😭😭😭

no matter how bad it gets, God always comes through for me. can't even join that bandwagon of "you really believe there's a God?" from being debt-ridden to sitting on 4 figures in ~3 hours. i believe in God. it has worked, is working, and will keep working for me. talk soon.

smh

demo video piped.video/WzgAuy__2Qo?si=x2pz… website exit-keepa-web.vercel.app

‣ Exit Keepa finished top 10 in the @KeeperHubApp Agent Economy Hackathon. yesterday was my first finalist judging demo and i thought I’d blow it, but i actually did quite well. also got some good feedback from the team (found out Exit Keepa scored 92/100) we go again 🤝

which one of them was I referring to?

‣ out of the three models that dropped yesterday, you only need one. 1 — GPT-6 Sol - what OpenAI said it’s for everyday coding, agents, professional work, it's not the flagship, Astra is. - price vs last Sol cheaper by half - better than last Sol? yes on the jobs OpenAI said it's for, it's not smarter. - how i know 1 - factual errors : 8.4% → 4.5%. almost half 2 - coding and agent scores went up a bit. 3 - independent intelligence is mehh - 47 → 48. - rank in its field mid-tier, it is the cheapest coder in OpenAI’s stack but not the best. - worth switching from 5.6 Sol? yes . same intelligence and fewer mistakes for half the price, gud tek. - when to skip it hard coding, long reports, or “get this right first time.” (one shot ting complexing stuff), just use Opus 5.5 or Astra for those ones instead. - models still above it - GPT-6 Astra - OpenAI’s real top model and it beats Sol on almost every hard test and costs 5x. - Claude Opus 5.5 - actually smarter. AA Index ranks it 58 vs Sol’s 48. - Claude Fable 5.1 - slight edge on a few coding tests (DeepSWE ranks it 69.9 vs Sol's 68.8). not that superior 2 — GPT-6 Luna - what OpenAI said it’s for high volume work. summaries, extraction, quick questions. not for hard coding. - price vs last Luna cheaper by half. $0.20 / $1.20 → $0.10 / $0.50 - better than last Luna? no. same level of intelligence but cheaper. - how i know 1 - OpenAI says AutomationBench moved it up 5.4 points and the task costs less 2 - independent intelligence is flat. 37.26 → 37.32. 3 - knowledge-work scores went down . coding index 43 → 41 - rank in its field cheapest seat in the OpenAI stack. good for volume. not a coding model. - worth switching from 5.6 Luna? yes, only because it's cheaper, it doesn't have better output. - when to skip it coding, reports, or for one-shotting things. use Sol or Opus 5.5. - models still above it - GPT-6 Sol — same family, built for coding and agents - Claude Opus 5.5 — a totally different league - GPT-6 Astra / Claude Fable 5.1 — a different league too, don’t compare Luna to these unless you like getting embarrassed 3 — Claude Opus 5.5 - what Anthropic said it’s for long agentic coding, computer use, knowledge work. Fable-level quality at a lower price. - price vs last Opus cheaper. $5 / $25 → $4 / $20. - better than last Opus? yes. this is the only real upgrade out of these three i'm reviewing - how i know 1 - Terminal-Bench 4.0 - 52.3% → 66.4% 2 - Cursor Bench - 46.6% → 57.8%. Automation Bench - 26.9% → 40.0% 3 - independent AA Index has it at 58. number one right now 4 - Sonar - almost the same pass rate as Opus 5, but 27% less code and 40% fewer tokens - rank in its field number 1 for coding and long work but not the cheapest - worth switching from Opus 5? yes. better and cheaper. no reason to stay on 5. - when to skip it cheap volume work, just use Luna. if you have a tight budget, use Sol. but if you have stupid money and want to go all at anything, use Fable 5.1 or Astra. - models still above it - GPT-6 Astra — wins in some areas. science terminal bench ranks Astra at 64.6 vs 58.7 of Opus 5.5. - Claude Fable 5.1 — much more ahead. - GPT-6 Sol — not smarter, but it's faster and cheaper.

Knowing what runs in your family is useful but knowing what to actually do about it is better Preventah turns your family history into one practical action for today, with a simple check your household can see try it - qualified-hummingbird-614.co… #AllGasHackathon

spun up the entire @thenotepadlabs website on Claude with just one prompt stay tuned.

ok

done

😂😂😂😂😂😂😂😂

son?

Google is giving away tools people still pay $20–50/month for. most people only know Gemini but here are 10 free Google AI tools worth actually using. Pomelli - labs.google/pomelli paste your website. it builds a brand profile then generates on-brand posts, campaigns and assets. Stitch - stitch.withgoogle.com describe an app or site. it gives you UI + frontend code and exports to figma. Opal - opal.withgoogle.com build AI workflows and mini-apps in plain english. no-code n8n with gemini baked in. Antigravity - antigravity.google google’s agentic IDE. you describe the feature, agents plan and write it. Mixboard - labs.google/mixboard canva meets pinterest. generate, combine and edit AI images on one board. Disco - labs.google/disco turns your open tabs into custom apps. waitlist and macOS-first right now. NotebookLM - notebooklm.google upload your pdfs, docs or youtube. get summaries, quizzes, mind maps and audio overviews from your sources only. Learn Your Way - learnyourway.withgoogle.com turns a topic into a personalized course. text, audio, slides, quizzes. Google AI Studio - aistudio.google.com the free gemini playground. build, test and run AI apps in the browser. Jules - jules.google asynchronous coding agent on github. assign a task, it plans and opens a PR. free tier is 15 tasks/day.

AI doesn't need human data anymore. for a long while, to scale AI meant feeding models more human data and information. books, code, web pages, images and human conversations. but the problem with that is humans can't consistently produce high-quality data. and this is where Recursive Self-Improvement comes in. an AI system that can generate training data, write code, run experiments, evaluate results, and use what it learns to improve the next version of itself. then that new version does it again. and again, and again. model improves → better model helps → create better training → better training produces a better model → repeat not trying to sound cool but this doesn't even excite me, I want to know if each improvement makes the next improvement easier. because if it does, these models will be able to develop themselves independently and if they can do that, then.....

iffa slap two of you

what killed your parents will kill you if you don't actively work towards preventing it. that's why I built Preventah on @nimiq it turns hereditary health risks into a daily prevention routine • personalized plan • check-ins • measurements • reminders • USDT commitment try it - preventah-nimiq.vercel.app

The Clarity Act failed because of Trump. no need to start pointing fingers at anything or anyone. I say this because the Clarity Act survived - a 294–134 House vote. - a 15–9 Senate Banking Committee vote. - more than 100 Democratic edits. - a 600-page compromise. and it couldn't still get 60 votes yesterday. why? because you're asking senators to pass a law regulating an industry while the sitting president had disclosed more than $1.4 billion in crypto-related income for 2025 and still had major financial interests tied to crypto. simple as that. i've read so many think pieces and all of them are either nitpicking or ignoring the Trump in the room (there's no fkn elephant, it's trump) “It's the stablecoin yield.” “It's the banks.” “It's CFTC vs SEC.” “It's the state AGs.” “It's the midterms.” all of them weighed a lot and pulled their weight and that's how negotiations work. but Trump is really what nerfed us even Sen. Mark R. Warner said it plainly after voting NO “we cannot pass landmark legislation governing this industry while allowing the President of the United States to personally profit from it.” and he said the failure to address that conflict of interest made it impossible for him to support moving forward. so you can go ahead and say whatever about the Democrats but they are fucking right. and we can't even say "The Congress hates Crypto" this same Congress passed the GENIUS Act and The House passed Clarity by 160 votes. this is the farthest we've ever gotten to in Washington, and Trump ruined it.

claimed my creator fees on @ponsdotfamily for launching a meme. also 11% up in profits on the meme itself gud tek.

‣ contrary to public opinion, a lot of people are waiting for ARC. to get USDC on Arc costs 1.8× the normal cost plus a 3% fee, and as we speak @unstable_eco has already settled 11,684 trades for $5.46M USDC, and the live price is still around 1.91×–1.95×. ‣ private mainnet also pushed 3x the transactions it did the day before Sept 11 → 120k txs Sept 12 → 126k Sept 13 → 131k Sept 14 → 407k and 116k new addresses were recorded in 24h. all on the private mainnet. ‣ lastly, Uniswap, Bitrue and Gate have confirmed day one support for Sept 16. same for edgeX, Fomo, Binance Wallet and OKX Wallet. Arc is coming.

made me giggle lol

there’s a best AI tool for almost every job. so i mapped out the category leaders. we all agree with this yeah?

W analysis

‣ too many things happening at once - apparently everyone and their grandma is making money on memecoins - AI is moving faster then we expected and can nerf humanity in 6 to 12 months - 2,000 new launchpads in crypto everyday can we just press pause on everything?

‣ The CEO of Anthropic just said that within 6–12 months, an AI agent swarm with stronger capabilities could potentially take over the internet. yes. the co-founder of Anthropic is saying the timeline for some AI risks may be measured in months, not decades. but that's not even the reason he wants to slow things down. it's because AI is increasingly being used to help build better AI. so if that loop keeps accelerating, our ability to understand and control these systems has to accelerate even faster. otherwise, we may keep trying to put brakes on a car that is learning how to build a faster engine while it's driving. are you ready?

‣ last payout on X just came in earned 3 figs from it gud tek

always knew you were the lost long son of a Saudi Billionaire

‣ There's $7 million waiting for anyone smart enough to claim it. it's been sitting there for 26 years but only one person has ever managed to solve 1 of the 7 problems behind it. yes. in 2000, the Clay Mathematics Institute selected seven mathematical problems considered among the hardest problems in the world and put a $1 million prize on each one. I went through all 7 this morning and I understand why they haven't been solved. 1 - Poincaré Conjecture SOLVED. Grigori Perelman solved it in the early 2000s. It was the first and, until recently, only Millennium Prize Problem to be solved. 2 - Riemann Hypothesis UNSOLVED. proposed in 1859. it deals with a hidden pattern in the distribution of prime numbers. mathematicians have spent more than 160 years trying to prove it. 3 - P vs NP UNSOLVED. a surprisingly simple question “if we can quickly check whether an answer is correct, can we also quickly find the answer?” a YES or NO answer to this question will change computer science forever. 4 - Yang-Mills and the Mass Gap UNSOLVED. Yang-Mills theory helps physicists describe some of the basic forces of nature. it works extremely well in physics, but mathematicians still can't prove why it should produce particles with mass. that missing proof is the mass gap problem. 5 - Navier-Stokes CLAIMED, NOT SETTLED. these equations describe how fluids like water and air move. the question is whether fluid motion can become so extreme that the equations break down. this week, OpenAI published a claimed proof showing that such a breakdown can happen. the proof was also checked with Lean, a system that verifies mathematical steps. but the Clay Institute has not accepted it, so the problem is still officially unsolved. 6 - Hodge Conjecture UNSOLVED. this is about the hidden structure of geometric shapes. Mathematicians can detect certain shapes using equations, now The Hodge Conjecture asks whether those shapes must also correspond to actual geometric pieces. some people have made good attempts at this but the questions haven't been solved. 7 - Birch and Swinnerton-Dyer UNSOLVED. this problem is about a type of equation called an elliptic curve. I.e - y² = x³ + x + 1 can have some solutions where both x and y are simple fractions or whole numbers. the problem asks whether we can figure out how many of these solutions exist by studying the equation itself. no Mathematician has an answer for it yet.

‣ we are beyond cooked at this point because now the scary part about AI isn’t that it can hack. it’s that you no longer need to be a sophisticated hacker to run a sophisticated operation. holy fuck.

‣ finding good hackathons is becoming a hackathon itself. @devpost has some, @DoraHacks has some , @devfolio has some @ETHGlobal has some. then there are bounties, grants, builder programs and AI competitions scattered across even more platforms. so I went looking for the tools trying to fix that problem. here are 7 hackathon aggregators you need to never miss a hackathon. 1. HackListHQ - hacklist.io AI + Web3 focused. It pulls together hackathons, bounties, grants, builder programs and jobs, then sorts them around prizes and deadlines. Best if you spend most of your time in Web3 or AI and don't want to check five different boards. 2. HackOdds - hackodds-live.vercel.app probably the most interesting one. it pulls events from Devpost, DoraHacks, Devfolio, Hackquest, HackList and others, then ranks them using prize money relative to registrations. in simple terms, it helps you answer "Is this hackathon actually worth entering?" you can also connect GitHub for a personalized watchlist and ranking based on the languages in your public repos. 3. HackathonRadar - hackathonradar.com The wide-net option. It pulls from 50+ platforms, including Devpost, Devfolio, MLH, ETHGlobal, Unstop, Kaggle, Hack Club and DoraHacks. ut also has quality scores, a map, API, MCP and webhooks. 4. EntrantHub - entranthub.com goes beyond hackathons. it also covers coding contests and AI/ML competitions across multiple platforms. 5. Hack Club Hackathons - hackathons.hackclub.com this one is more niche. it's focused mainly on free, student-led hackathons, with filters for location and virtual events. It also has a public JSON API, so the data it has is more useful beyond the website itself. 6. Hack Events - hackevents.net a broader tech-event search engine covering hackathons, meetups and conferences. it has an API, calendar and MCP, so it's more useful for agents and automated workflows. 7. Hackamaps - hackamaps.com this one takes a different approach. it's map-first. if you care about physical hackathons and want to plan travel around where builders are meeting, this is the most useful one for that

just got my @world_xyz airdrop gWorld.

bruv will rest it till they deploy fixes

this isn’t new or even breaking news. Stuart Russell already said this to @StevenBartlett on Diary of a CEO last year. the people at the forefront of AI already know it can kill everyone, they just don't care. and yes, we are the last generation to experience what life felt like before AI.

mfer switched up on us

the only people that came close they were all generated with GPT 2.5 none of them are real. it was fun doing this with you all

1 of these 4 images was generated with GPT 2.5 first to find the generated image gets a 100$

these are 10 repos people use to make money while they sleep. i've seen so many mid repos circling about so I decided to bring in the leaders of each money making category. 1 - n8n is an open-source workflow automation that builds custom automations for businesses and all you have to do is charge for setup and maintenance. github.com/n8n-io/n8n 2 - TradingAgents is a multi-agent trading framework with separate roles for analysis, trading and risk management. github.com/TauricResearch… 3 - Freqtrade a mature open-source crypto trading bot with backtesting, optimization, live trading and Telegram support. github.com/freqtrade/freq… 4 - MoneyPrinterV2 it automate parts faceless content workflow, from generating topics to producing videos. github.com/FujiwaraChoki/… 5 - Medusa open-source Shopify, and this time you keep the margin instead of giving it to a platform. github.com/medusajs/medusa 6 - Cal.com is an open-source Calendly. you can fork it, white-label it, and charge businesses a monthly fee to handle their booking. github.com/calcom/cal.com 7 - Plausible is a privacy-first Google Analytics alternative. simply self-host it and resell hosted analytics to agencies per client every month. github.com/plausible/anal… 8 - Ghost is an open-source Substack. you keep 100% of membership payments instead of giving a cut to the platform. github.com/TryGhost/Ghost 9 - Dify is an open-source platform for building AI agents and chatbots. self-host it, white-label it, sell the product and keep the margin. github.com/langgenius/dify 10 - Postiz it’s an open-source Buffer. people can host it for free, while you charge for a managed version with a monthly subscription. github.com/gitroomhq/post…

yay?

whether you are a market maker whether you are on-chain or off-chain whether you have the strategy or the capital doesn't matter, Just Flint It.

You already have the recipe. Flint has the restaurant. Bring the strategy, Bring the capital and leave the infrastructure to Flint.

neat

A hedge can look good on paper and still leave you exposed. Watchman checks the actual fill against the position and shows whether it really offset the loss. watchman-beta.vercel.app

bro thought we were gonna miss this

thesis confirmed, World is coming 9/9 be there.

‣ there’s no airdrop. go home to your families on 9/9, Solana will be reducing account rent, basically like a large-scale @solincinerator event. unused token accounts have SOL locked in them will be freed, and about ~3M SOL could become claimable. if $SOL stays around $100, that 3M SOL = $300M+ that's what is happening, there's no airdrop. they just took that rumor and ran with it, and the timeline is running with them too. ‣ the 9/9 countdown they’ve been pointing at September 9 since August. so that timing is adding even more FOMO to it. W marketing and i respect it. ‣ here are the facts I’m basing my prediction on - World is a real product, and an onchain prediction market at that. - there is no official token. No TGE is in sight, no disclosed treasury, and no tokenomics from the team. - all $WORLD coins on Pumpfun are copycats. - there’s no public raise we can tie to them that supports a $319M airdrop. - “they followed 4,000 accounts, so those wallets will get the drop.” my brother in Solana, no team in their right mind allocates nine figures worth of token supply based on who they follow. wake up, 2021 is gone and not coming back. ‣ again, here are my predictions 1 - 9/9 is a product event they’re building momentum toward, with CT hoping for an airdrop. 2 - a standalone app or public waitlist opening. 3 - a points program. Season 0 of God knows how many seasons. 4 - maybe a token or TGE tease. 5 - even if they do drop a token, there will be criteria. for example, you might need to have traded a World market on Phantom, be a regular Phantom user, be on the waitlist, hold an affiliate badge, or meet some other criteria that nerfs 90% of the people waiting for the airdrop. ‣ what happens to prices? - the copycat $WORLD coins will pump, devs and insiders win again, while you and I become exit liquidity, as always. - SOL won't reprice because of this. that will be all, thanks for coming to ma TED Talk.

‣ Happy Sunday - lost a 3-fig deal because I didn't have PayPal - @ChogNFT FP is holding up nicely (Benja, send it haya) - building 3 hackathon submissions and debugging another 3 (send help) - X brought content scheduling to mobile (yay!)

minted my Chog. again, ty to @crexsol and @1stBenjaNAD for the opportunity. Chog World Order.

so fucking pissed

ok bro

by what standards?