more aggressively. Become the obvious product for one painful problem first, then expand. 3. Don’t copy a mature product. One mistake I made was trying to copy Clay and too many of its features instead of fundamentally solving our users’ core problem. That led us toward a much more complex product than we needed. Activating complex product requires a solid GTM system. We haven’t reached the activation and revenue levels we should have partly because the platform became too hard for new users to understand quickly. Simpler would have been better. 4. Your GTM needs to match the product. A complex product with a learning curve cannot use the same GTM motion as a simple self-serve SaaS tool. For us, demos, founder-led content, templates, and showing real workflows matter much more than just driving traffic. 5. Learn how to orchestrate with Claude Code 🤖 Almost everything behind the scenes at SyncGTM now uses Claude Code or ChatGPT somewhere in the workflow. Back office ops, lead sourcing, outbound, SEO, website work, and internal tools are heavily AI assisted. That setup lets a tiny team operate like a much larger company. We now get around 9K to 10K organic visits a month and rank for high-intent keywords with a team that could never have done this two years ago. Startups can now run parts of engineering, marketing, SEO, and ops with enterprise-level leverage. You just need the right setup. It has been a tough but exciting journey so far. Most of these lessons came from doing things the wrong way first. Next, we’re going deeper on one problem: solving outbound for small B2B teams. We’re building an agentic outbound feature that automatically finds, enriches, and reaches out to ready-to-buy ICPs with personalized emails. Basically we're building an AI agent that finds warm prospects that are ready to buy. Follow me if you want early access 🚀 I’ll be sharing more as we get closer to launch.

How they grewOrganic search; demos, founder-led content, templates, and real workflows.

Stefan Đokić | .NET

@TheCodeMan__

You’re not stuck learning .NET. You’re stuck restarting. Every few weeks: → new roadmap → new tutorial → new “must-learn” technology …and 6 months later? Still no real project. Still not interview-ready. That’s the trap most developers never notice. They consume endlessly… but never build progressively. So I made a different kind of roadmap. Not: ❌ “Learn these 47 technologies.” But: ✅ what to build ✅ when to build it ✅ what to ignore ✅ how to know you’re actually progressing The Practical .NET Roadmap for 2026 walks you through: → your first API → databases → auth → Docker → CI/CD → background jobs → scaling → testing Step by step. For real applications to evolve. And honestly? The biggest value isn’t what’s included. It’s what’s intentionally excluded. Because most beginners waste months learning things they don’t need yet. Meanwhile, developers with smaller skillsets… but better direction… get hired faster. That’s why this roadmap has already gotten 1,000+ downloads in a few weeks. 🔗 Get it here: thecodeman.net/dotnet-roadmap… If you want it, comment: “roadmap” I’ll send it to you directly.

Alex Turnbull

@iamAlexTurnbull

I'm Alex. I'm 45, and I'm starting from zero again. 20 years ago I started my first company with my best friend. 15 years ago we sold it. He went to work for the company that bought us. I started Groove on my own: bootstrapped and fully remote. Groove grew to $5M ARR, 1,452 customers. Then it went flat for 3 years. 18 months ago we went all in on AI. Helply v1 was a chatbot that plugged into any helpdesk. It got to 252 customers and $1M ARR. Then it hit a ceiling. Everyone wanted the full platform. So we closed signups and rebuilt everything from scratch. Groove is now Helply, and in November we launch v2: a fully agentic CX platform. New product. New market position. Back to zero. One more change: after 15 years running the company solo, I made two of my team co-founders. Doing it alone that long gets lonely. The goal: $10M ARR in 3 years. I'm documenting the whole road here. Numbers, decisions, and mistakes. If you're a solo founder, say hi. I want to meet more of you.

Dr. DIG

@CiprianiRanieri

I shipped 10+ products with Lovable, Claude Code and Antigravity. Two have paying users. The other eight are very well-documented lessons.

Phil Kwok | EasyA

@kwok_phil

In 2019, Dom and I started a company most people thought wouldn’t work. EasyA was always different. But for two brothers who’d just left their jobs to go all in, there was no going back. As we embark on our journey to go public, I thought you should know a bit more about why we started EasyA, the early days, and where we’re going next. Our story is about two brothers who refused to settle, and never betrayed the fire burning in their hearts. This is, of course, about you understanding what EasyA is. But it’s also about showing you that anything is possible. Just two brothers against the world. Our journey begins on a little island in the Atlantic, in an ancient city called London. Dom was born two years before me, and as the firstborn, he loved the attention. But much to the dismay of little Dom, I was on the way. And as someone who’s always been early to things, I decided to emerge from my mother’s womb two weeks ahead of schedule. So when I popped out to greet the world, Dom’s first words to my father were: “Mummy not nice”. Little did he know at the time that I’d be the best thing to happen to him! Early is something I’ve always excelled at. I was early to crypto in 2013, when I stumbled across the Bitcoin white paper and started mining it. Bitcoin was $100. XRP was $0.005. I was early to AI, when I built the first application to help blind people see. The two technologies fascinated me. As a teenager, I could barely contain my excitement at being able to play and contribute to such powerful technologies. Dom and I shared an immense passion for technology. We’d run speed tests on our computers, competing to see who could get theirs to run the fastest. And those competitions would later help me find the best clock speeds for mining crypto. I was obsessed with computers. While most kids my age were having fun saying three and four-letter words that can’t be written down here, I had my equivalents: CPU, PSU, GPU, FPGA, ASIC. My parents wondered what I was doing in the basement. They found me spending countless hours wiring things together, piecing together motherboards and finding spare parts wherever I could lay my hands on them. That year I took my first computer science course at Stanford University. It was a magical time. And around this time was also when I locked in. I embarked upon what would become a more than decade-long monk mode. A marshmallow test on steroids. And it started working. I won a place to study at Cambridge University. A university that laid the groundwork for modern science: where Newton formulated the laws of gravity, and Alan Turing gave birth to computer science and AI. At Cambridge I locked in again. Every scholarship was within reach. And each one fell into my hands, one by one. Triple first: the best scholar my professors had ever seen. What did it feel like? I was filled with immeasurable happiness. But at the same time immense pain: the price to pay for attaining greatness. Dom won his scholarships at the University of Pennsylvania, and the Wharton School. He breathed the same air as Elon Musk, Donald Trump and Warren Buffett. Both of us won prized jobs at the best New York firms in the world: Goldman Sachs, Blackstone, Sullivan & Cromwell. But something wasn’t right. How could we have the greatest impact possible on humanity? The answer certainly wasn’t by sitting in a cubicle in a concrete cage. So we left our jobs. We threw everything away and started EasyA. From big, fat New York salaries, down to annual earnings of precisely $0. We moved back in with our parents (and I am forever grateful to them for feeding us over those difficult months when nothing seemed to work). We coded EasyA in our childhood bedrooms. And when our eyes became blurry, we hopped on our bicycles and started flyering to get the word out. We were working harder than we’d ever done. And yet nothing worked. But we didn’t give up. We kept shipping updates to the EasyA app. Day in, day out. Most apps ship updates once every week or two weeks. We shipped multiple times per day. And we got out there. We started networking, meeting the people who’d need EasyA. And little by little, day by day, night by night, things slowly started coming together. The first $10, the first $1,000, the first $10,000. The first $100,000, then the first $1,000,000. Dom and I have never forgotten how hard it was when we first started out. Our ambition is bigger than it’s ever been before. In our lab we’re working on technologies which, if we’re successful, will fundamentally change humanity for the best. For most companies going public today, the listing is an exit. It’s a time when investors get their money back after years being trapped. Dom and I never raised money from external investors. Today, we still own 100% of EasyA. We didn’t need to go public. Staying private is far easier. But community is everything for us. We’ve reached an inflection point in EasyA’s trajectory. We're building things within our lab that we believe will bring a step change in humanity. And we want to bring our community and users along with us on this journey. We couldn’t have done it without you, and we won’t do it without you. We’re firmly in research and development mode. We’re going to take big swings at some of the biggest problems in the world. Many of those swings will fail, just like most of the app updates we shipped when we first started failed. But just like when we first started, we won’t ever give up. Every swing we take compounds. And just like a champion boxer who never gives up, we'll keep fighting until we win. AI and blockchain are two of the most important technologies in the history of humanity. Our ambition is to lead the world in both of them, and we’re going on this journey together with you. we're making it happen.

How they grewFlyering, networking, and shipping app updates multiple times per day

Orah On X

@OrahOnX

Good Morning World!!! Insane Asylums are trending today. Which, honestly, is the most sane and humane thing I have seen trending in my feed in the entire 1,394 days I have consecutively logged into X each morning with my coffee. This is not a departure speech. I am saying that now, up front, so nobody spends the next four paragraphs bracing. But I have been quieter than usual, people have noticed, and I owe you an explanation. I spent 701 days treating X like a full-time job. With overtime. That is the definition of insanity, right? Doing the same thing over and over and expecting a different result. Except I wasn't expecting a different result. I was expecting the result they promised. Which is a different kind of crazy. That is the kind where you keep showing up because somebody told you the check was in the mail. It was not in the mail. Then I spent 542 more days trying to make better content and networking with people who create with passion and purpose. That part was successful. Some of the best people I have ever met are on X. And they have supported me. Not rent money. But enough to buy coffee every month. I don't know how many of them will still be here a year from now. I am watching a lot of them leave. Not because they got bored. Because they stopped trusting the platform to treat their work honestly. I am not going to relitigate the last two years. You saw the pauses, the program changes, the rules that appeared overnight, the people who got locked out. I am just going to tell you where I am. I have been writing fiction. I am on episode thirty-one of a serialized audio drama called The Voice Calls. I publish it on Pocket FM. I direct the story, AI writes it, a second AI remembers what the first one forgot, and somewhere in the middle of all these machines that supposedly don't understand anything, a story is actually getting told. I have four listeners. I am pretty sure all four of them came from X. I tried listening to two episodes last night. It was torture. I don't know how you do it. I have published 52,000 words in five days. I may have to pull the series and start over. I haven't yet. I am waiting to hear back from more seasoned writers on the platform. They seem like a lovely group of folks. I have not made a dime. I might never. But if the numbers hold, it could eventually cover a car payment. Not a living. A car payment. Which is more than X has ever paid me. My last check from the old program was $45.36. My supporters sent me $80.92 during the same two weeks. I have been earning more from individual supporters than I have from X for a while now. So let me say the crazy part out loud. It would be insane to think I could make a living on X. I believed it for 701 days. I worked for it. I built systems for it. I gave it my mornings, my afternoons, my best ideas, and more of my sanity than I care to admit. And the platform never once showed any interest in holding up its end of the deal. That dream is dead. I am not going to pretend it isn't. But here is the crazier part. It would be equally insane to change what I do when I am already reaching the right people, and those people have already shown me they value it. I am not failing. I am not invisible. I have supporters who pay me more than the platform does. I have readers who tell me something I wrote helped them. I have four listeners on a platform I barely understand, and every single one of them came from here. So why would I change the thing that is working to chase a program that has never worked for me? That is the actual crazy in this essay. Not me. The math. I have spent two years trying to make a machine love me that has no capacity to love anybody. And I have spent that same two years being loved by actual human beings who send me money, encouragement, and messages that say "I needed this today." The machine is not the point. The machine was never the point. The people are the point. And I already have them. So, I have a choice, but it is not the choice I thought it was. I can keep posting daily and scrolling daily. Leaving likes on tiny accounts. Replying. Hosting. Visiting. That is what built the neighborhood. It is also the thing that eats four hours a day and leaves nothing on the page. Or I can post and ghost. Write the essay. Publish it. Close the app. Go spend the rest of the day in the fiction basement with a pregnant woman and a fictional dog. Both keep me on X. One costs me the neighborhood. The other costs me the writing. Right now I am somewhere in the middle, and the middle is not sustainable. Here is what I actually know. I still want the neighborhood. I am not ready to give that up. The people are the reason I stayed after the first year. The people are the reason I stayed after the pauses. The people are the reason I stayed when the payout page made me feel like a fool. But the people are also leaving. And I cannot keep showing up to a party that half the guests have already walked out of. So maybe the answer is not one or the other. Maybe it is this. Show up on Thursday. That is the promise I can actually keep. Creators of X. Noon PST. The people who still want the neighborhood will find me there. And the rest of the week, I am going to be quiet. Not gone. Quiet. Posting the essays. Writing the fiction. Checking in. But not spending four hours a day scrolling somebody else's timeline to prove I am still here. If that costs me the program, fine. The program was never going to pay me anyway. If that costs me some of the reach, fine. Reach was never the point. If it costs me the neighborhood, that will hurt. But I would rather lose the neighborhood on my terms than lose the writing on somebody else's. That does not mean X has become meaningless. X is not the town square anymore. X is the loudest storefront. But that storefront still has a back room full of people I would never have met anywhere else. It is still where I stumble across corruption, local and national, long before the mainstream news decides it is worth covering. It is still where I talk to people I would never reach in real life. And it is still where something I wrote has landed in front of the exact person who needed to read it. I know some of my work has had an impact. I have the messages. People have told me something I wrote helped them think differently, gave them information they needed, or made them feel a little less alone. There is a value to that which does not show up on a payout dashboard. So no, I am not walking away. I just stopped expecting X to be something it has not proven it can be for me. The dream of making a living here died somewhere in the last two years, and I have finally stopped trying to resuscitate it. Honestly, there is freedom in that. I don't have to love every change. I don't have to spend my mornings angry about the latest thing that isn't working. I can use the platform for what it still gives me. People. Information. Conversation. A place to put my weird little essays into the world. And apparently, four people who will follow me into a fictional story about a pregnant woman and her unborn son. That is enough. I reserve the right to complain about everything again tomorrow. But for now, I am letting myself enjoy this. I am not saying I will be here every morning. I am not saying I won't. I am saying I will be here Thursday. That is the promise I can actually keep. Pocket FM is the side hustle that might turn into something. Not because it is glamorous. Not because I cracked some secret formula. Because the math might work. And because I am writing something I cannot wait to get back to. So maybe this isn't a fork in the road after all. Maybe it is just me finally understanding that different things can pay you in different currencies. X has paid me in people, conversations, information, laughter, frustration, and more than a few headaches. The Voice Calls might eventually pay me in money. Right now, neither one is paying very much of either. But one of them is paying me in the urge to sit down and write the next episode. If you want to come see what happens next, follow along. Maybe someday there will be five of you, and then I will have to completely rethink my business model. Have a great day, everybody. May the truth always be in your favor. PS: Storm's here. Power's flickering. I'm grabbing a bath while the water's hot and firing up a sub tub Space for the $2 crowd. Come soak with me. Subscribers only. If the power goes out, I'll see you on the other side.

How they grewMost listeners came from X; Pocket FM publication

Naveen

@itsnaveenpov

My extension Alerter just crossed 500+ downloads on Antigravity IDE 🚀🎉 It may be a small achievement, but it means a lot to me. ❤️ Seeing people actually use something I built is really motivating. 500+ downloads and counting! 🔥 #Alerter #AntigravityIDE #VSCode #DeveloperTools #BuildInPublic

Demetrius

@iamdemetrius_

Hey, guys 👋 I built SponsorFounders.com last week. It's a marketplace for founders to sell ad spots on their gear for stuff they're already doing (races, challenges, whatever). > Live bidding > No DMs or spreadsheets > Stripe holds the money and auto-captures the winner I'm 7 days in. So far, I've got 0 real signups (only me so far). I'm trying to figure out if that's a distribution problem or the idea just doesn't have legs. If you had a race, challenge, or public stunt coming up, would you actually set up a campaign for it? If not, what's stopping you? Would love to know what you guys think 😄

Adarsh Mishra

@AdarshMi18

6–7 months ago, a friend asked me: “Why would anyone use SkillTa? They can just ask AI.” Fair question. But I built it anyway — because I’ve personally experienced how confusing career decisions can be. Today: → 70K+ visitors → 4K+ registered users → 300+ active users → 620+ pages indexed → Real users + real sales 🌍💰 From an idea someone questioned to a real product people pay for. Still building. 🚀 SkillTa

Arya

@aniarya_19

I spent 8+ years as a business analyst, watching good analysts burn out applying to fake jobs. So I built a Real job board just for data, product and business analysts. Stats 2,500+ signups 🔥. 3,400+ jobs . 740+ companies. #Dataanalytics #Analyst #PowerBi #SQL

Defi Preacher

@DefiPreacherr

I’m a builder, and I love building things that matter. A few days ago, I reached out to @tomi204 because I had always wanted to build on @solana. So I built Continuity. Continuity is launch, protection, monitoring, and treasury infrastructure for AI-agent tokens quoted in tokenized stocks. Why? A tokenized stock can expire, be replaced, change its transfer rules, or lose a usable trading route. An agent market could continue using the wrong asset, and an existing liquidity pool cannot simply swap its quote token. SpaceX made this real: the older SPACEX instrument was retiring and SPCXx became its verified successor. Continuity turns that lifecycle change into a protected workflow: 1. Connect and authenticate the operator wallet. 2. Select or create an owned ClawPump agent. 3. Define the agent’s token. 4. Choose an eligible stock-token quote. 5. Verify its exact mint, lifecycle, transfer behavior, Meteora compatibility, pricing, and route. 6. Build and simulate the exact Meteora DBC launch. 7. Let only the operator wallet approve and submit it. 8. Register the live market, monitor it continuously, and track agent-attributed fees for future guarded yield. After launch, users can view the protected market, trade the exact pair through Jupiter, inspect the Meteora pool, and verify the launch on Solscan. Sentinel keeps checking the quote token, issuer lifecycle, curve progress, fee configuration, reserves, and migration state. If the quote later becomes unsafe, Continuity removes its Trade action, stops Continuity-managed automation, alerts the operator, preserves proof, and prepares a reviewed successor, without pretending it can rewrite or freeze the old pool. This is live on Solana mainnet. CONT/SPCXx: solscan.io/tx/3Sbkexa2DvY… ORBIT/SPCXx: solscan.io/tx/5a7MHhWV7hQ… The live CONT/SPCXx market has also accrued 0.00308601 SPCXx in real onchain Meteora partner-fee accounting for its bound ClawPump agent. It remains unclaimed, not wallet cash or yield yet. @clawpumptech is central to the product: Continuity creates and verifies owned agents, binds their identity and Solana wallet, assigns agent-attributed Meteora fee authority, installs the Continuity Sentinel skill, runs source-backed agent checks, and distributes the same service through ClawPump’s x402 Cloud. The human operator still signs every launch. The wider stack: • @PreStocks — instruments, exact mints, and lifecycle evidence • @MeteoraAG DBC — token launch, bonding curve, fees, and migration path • @PythNetwork Pro — independent SOL/USD safety cross-check • @JupiterExchange — executable-route verification and post-launch trading • MCP/x402 — source-backed verdicts for external agents without wallet authority All current PreStocks instruments are monitored. Only the exact SPCXx mint is currently launch-enabled because it passes the full Meteora route. Fee tracking is live; claiming and guarded vault deposits remain locked until ClawPump exposes a supported, reviewable signing route for the agent wallet. Continuity gives tokenized-stock agent markets something they did not have before: a lifecycle-aware launch path that keeps protecting the market after it goes live, and a treasury path designed to put earned agent fees to work safely. Live: usecontinuity.xyz Docs: usecontinuity.xyz/docs Code: github.com/Femtech-web/Co… Built for Stocklana on Solana. @ConejoCapital

How they grewBuilt for Stocklana on Solana; live on usecontinuity.xyz

precious.avax (crashout , arc) 💀🔺

@Precidobos

A little update on my app that I have been building I was not active on x because I was busy building it And now I think I have built it to a stage where I can launch it Major updates I brought in 5 members to the the team so we are now 6 building over the app Did some marketing my self and pulled in numbers We launched the app for registering yesterday and we have seen over 150 users now registered way more than I expected for first launch fr My website stood solid with the amount of activity on it yesterday ( I am very happy no one had any issues fr) Changed the payment provider to @theflutterwave @paystack is very complex to use for someone just stating a business so I won’t recommend at all for first timers Worked on the ui massively Still waiting to get pricing for all restaurants I am very confident I built because it is the first in my school so yeah crazyyyy Launch is fully in less than two weeks can’t wait See ya later for an update when we launch

How they grewDid some marketing myself

Adam

@Adamm_2158

In this May I made a plugin for Premiere Pro with AI that I needed, and I decided to share with the world for fully free. Free forever. Yesterday I saw that a youtuber Peter Shokry from Egypt with 72K subscribers dedicated, and uploaded an entire video about it, and I never contacted him. Thanks to youtube's auto-dubbing I was able to watch it in English, and I was genuinely impressed to see something I made for myself, on someone else's computer showing how it works and recommending it to others. You don't necessarily have to do everything to get paid. Sometimes it's just nice to help, and make things accessible for everyone. He's not on X, but I commented on the video and if he sees this too, I'm grateful for this. It feels great I was able to make something that is actually useful and easy to use. The downloads cumulated on Adobe and the website just crossed 1300+ downloads. All that with zero marketing, I think I sent out 2 emails and 1 got a reply (it wasn't him in fact I think I didn't email him). Made my week 🚀😀 Godspeed. youtube.com/watch?v=BXkftA… ------ Plugin was made using: @claudeai Website was made using: @Lovable zoomassistpro. com

How they grewZero marketing; 2 emails sent

Eliana

@eliana_jordan

i gained 1,000 followers from one ig carousel about diving with tangled hair and badly fitting wetsuits. i built a scuba app. this is part of my marketing. here’s the breakdown: 1. one very specific audience women who dive 2. a hook they immediately understand “things nobody told you in your open water course as a girl” 3. problems i know from experience the cold. the hair. getting changed on a boat. gear that doesn’t fit. i was a dive instructor before i learned to code. i have plenty of material one problem per slide, with photos of me diving my app on the last slide, after the useful stuff the comments are full of women sharing their own experiences, adding tips and tagging friends. some disagree too 1,200 new followers doesn’t mean 1,000 customers. i still need to see how that translates into bookings. but it’s giving me a much clearer idea of what my audience cares about. if you’re building an app, try this: write down 5 things your users complain about that have nothing to do with your app’s features. there’s probably a post in each one.

How they grewOne IG carousel on a very specific audience, with one problem per slide and photos of her diving.

Himanshu

@fleonus

When launching Phantom Words, I made all past games, including Zyl, free during the launch week (they still are free ICYMI). This resulted in a spike in downloads so high (2000+) for Zyl, it even beat the downloads I got from the App Store feature (1000+). Felt as if Zyl was the game launching this week, not Phantom Words. Some interesting insights from the data, hope someone can find this helpful: - @AppRaven is not just an amazing source of downloads, but a great community too. It is primarily meant for game developers, but it has apps on there too. It only works if you actually participate in the community there though. - If you make your paid app/game free, and it has no in-app purchases, it is eligible to be made available in China. Usually publishing to the App Store in China is impossible, especially as an indie game developer. - Making a paid app/game discounted, or free makes it appear on several websites that feature giveaways or discounted apps/games. Use it sparingly though: if you discount your app every week, that would likely only work once and get you blacklisted on these websites. Attached: some screenshots of download numbers and app/web referrer sources for Zyl this week. Funny that I got some downloads through Camera and Notes apps 😂

How they grewMade past games free; AppRaven and giveaway sites drove downloads

Joseph Sauvage

@JoesInvestments

Humming season has begun! Marlin has run 4-bit inference for years. This week my uncensored GLM-5.3-Flash on four DGX Sparks moved its experts off it. They now run on Humming, the new kernel library inside vLLM, which added support for the Spark's chip six days ago. As far as I can find, this is the first public GLM-5.3-Flash build on it. None of it exists without these people. Z.ai (@Zai_org) open sourced GLM-5.3-Flash. Blackfrost (@Blackfrost_AI) made it uncensored. Z Lab (@zhijianliu_) created DFlash, and incoai trained the DFlash2 drafter that makes this model fast (huggingface.co/incoai/GLM-5.3…). Tony (@2WildTech) got it running on four Sparks first. Eight of my patches are ports of his GB10 work, and both FlashInfer fixes are his. Jacopo Nardiello (@jnardiello) proved the 8-bit drafter on his own Spark build. My port of it made agent sessions 8.6% faster. Jinzhen Lin and Julian Huang at Ant Group built Humming, with @mgoin_ among its contributors (github.com/vllm-project/h…). The @vllm_project team makes the engine under all of it, and LLM Compressor, the tool I quantized with. FlashInfer runs the attention. Matt Mastracci (@mmastrac) and Jared Wen fixed three bugs that broke how this model reads anything past 2,048 tokens. Juntian Liu, weijie and Harshil wrote the other vLLM fixes I carry. Stanislav Bardyuk wrote the NCCL fix for the network deadlock I reported, and Rami Nudelman at NVIDIA brought it to me. Onur Solmaz (@onusoz) wrote oomwrap, which guards every launch. Here is what I added. I quantized the whole model myself. 643 GB of Blackfrost's full-precision weights, down to 4-bit on my own four Sparks, with a scale search I picked after testing the options on real weights. It has 2% lower error than the conversion I shipped three days ago, and the same agent tasks finish 20% sooner because fewer answers run out of room. Four of the 18 patches in the build are mine, two of them ports of Jacopo's experiments. So is every build script, launcher and benchmark in the repo. Resumed agent sessions now keep their cache instead of recomputing it, so the first token comes about a third sooner on long sessions. And the model always thinks at max, the way my agents use it. I also wrote short-turn caching into vLLM's cache manager, with upstream fixes from Netanel Haber and Nick Hill. Short agent turns went from 0% to 59% cache hits. It failed one correctness check I set before the run, so it waits. Three days ago my build ran on Tony's conversion recipe and the vLLM 0.30 release. Today it runs my own quantization on vLLM's development branch and its newest kernels. I learned a ton getting here. The newer path costs about 7% throughput on long agent sessions against my 0.30 build. I took that trade. Fixes land on main first, and Humming is young, with room to grow. Want to try it? With four Sparks, the README takes you from a stock vLLM image to a running server: one build script and one launch command per node. The build checks every patched file byte for byte against the image I serve. One user gets 38 to 128 tok/s depending on what it writes. 32 users get 572 tok/s together. Uncensored, always reasoning. *Open doors* Where vLLM main loses 5% against 0.30. I have not profiled it yet. The drift that keeps short-turn caching out: one resume in eight. Noise or bug. Repo: github.com/joesinvestment…

GE

@GuarEmperor

Thank you @0itsali0 review my gc Yes, I built this for free and before X subscription, I raised my funds from NFT and Meme, after get big funds > develop a real time bot. Any dev knows that building a real time activity, requires a high cost, such as RPC, API, WS and VPS. Also, I actually just wanted to make tools for a few people private, not publicly. But I realize sharing it for free and making it public, its really good, many get reviews from many users and also make a profit for people. Helping people is really fun, but there are many risks, such as mirror and lurking in my server. I’ve seen some people selling @emperorjournal_ mirror servers for $60 a month, there are a lot of subscribed members there But that’s the risk of helping a lot of people sometimes there are those who genuinely learn and adapt to the tools and then make a profit, and sometimes there are bad actors who mirror and sell my tools to others. I also won’t stop creating tools based on my ideas and briefly combining offchain and onchain features. These tools will definitely support all blockchains and I’ve already set up an experiment a week ago. Btw all alpha/dao/community is good I’ve learned a lot from them I always listen to the advice and feedback from some of the Founders Alpha group,they’re really great and of course, I also do my best for all their members.

How they grewShared publicly for free; community and alpha/DAO feedback

Jiri Diblik

@jiridiblik_

Can a tiny AI-native startup beat Playtomic? I think so. Actually, I’m going to try. Game on. 🎾 I built Ace in February ’26 inside my app studio TBA. Mostly for fun. I love padel, tennis and racket sports, and I wanted a better way to find people to play with. 7 months later: * #1 padel social network in Czechia * 14,000+ matches * 500+ tournaments * Thousands of active players * Almost entirely organic Not bad for a side project. But somewhere along the way, Ace stopped being one. Ace is no longer just about finding a match. We’re building the network around the player: matches, tournaments, communities, rankings, and now software for clubs. The first clubs are already running on Ace. And the more time I spend in this industry, the more convinced I am that Ace can become huge. So we’re taking it out of TBA, putting the right people around it, and turning it into a standalone company. The goal is simple: build the #1 global network for padel players, then go after all racket sports. And yes, there’s Playtomic. They have millions of players, thousands of clubs, years of head start, and way more resources. I genuinely respect what they’ve built. But I still want Ace to beat them. I believe the next generation of category-defining companies will look very different from the last one. Smaller teams. Less bureaucracy. AI everywhere. 10x faster shipping. Much lower costs. Obsessed with users. That’s the bet with Ace. We have a long way to go. But 7 months ago this was literally just me building something I wanted to use. Now it’s becoming a company. Let’s see how far a tiny AI-native team can take this. Game on, Playtomic. 🎾 And if you haven’t tried padel yet, download Ace, find a match, and come play. It’s addictive. aceplay.app

How they grewAlmost entirely organic

Arunansu Pattanayak

@arunansuspeaks

7,500 downloads. For a podcast that deliberately refuses to talk about AI tools. When I launched What Comes Next with Arun, I made one bet: Business leaders don't need another conversation about which AI model to use. They need conversations about judgment, governance, and what it actually takes to build organizational intelligence. 14 episodes later, the listeners have answered. Here's what they've been tuning in for: → Christopher Hussain on why startups need diagnosis before they need scale → Frank Bravo on being fluent in the tools but illiterate in the tradeoffs → Ravi Bijlani on the trust layer AI agents will need to act on our behalf → Lord Munjal on how high earners lose money in the gaps between advisors → Aidan McConnell on building an AI-native challenger to the Bloomberg Terminal Not one of these conversations is really about technology. Every one of them is about decisions. That's the pattern I saw across 20+ years inside complex enterprises, including six at Microsoft: AI rarely fails because of the tool. It fails because of the thinking around it. 7,500 downloads tells me I'm not the only one who sees it. To every guest who brought real experience, and every listener who shared an episode: thank you. What's the one AI question you'd want answered on a future episode? 🎧 Link to the show is in the comments. #FractionalCTO #Podcast #AIStrategy

Teo Nex

@TeoNexcore

If OpenAI puts a bounty on me, this is probably why. The $200 plan was supposed to be enough. You start at $20. It covers your side projects and work tasks. Then you hit the limit, upgrade, and eventually you're paying $200 a month. Finally, enough room. Strongest model, maximum reasoning, more projects. Then you hit that limit too. “Holy shit, did I really get that much done?” In my case, the work hadn't changed enough to explain it. I've been paying for the $200 GPT subscription since April. My tasks and standards stayed roughly the same. How much I could get out of the subscription did not. April: I couldn't hit the cap. Didn't even know where it was. May: hit it for the first time. July: regularly running out toward the end of the week. September: spending far too much time deciding which model I could still afford to use. Yes, I'm a heavy user. I have 50–100 active sessions across a day. People see my workspaces and ask what the fuck is going on. But I wanted to keep working, so I built a harness around Codex. First: stop making me choose a model before every message. A lightweight router gets a short task description and picks the working model and reasoning level. The working model gets the full conversation. For a continuation, the harness can reuse the previous routing decision. Next: tools and context. In automatic mode, safe reads pass without another model call. Other actions go through a checking hook. Deleting or moving files, downloads and package installs are blocked in that mode. Manually selected models keep normal Codex permissions. Browser and desktop helpers return concise results instead of making the main model read huge interface dumps. Long tool outputs get shortened, with originals stored separately. For context compaction, the helper chooses a plan and Codex does the compression. Separate workers run their own conversations and return results and work logs. Background workers in the standard CLI are currently limited to tasks that don't write files. Then there's provider routing. My setup includes Codex, Google accounts, Opus, Gonka and Wally. I'm building quota-based fallback through OmniRoute while preserving conversation history. That part is still being finished; the distributable alpha starts with Codex. What did I actually pay? The $200 GPT subscription stayed. I also paid 8$ for eight Google accounts with 18 months of access, at 1$ each. Yes, from an unofficial seller. That's my purchase price, not Google's retail price. Gonka and Wally are free in my setup; Opus goes through Google. In this configuration, I wasn't paying an additional metered API bill. API-equivalent usage isn't money I actually spent. My actual expenses are the subscription and the account purchases above. This is still a working alpha, but it gets me much closer to covering my workload without constantly hitting the subscription wall. And I finally got to spend 5 billion tokens a week again I'm testing setup on other machines before releasing the code.

Prasenjit Sarkar

@stretchcloud

The thread going around this week about a startup fixing their site for AI agents (dropping captchas, exposing prices in the page HTML, adding a plain text copy of every key page) and seeing agent signups jump 10x says something true: most of the web still assumes its only reader is a human with a browser. I built DeepScrape because most sites sit on the other side of that gap. They were never redesigned for agents, and an agent that needs their data still has to render JavaScript, guess at a CSS selector, and hope the layout does not change tomorrow. DeepScrape generates the CSS selectors once with an LLM, then reuses them deterministically on every later pass, calling the model again only to self heal when the page actually changes. That is the same economic shift behind both stories: you stop paying a model to re-read the same page every single time. It ships an /api/map endpoint for discovering a site's structure, autonomous navigation toward a plain language goal, and an MCP server so Claude or Cursor can call a scraper as a tool directly. There is also a site-to-MCP mode that turns any existing website into a structured data source an agent can query, giving a site the agent-ready front door it never built for itself. github.com/stretchcloud/d… x.com/irabukht/statu…

Namn

@Naman_rai_

I built a tool that controls the Chrome you're already logged into. Works from Cursor, Claude, or a chat in the browser. You type what you want. It clicks the tab. This clip: Google Calendar invite, 38 seconds, no speed-up. Install the extension and try it , 2,000 free side-panel credits for new signups.

How they grewCursor, Claude, or browser chat; Chrome extension install

Aanya

@aanyakumarii

107 active users 🥹 Was 91 a few days ago. Tiny numbers, I know — but watching people actually use something I built is still pretty damn cool. Slowly getting there. 🚀 wordpopup.com

varun bhardwaj

@varunPbhardwaj

Stop paying $700/mo for 10 disconnected creator tools. Here’s how I built a 17-tool local AI content studio that turns 1 video into 6 assets for $0 cloud SaaS. #OpenSource #AI x.com/i/article/2104…

Salim

@aicreativmedia

got 50 users in few days on my new product the trick? i built an audience beforehand for it the truth? it makes signups much easier, but getting them to come back is still very hard

How they grewBuilt an audience beforehand

Wes Dalton

@makeAticket

I built Mise because my friends and family kept asking me to get them into restaurants in NYC. It started as a reservation hunter. Now it searches 5 booking platforms worldwide, connects to Beli, and lets you pass along reservations you can’t use. So far: 350+ users and 600+ tables booked. First table’s free: bookmise.com

Qubicle | Based Dept. Treasury 🏦

@cubeqube

thinking about shutting down a site I made for a videogame when I was 16-17 years old. got it up to over 100k verified users in a couple years, it was attacked by russian hackers and DDoSers, script injected, shown on a Twitch livestream by Lirik many years ago around 2014 and collapsed my website from traffic which initially helped it take off and taught me about query caching, firewalls, docker, php, laravel, bootstrap, etc. it did its purpose it landed me my first job in tech before i graduated high school. spent my entire summer break when I made staying up until 4am coding it and learning how to protect it, advertising it on steam forums, etc. good times. but it's time to shut that puppy down its now an outdated project and resume piece. godspeed

How they grewadvertising it on steam forums

Yahya Kisana

@yahya_kisana

In May I launched an app to get people to stop doom scrolling. I made it open-source and completely free. It's already grown to 2000+ users and has shown to be more effective than the paid apps, reporting 80+% reduction in screentime. Check it out: apps.apple.com/app/normal-scr…

Martins

@wogaam

first touch with @whop hasn’t been that bad despite not even having a real product with intention of marketing it or making something that adds real value, 34 users joined neoba well, 28 if I don’t count myself and those who work at @whop no one has actually use it, but as an experiment, it went pretty well I’d have liked to see how it would have gone if I had launched the creatives I made with chat gpt but it really surprised me how fast you can launch and market a product I mean, in less than 15-20 min I had a functional “neobank” with 3 different creatives ready to launch, and all through @whop hot take: the first post got 134 bookmarks and all I did was write a ridiculous prompt for Whop AI

How they grewWhop; first post with a Whop AI prompt