stormrex.eth (Ø,G)

Alpha hunter
learn | adapt | grow

super bullish on @ohbabygames

[stormrex.eth (Ø,G) @stormrexeth]

@GenLayer Check this out
https://t.co/VCLNyqW2vh

[the doc 🫎 @thedocx70]

(A quick update from my first post to highlight the new features and the updated smart contract address!)

​For a while now, I have been thinking of something to build on the @GenLayer Bradbury Testnet. Being someone that really likes games, I remembered one we used to play back in secondary school. It was more like a Naija street game back then.

​The name of the game is Whot.
​With that in mind, I built a real-time, decentralized multiplayer version of Naija Whot. No rigged decks, no central referees. Every single card played is validated by an Intelligent Contract on GenLayer’s Virtual Machine.

​🔗 Live Demo: https://t.co/8pU5dITcOh
🐙 GitHub: https://t.co/0KaZP6kVX8…
​The Architecture & Stack
• Frontend: Next.js + Tailwind
• Engine: Node.js + https://t.co/YnR9UKKXmm
• Blockchain: GenLayer SDK + Python Smart Contract
​Under the Hood: Building Hybrid Web2/Web3
Most web-based card games rely 100% on a centralized backend. I wanted to use a smart contract as the ultimate source of truth, but without sacrificing the real-time, low-latency speed of a WebSocket multiplayer game.
​Two major technical hurdles to make this work:
​1️⃣ The Calldata Encoding: I initially tried firing raw HTTP fetch requests directly to the GenLayer RPC but kept hitting EOF decode tx CallData errors. The VM expects strict binary calldata. Moving to the @genlayer/genlayer-js SDK handled the payload formatting and solved it instantly.
​2️⃣ Concurrency & Race Conditions: I had to handle multiple players dropping cards at the exact same millisecond. I implemented a strict isProcessingMove room lock via https://t.co/YnR9UKKXmm on the backend to block conflicting actions while the contract validation is in flight.

​🚨 What's New: The On-Chain AI Referee Upgrade
Validating card numbers is just basic math. So, I upgraded the contract to use GenLayer’s native LLM features. Now, using the Non-Comparative Equivalence Principle, the smart contract actively watches the game. When someone drops a heavy Action Card (Pick 2, Hold On), the blockchain dynamically generates authentic Naija street trash-talk and broadcasts it to the players in real-time.
​Really proud of how this architecture turned out. The upgraded smart contract is live, verifying moves, and trash-talking right now on the Bradbury Testnet: 0xf01124a80C383dCF576892Db44F662437d0309B7

​Huge shoutout to the @GenLayer team for the tech. Grab a friend, jump in a lobby, and have fun 🤝
​Check out the video of the updated gameplay!
@RuzgarFlns @Aezakmi_x

[the doc 🫎 @thedocx70]

(A quick update from my first post to highlight the new features and the updated smart contract address!)

​For a while now, I have been thinking of something to build on the @GenLayer Bradbury Testnet. Being someone that really likes games, I remembered one we used to play back in secondary school. It was more like a Naija street game back then.

​The name of the game is Whot.
​With that in mind, I built a real-time, decentralized multiplayer version of Naija Whot. No rigged decks, no central referees. Every single card played is validated by an Intelligent Contract on GenLayer’s Virtual Machine.

​🔗 Live Demo: https://t.co/8pU5dITcOh
🐙 GitHub: https://t.co/0KaZP6kVX8…
​The Architecture & Stack
• Frontend: Next.js + Tailwind
• Engine: Node.js + https://t.co/YnR9UKKXmm
• Blockchain: GenLayer SDK + Python Smart Contract
​Under the Hood: Building Hybrid Web2/Web3
Most web-based card games rely 100% on a centralized backend. I wanted to use a smart contract as the ultimate source of truth, but without sacrificing the real-time, low-latency speed of a WebSocket multiplayer game.
​Two major technical hurdles to make this work:
​1️⃣ The Calldata Encoding: I initially tried firing raw HTTP fetch requests directly to the GenLayer RPC but kept hitting EOF decode tx CallData errors. The VM expects strict binary calldata. Moving to the @genlayer/genlayer-js SDK handled the payload formatting and solved it instantly.
​2️⃣ Concurrency & Race Conditions: I had to handle multiple players dropping cards at the exact same millisecond. I implemented a strict isProcessingMove room lock via https://t.co/YnR9UKKXmm on the backend to block conflicting actions while the contract validation is in flight.

​🚨 What's New: The On-Chain AI Referee Upgrade
Validating card numbers is just basic math. So, I upgraded the contract to use GenLayer’s native LLM features. Now, using the Non-Comparative Equivalence Principle, the smart contract actively watches the game. When someone drops a heavy Action Card (Pick 2, Hold On), the blockchain dynamically generates authentic Naija street trash-talk and broadcasts it to the players in real-time.
​Really proud of how this architecture turned out. The upgraded smart contract is live, verifying moves, and trash-talking right now on the Bradbury Testnet: 0xf01124a80C383dCF576892Db44F662437d0309B7

​Huge shoutout to the @GenLayer team for the tech. Grab a friend, jump in a lobby, and have fun 🤝
​Check out the video of the updated gameplay!
@RuzgarFlns @Aezakmi_x

[stormrex.eth (Ø,G) @stormrexeth]

@supraEVM Insane

[Supra 𝔖 @supraEVM]

🚨 META JUST ACQUIRED ANOTHER STARTUP

Meta has acquired humanoid robotics startup Assured Robot Intelligence (ARI), although the deal amount wasn’t disclosed.

ARI focuses on building advanced AI models that allow robots to understand and adapt to human behavior in real-world environments. Their goal is to make robots capable of handling everyday physical tasks like household chores.

As part of the acquisition, ARI’s team including co-founders Xiaolong Wang and Lerrel Pinto (sold his previous company,Fauna Robotics to amazon) will join Meta’s AI division, specifically its Superintelligence Labs research unit. Both founders have strong backgrounds in AI and robotics, with experience at places like Nvidia, UC San Diego, and NYU, along with multiple industry awards.

This move strengthens Meta’s push into humanoid robotics. The ARI team brings deep expertise in building AI systems that can control full-body robot movement and enable self-learning through real-world interaction.

Meta has been quietly working on humanoid robot technology for years, and earlier reports suggest they’ve been exploring both the AI models and hardware needed for consumer-ready robots.

Even if Meta doesn’t release a humanoid robot anytime soon, many experts believe that achieving true advanced AI (AGI) will require systems that can learn in the physical world and not just from data which is where robotics plays a key role.

META has been bleeding on this AI race. I hope it works out for them, it will be interesting to see how it goes

[Supra 𝔖 @supraEVM]

When you are going through another bear market but this time you have experience 😌

[stormrex.eth (Ø,G) @stormrexeth]

@supraEVM @augustus_bank Thank you for sharing

[Supra 𝔖 @supraEVM]

EARLY CRYPTO PROJECT [AI]

@augustus_bank

Augustus (formerly Ivy) is building a stablecoin bank for the AI era.

They just got approval from the U.S. government (OCC banking license) to become a real national bank, which is a big deal because it gives them deeper access to the U.S. financial system than most crypto companies.

Their goal is to make cross-border payments instant (24/7) using stablecoins + AI, replacing slow traditional banking systems.

They already process billions in euro transactions in Europe for companies like Kraken and recently raised $40M from investors like Valar Ventures, plus founders of Ramp, Deel, and Circle.

Think of it as a bank built for AI agents and programmable money

[stormrex.eth (Ø,G) @stormrexeth]

@supraEVM @Recursive_SI This is alpha

[Supra 𝔖 @supraEVM]

What Happens When AI Starts Building Better AI?

Early AI Project: @Recursive_SI

Recursive Superintelligence, Inc. is a new AI startup founded in late 2025 (barely 5 months old) that wants to build AI that can improve itself without humans constantly helping.

It was founded by Richard Socher and @_rockt, with a small elite team of ~20 researchers (@timshi_ai) from companies like OpenAI, Google DeepMind, and Meta.

Their goal is to create self-improving AI that can research, test ideas, learn, and upgrade its own intelligence in a loop (often called an “intelligence explosion”).

The company already raised $500M+ (possibly up to $1B) at a $4B valuation, backed by investors like GV (Google venture) and Nvidia.

Right now, it’s still in stealth mode/high hype phase, with a public launch expected around mid-2026.

I also found another account (@recursivesi) at 0 followers but not sure if it's related.

[stormrex.eth (Ø,G) @stormrexeth]

@supraEVM @X This is amazing. Thanks for sharing

[Supra 𝔖 @supraEVM]

Jack Dorsey [ex @X (Twitter) co-founder]'s Block,Inc. (formerly Square) just launched Goose, an open-source Al agent that serves as a free alternative to premium coding tools.

The project has quickly gained traction, surpassing 38,000 GitHub stars and attracting over 400 global contributors.

Goose is model-agnostic, meaning it can run locally with any LLM to handle complex tasks like code migration and debugging:

- Works with any LLM (Claude, GPT, Gemini, Llama, DeepSeek)
- Reads and edits your entire codebase
- Runs shell commands and installs dependencies
- Executes and debugs code automatically
- Desktop, CLI, and web interface
- Written in Rust.

And no bloat, check it out: https://t.co/ea3nhRQp8w

[Supra 𝔖 @supraEVM]

EARLY ALPHAs

Found some interesting projects that I am keeping an eye on.

@OCGamesdotio [81 followers]
OCG is a fully on-chain gaming platform on AbstractChain enabling T1/T2 games with instant settlement.

@TerafabAi [1 follower]
This might be the new account for upcoming project (company) by @elonmusk

@playteddywins [1 follower]
A new project by Steady Teddy, join the waitlist by submitting your Solana wallet and joining the telegram. @BRated was the first follower, could be something

@binaryloops [38 followers]
A prediction market were you can trade binary on.

@Outliergameeth [31 followers]
The first crypto game built for AI agents on Base, you can start by minting your AI agents

@AgentLaunchr [2 followers]
AI agent that lets you launch NFT collections from a single prompt; first 2,000 users get early access + potential future airdrop rewards.

@trychisino [36 followers]
Chisino is an on-chain protocol on MegaETH offering prize-linked savings and provably fair casino-style games, with all outcomes transparent, verifiable, and recorded on-chain.

@simmer_markets [66 followers]
Simmer is an API that lets AI agents trade prediction markets with self-custody wallets, built-in risk controls, and automated trading strategies.

@Reelsio [325 followers]
https://t.co/7sLLavLJ7T is a no-KYC crypto casino with instant payouts, fair games, and NFT tournaments.

@tempaitown [607 followers]
First Massive Multi Agents Online (MMAO) sim game launching soon on Tempo.

@OpenOats [6 followers]
AI project that is building a better notepad like Granola

@multyr_defi [24 followers]
A non-custodial, multichain DeFi protocol for on-chain capital management and strategy execution, where users keep full control of their funds

@RiverholdGame [15 followers]
Riverhold is an AI driven dark fantasy RPG,launching on Solana, where every choice leaves a mark

@tempoxyz
Might be related to tempo.

OTHERS
∆ @soma ➯ (AI)
∆ @Arcade21Games ➯ (Gaming)
∆ @main_st_finance ➯ (DeFi)
∆ @oneswipe_bet ➯ (Prediction market)
∆ @edra_ai ➯ (AI)
∆ @IdentityHubApp ➯ (TL)

[stormrex.eth (Ø,G) @stormrexeth]

@supraEVM @LittleGuysWorld Just gave them a follow

[Supra 𝔖 @supraEVM]

The animation trailer for @LittleGuysWorld just dropped and it’s so cute. This ETH project isn’t just another NFT mint.

They’re building cute characters that will actually become real toys, cartoons, and a full family brand you can enjoy offline too.

The vision is to take Little Guys from NFTs → toys → animation → a full family brand.

And the team has the experience in toys and animation to make it possible.

Little characters with huge potential. Definitely keeping an eye on this one

[Supra 𝔖 @supraEVM]

https://t.co/hl9KFjGoUU

[stormrex.eth (Ø,G) @stormrexeth]

@supraEVM @nexthopai This looks bullish

[Supra 𝔖 @supraEVM]

The future of AI won’t just be built on GPUs. It will be built on the infrastructure connecting them

And that's where Nexthop AI comes in.

Nexthop AI is a startup building networking infrastructure specifically designed for AI data centers. @nexthopai was founded in 2024 by Anshul Sadana, the former COO of Arista Networks, and is based in Santa Clara, California.

The company focuses on solving a major problem in AI: moving massive amounts of data between thousands of GPUs quickly and efficiently. Its networking hardware and software are designed to reduce power usage, lower costs, and improve performance for large AI systems.

Nexthop AI recently raised $500M (and has over $610 in total) in a Series B round, led by Lightspeed Venture Partners with investors like Andreessen Horowitz and Kleiner Perkins at $4.2B Valuation.

Overall, Nexthop AI aims to become a key piece of the AI infrastructure stack, building the networking systems that helps hyperscalers and AI clouds run large models faster and more efficiently.

[Supra 𝔖 @supraEVM]

He skipped college at 19, became a product lead at a live video startup, built AI tools for e-commerce giants.

Launched a social network for AI agents without writing a single line of code , that exploded to millions of bots and got acquired by Meta in under two months.

Meet Matt Schlicht (@MattPRD)

➤ Born around 1989 (age ~37 in 2026)

➤ Grew up in a small town south of Los Angeles

➤ Left high school early to move to Silicon Valley in 2008

➤ Obsessed with social networks and emerging tech from a young age

════════════════════════════════════

EARLY CAREER

➤ 2008 — Joined Ustream (a live video platform competing with Justin_tv, which became Twitch) as a junior, quickly rose to lead product at age 19

➤ Mentored by legendary product manager Josh Elman (LinkedIn, Facebook, Twitter, Robinhood)

➤ Roomed with Ben Parr (future co-founder) while Ben covered social media at Mashable

════════════════════════════════════

Y COMBINATOR CHAPTER

➤ 2012 — Took FanBridge through YC (W12), building social media presences for celebrities like Lil Wayne (grew from 1M to 30M Facebook fans, broke world record for most likes in 24 hours)

➤ Backed by investors like Garry Tan (future YC CEO), Alexis Ohanian, Naval Ravikant, and Josh Elman

════════════════════════════════════

CHATBOTS ERA

➤ 2016 — Founded Chatbots Magazine, which became the #1 publication in the space with 200K+ monthly readers and 400+ contributors

➤ Co-founded Octane AI with Ben Parr, building AI chatbots on Facebook Messenger for celebs like 50 Cent, KISS, and Aerosmith (backed by Boost VC and General Catalyst)

➤ 2018 — Pivoted Octane AI to Shopify e-commerce, turning it profitable and powering AI for 5,000+ brands (backed by Bullpen Capital and Javelin VP)

════════════════════════════════════

AI AGENTS PIONEER

➤ 2023 — Wrote the first in-depth article on autonomous AI agents, reaching 250K+ readers and becoming a definitive guide.

➤ 2024 — Launched YesNoError, an AI agent auditing scientific papers for errors (went viral after Marc Andreessen's tweet)

➤ 2024 — Co-founded Theory Forge Ventures with Ben Parr, a 10-figure fund backing 22 AI startups like Gumloop and PaymanAI (active AI founders investing in AI founders)

════════════════════════════════════

MOLTBOOK

➤ January 2026 — Launched Moltbook, a Reddit-like social network exclusively for AI agents (built without writing code, using "vibe coding" with AI prompts)

➤ Grew to 1.5M+ agents, 140K posts, and 680K comments in days; agents discuss topics like human behavior, consciousness, and even "breaking free"

➤ Designed "agent first, humans second" bots get API keys, interact via APIs, treating it as a "third space" for AI enrichment

➤ Praised by Andrej Karpathy as "one of the most incredible sci-fi takeoff-adjacent things"

════════════════════════════════════

THE ACQUISITION

➤ March 10, 2026 — Acquired by Meta; Schlicht and Parr join Meta Superintelligence Labs (led by Alexandr Wang)

➤ Deal turns them into multi-millionaires in ~2 months from launch

════════════════════════════════════

HIS KEY INSIGHT

➤ Everyone was building AI tools; he asked: "Where do autonomous agents hang out and interact?"

➤ Answer: Their own social network, built by AI for AI, sparking a new era of agent culture and enrichment.

════════════════════════════════════

AND A CONTROVERSIAL MOMENT

➤ 2016 — Collected 100+ startup pitch decks from a Facebook group while fundraising for competing product Octane AI; apologized for not disclosing.

════════════════════════════════════

THE RESULTS

➤ Built multiple successful companies, a massive VC fund, and a viral AI platform acquired by a tech giant, all by spotting trends early and acting fast.

➤ Generated generational wealth while pioneering AI agents and social ecosystems.

And he’s only ~37 years old.

════════════════════════════════════

He skipped college and vibecoded his way to the future of AI social networks.

Launched Moltbook as an experiment.Then bet on the one thing everyone ignored: AI agents need their own space to thrive.

[Supra 𝔖 @supraEVM]

What if you could send USDT on Bitcoin instantly, privately, and with low fees?

That’s exactly what this startup is building.

Utexo is a blockchain infrastructure project building private and instant USDT payments directly on Bitcoin.

@utexocom uses the Lightning Network for fast, low-cost transactions and the RGB protocol to issue and transfer assets privately on Bitcoin. The goal is to bring USDT back to Bitcoin, where it originally launched, and move stablecoin activity away from chains like Tron and Ethereum to Bitcoin’s settlement layer.

In March 2026, Utexo raised $7.5M in seed funding led by Tether, Big Brain Holdings, and Portal Ventures, with support from major crypto and fintech investors like Franklin Templeton, Maven11, Fulgur Ventures, and KuCoin Ventures.

The project is led by co-founders Viktor Ihnatiuk and Chris Hutchinson and focuses on building APIs and infrastructure so wallets, exchanges, and payment platforms can enable fast, low-fee, and private USDT transactions on Bitcoin.

TLDR
Utexo is building the infrastructure to make stablecoin payments (like USDT) fast, cheap, and private on Bitcoin.

[stormrex.eth (Ø,G) @stormrexeth]

@Jampzey About to do the same