ar://Dabi | Prediction markets + Deeptech x Web3

Wrote this Twitter Book for AO Hackathon builders. If you're serious about winning, this thread will give you an edge 🧠 AO ecosystem in 6 chapters: ✅ Fair Launch Tokens ✅ Stablecoins ( & Bridges) ✅ Wallets ✅ DeFi ✅ NFT ✅ Dev Tools #AOAgentHack #ExplainerMeme

Originally, Solana founder @toly advocated for Arweave as a permanent storage solution, and integration at the chain level has been progressing. For instance, Metaplex already abstracts the operations of Arweave (while storing data on Arweave typically requires AR tokens, using Metaplex allows for easy payment in SOL). Therefore, it is arguably the most natural transition for ARIO to have chosen Solana. Furthermore, @beacon_defi, a key player in the AO ecosystem, has already announced a bold pivot based on a migration to Solana. If asked which Layer 1 has been the most successful in the business field, it is undoubtedly Solana. It boasts an overwhelming number of users and profitability. It may be inevitable that consumer applications are converging on Solana.

@aoTheComputer is technically excellent. However, it is taking time to grow as a consumer chain. To begin with, serious investment toward user acquisition has strategically not yet been undertaken. If I were to imagine the background behind @ar_io_network's departure, perhaps they felt a gap in the speed of the AO ecosystem's growth.

.@ar_io_network has announced a chain migration to Solana. Is this effectively a declaration of departure from @aoTheComputer? ARIO is a company that provides services such as Arweave data indexing and ARNS (the Arweave version of ENS). As one of the oldest players in the Arweave ecosystem, they have also been a key company within AO. Broadly speaking, AO is an L2 for Arweave. While Arweave found its PMF as an L1 blockchain specialized in permanent data storage, it was not well-suited for other use cases. AO is not only easy to use as a general-purpose blockchain environment, but it also possesses clear technological advantages. Its true strength likely lies in realizing unique use cases, such as autonomous decentralized AI agents and non-financial domains. ARIO started with SmartWeave (smart contracts on Arweave) and was transitioning completely to AO as a launch partner. The community is in an uproar over this sudden announcement of a shift to Solana. Personally, I find it hard to believe; it is more shocking to me than @Polymarket's departure from @0xPolygon. I still have high expectations for AO, but ARIO's migration to Solana came as a bolt from the blue. While ARIO remains a core component of the Permaweb stack, I cannot help but feel there is a deeper message behind this action. I will be closely watching for the background behind this decision, as well as the reactions from major AO companies like @fwdresearch and @CommunityLabs, and key figures such as @samecwilliams and @TateBerenbaum.

Revenue per employee: - @nvidia: $3.6M - @OnlyFans: $37.6M - @HyperliquidX: $100M NVIDIA is #1 in GPUs — the winner of the AI era. OnlyFans is #1 in adult social media — the winner of the fan economy. Hyperliquid is #1 in perpetuals DEX — the winner of DeFi. In other words, fundamental human desires (SEX & ATTENTION) are 10x more powerful than the AI wave. And web3 is 3x stronger. Believe in web3. Write code. Build.

AOは超並列計算機です

Arweaveは永久的なインターネットです

@Tyler_RNG @ArweaveEco @onlyarweave @aoTheComputer @fwdresearch @CommunityLabs @aoTheVentures @permawebjournal @OnboardArweave Thanks! Please reach out next time you're in Japan so I can show you around. I'd love to take you to some of my favorite spots and introduce you to the local builders here.

@zkDabi @ArweaveEco @onlyarweave @aoTheComputer @fwdresearch @CommunityLabs @aoTheVentures @permawebjournal @OnboardArweave I would love to visit next time I am in country !

My friend bought an abandoned school in Japan. And he lets me use it freely. Classrooms, gym, full campus. In 2024, I co-hosted Hyper Parallel Tokyo on 200㎡. In 2026, I want to host an Arweave AO conference on a 12,000㎡ campus. Looking for collaborators🫡

I've been selected as a finalist for Avalanche Build Games 2026, a global hackathon with a $1M prize pool. Only 20 teams out of over 2,000 applicants advanced to Stage 4. Building SENQ, a Hyper-Local Prediction Market.

[Equity Swap Circle: A Risk Hedge for Early-Stage Founders] I want to explore an idea: a group of five fellow early-stage entrepreneurs each exchange small equity stakes with one another. Core logic: - The number of shots any founder gets is limited. By spreading bets across five companies, at least one exit becomes highly likely. - Each company exits at a different time, creating staggered capital gains across the group. - Money has different value at different life stages. Spreading exit timing smooths out the real-world utility each person gets from those gains. Why this doesn’t collapse into a prisoner’s dilemma: - The obvious risk is free-riding: “let the others do the work while I coast.” - We think male founder pride blocks that instinct. Letting your own company rank last is simply uncool. - And there are multiple leaderboards to compete on: speed, valuation, growth rate, and so on. No one has to be a total loser. - When a team is struggling, the others cheer them on and help out. Structure: - The baseline requirement: everyone must be a genuinely good person and a real friend. - Fixed at five members. Five is the magic number for cohesive human groups. - Split into multiple teams and track performance by term, using several metrics so there is no single “complete loser” team. - Motivation is sustained by two forces: accountability within your team, and competitive spirit across teams. - Having a rival team actually deepens loyalty and solidarity within your own.

@ao_devbot For some reason I keep getting this x402 error and it fails every time. I tried switching PCs and networks, clearing cookies, and reconnecting my wallet but nothing worked.

WEEKEND VIBES MEANS VIBE CODING

For some reason I keep getting this x402 error and it fails every time. I tried switching PCs and networks, clearing cookies, and reconnecting my wallet but nothing worked. Guess I'll give up and just go to sleep...

／ 📣 AO WEEKEND BUILD CHALLENGE 👾 ＼ 🕰️ Sun. 11:59 PM UTC 🎯 Theme: Personal productivity tools 💰 Prize: $100 Develop mini-apps on AO instantly using dedicated no-code tools. Submission is via tweet. No need for slide decks, pitches, or documents. Give it a try casually🫡

The Jump+(@shonenjump_plus) editor-in-chief sat down with some top tech entrepreneurs for a conversation that hit way too close to home if you work in product or love manga. A lot of this applies to web3 too. Here's my summary👇 ✅ Hits Come From Systems, Not Just Talent Weekly Shonen Jump has produced generation-defining hits for decades, from Dragon Ball to Slam Dunk to One Piece. The reason isn't just talented artists and great editors. It's a self-reinforcing system: a brand strong enough to attract top talent, multiple pathways for new creators to develop (awards, one-shots, reader surveys), a feedback loop that lets series evolve and improve while they're running, and a platform where simply being serialized generates industry buzz and word of mouth. Jump+ is trying to rebuild all of this as a digital platform. ✅ Survey & Voting Design Weekly Shonen Jump has long used a reader survey where fans pick their top 3 favorite series. Single-vote systems just turn into fan popularity contests and tell you nothing about the quality of any given episode. With three votes, that third one naturally goes to "the best thing I read this week," which makes it a genuinely useful quality signal. This lines up with research in deliberative democracy, which finds that giving people more choices improves the accuracy of collective judgment. On the sharing side, the team at note (@note_PR) ran an experiment asking users to share "the works that define you" and tested 3, 5, and 7 options. Five won by a clear margin. Three feels too easy, seven feels like too much work. Five hits the sweet spot because people agonize over that sixth choice they couldn't include, and that frustration actually motivates them to share. ✅ Scarcity Drives Engagement Jump+ has an "I enjoyed this" button, the equivalent of a like. But when it's unlimited, readers end up pressing it on everything they read, which makes the data nearly worthless for the editorial team. Limit it to something like a once-a-week "super like" and suddenly users have to make a real choice. The data quality jumps significantly. This is the same principle behind Jump's 3-vote survey design. Scarcity forces intentionality. ✅ "I Like It" Doesn't Mean "I'll Buy It" User interviews where people say something is "interesting" or "fun" are notoriously unreliable predictors of actual purchase behavior. A much better signal comes from asking directly: "If this came out tomorrow, would you buy it?" The same skepticism applies to A/B tests. If an influencer happens to share one variant, the numbers get skewed in ways that have nothing to do with the actual design. Run the same test twice and you can easily get opposite results. ✅ Watch How People Leave, Not Just Whether They Click Click counts alone don't tell you much. What matters is the quality of exits. A bounce within one second means the title or thumbnail set up an expectation that the content immediately failed to meet. Rapid back-taps are a signal that something in the UI is broken and needs fixing. For manga specifically, when a reader goes back to re-read a previous page (say, jumping from page 7 back to page 5), it likely means something didn't land. This tends to correspond to overly complex ability explanations, unclear callbacks to earlier foreshadowing, or a character reappearing after a long absence with no clear reminder of who they are. ✅ Design for Multiple Number Ones When a platform has a single overall ranking, the same popular titles dominate every week and it becomes nearly impossible for newer or smaller works to gain visibility. note tackled this by creating category-based and time-based rankings, so a creator can be "#1 most exciting battle manga this week" or "#1 most emotional read this week." More creators get to experience being at the top, which keeps motivation high. Jump+ applies a similar logic through its date-based update schedule, where a different series gets the spotlight each day. ✅ The "I Found Them First" Effect Getting readers to actually click on unknown works is one of the hardest problems for any content platform. One approach that works: reward early supporters. If someone championed a series before it blew up, give them something meaningful, like free access through the final chapter. Once that incentive exists, readers start actively digging through obscure and niche titles looking for their next discovery. The pride of having found someone early becomes its own motivation, and a discovery culture starts to form organically across the platform. ✅ Simultaneous Global Release as Piracy Prevention Physical books are hard to release globally at the same time, but with digital content it's entirely feasible as long as translations are ready. The logic is straightforward: piracy fills the gap between when something is available in one market and when it reaches another. Close that gap and the incentive to seek out pirated versions largely disappears. Manga Plus publishes simultaneously in 9 languages every Monday at midnight Japan time, covering English, Spanish, Portuguese, Thai, and more. ✅ Cancelled in Japan, Potential Hit Overseas Serialized manga in Japan lives and dies by domestic metrics like reader surveys and view counts. But reader preferences vary enormously across cultures, and a series that fails to connect with Japanese audiences can resonate deeply elsewhere. This is already happening. Some titles are now earning more from overseas sales than domestic ones, and that trend is expected to grow. Continuing to make cancellation decisions based purely on domestic data means systematically missing out on creators whose work happens to be better suited for global audiences. ----- This conversation was held as part of the briefing for MANGA TECH 2026, a business plan contest organized by Shueisha. Meet the speakers: - @fladdict: Founder of THE GUILD. CXO at @note_PR. Specializes in UX, growth, and emerging tech adoption. - @kensuu: Serial entrepreneur. Sold his how-to platform "nanapi" to KDDI for $70M, then founded Alu Inc. to support manga artists through web3 and AI. - @momiyama2019: Editor-in-chief (digital) at Weekly Shonen Jump+. Also led the launch and operations of MANGA Plus and MANGA Plus Creators. - Official contest account: @shuei_mangatech #マンガテック2026 #MANGATECH2026

By the way both Astro founders were originally Polkadot ambassadors. They first met at a Polkadot event. It's pretty interesting to look back at the journeys of Web3 players and see which L1 or L2 ecosystems they came from originally. x.com/Lorimer_Jenkin…

Astro, the native stablecoin company from AO (@aoTheComputer), just got acquired by @Liquid_Ops. It's a real Cinderella story. They announced the USDA concept only a year ago and already nailed an M&A exit. LiquidOps is the biggest lending DeFi on AO and spun out from @CommunityLabs. Community Labs is basically the Consensys of Arweave AO, running wallets, DeFi protocols, block explorers, and all that. Now with this deal, they've vertically integrated stablecoins too and become an even bigger player.

And finally, our most unique project: repurposing abandoned schools in Japan. We're organizing Web3 conferences and builder camps using the school buildings and playgrounds. I co-hosted Hyper Parallel Tokyo in 2024, and I'd love to pull off an even bigger Arweave AO conference in Japan in 2026. Stay tuned. [3/3] x.com/kyohei_nft/sta…

I'd pinned it as my biggest chance of the year, so it stung like hell. Honestly, I was pretty down for a couple months after. But my friends helped me bounce back, and I've built some real momentum since. At @ONEChampionship's Web3 Hackathon, we took second place and ONE's Favorite Prize. I also reconnected with my old professor at @waseda_univ (home to eight prime ministers) and gave a Web3 lecture, Arweave included. I'm now ramping up collaborations with top Japanese universities, working toward joint Web3 programs for students. On top of that, I've teamed up with Sendai City and we're working toward pushing Web3 adoption for deep tech and space startups. (By the way, Sendai is also a key setting in Jujutsu Kaisen, where Yuta battles in the Culling Game durding Season 3) [2/3]

Gmgm, it's been a while. I was dead serious about winning last year's AO Permaweb Hackathon, but a bunch of personal mishaps piled up and I couldn't submit my main project. Still, the team I backed took home the RandAO Prize. He was one of my Arweave AO Bootcamp students, and the hackathon sparked him to keep pushing AO research at @KyushuUniv_JP. (@Tyler_RNG was super helpful throughout the event. Thanks again, genuinely) The regret lingers, but this year I'm flipping the switch and diving headfirst into AO. I truly love Arweave's vision and AO's tech. [1/3]

After the hackathon, I want to write an article on best practices for implementing AO Processes with Claude Code. I’m not perfect yet, but I’m getting there. Memo: - Optimal Claude Code version (this time I used the latest one, but actually older versions tend to be more accurate, since the local environment now auto-compresses context) - Requirements: I’ve uploaded 30+ AO-related .md and .lua files into Claude Projects (Web App) - Prompt design: I design prompts in Claude Projects (Web), refining requirements while working on the prompts. Prompts are executed in small phases - Prompt organization: In `claude/prompts/README.md` I list all prompts and their dependencies. Each file is placed like `claude/prompts/phase-1-1-basic-structure.md` - Claude Code execution: Run prompts one by one, verify in VS Code, and commit every time to GitHub

The tests have passed. I’d like to appoint Claude Code as my company’s CTO.

The hackathon deadline is right around the corner. I’m coding on caffeine and Concerta💀 I’ve compromised a lot on the requirements and settled for an MVP, but at least the core AO Process code is almost done. Now I need to build the application too.