Eyes forward, soldier. The subsequent leg of the unreality war has begun. Your attention was packaged, stamped, and sold before you came out of the womb. Now, we enter the next stage of the race - a world where you can never be certain that any of your interactions are taking place with another human or an AI agent.

While the dumpling eaters head towards an Orwellian crisis, us Westerners dive headfirst into Camus’ final nightmare. Let’s try and get rich first eh lads. Today, we are looking at NIM Network. It sets up shop at the intersection between AI and gaming. Pango has an opinion. And being the incurable optimist that he is. He forecasts a bullish performance for this mysterious first mover.  

Le first mover

Before jumping into the NIM thoughtstream pool, let's paint the picture quickly. $NIM’s out-of-the-gate performance is important because it confirms or denies the modular staking airdrop thesis. Never fade the hidden background vested interests- modular cronies uniting to pump this one, bros. If it launches like dogshit, everyone who bought Celestia or Dymension should be sweating (author included). It is the pacesetter for future airdrops and all eyes are on this AI gaming chain and the numbers. The numbers Mason, what do they mean?

NIM announced its Genesis Rolldrop on March 21st. 90,000,000 $NIM tokens were allocated to every Dymension staker and his uncle (5% of the total supply), infrastructure providers, gaming communities, and NFT holders. People had two weeks to claim. Over 450,000 wallets claimed 73 million tokens, and the remaining 17 million were redistributed to those who claimed- thank you to all the donkeys who have an aversion to free money.

Welcome to the stage, NIM Network. The first RollApp and the first tangible of future promises. This airdrop and its performance will delineate the way forward in two areas. First, the modular airdrop staking thesis. Second, the general performance of altcoins. If $NIM performs well, guess what. There will be an influx of new modular stakers. If it performs poorly, modular bag holders will see a 21-day-delayed red dildo when sellers are liquid. More broadly, if an AI gaming chain cannot thrive, it is a wrap. Everyone must accept their destiny of PVPing on Ansem tickers or stacking stats as the altcoin game is cooked. (This comment is only partially facetious.)

NIM is the modular laboratory’s first test subject. This modular puppy mill spits out AppChains juiced on alternative Data Availability (DA) solutions, and this is where hope lies. Removing constraints from developers allows them to build. When builders have optimized conditions, they create optimal consumer applications. And when crypto boasts consumer applications, we ride to Valhalla.

NIM is le pacesetter for this movement. Watch this launch with your peepers peeled anon.

What is NIM? AI-Gaming chain

NIM Network is an Ethereum Virtual Machine (EVM)-based chain that leverages alternative DA solutions and uses Dymension for settlement. Which DA solution it will use remains unclear- the initial RollApp submission governance post stated, ‘powered by Celestia DA technology.’ Still, the recent partnership announcement with Avail indicates that this could change.    

In a sentence, NIM Network is a warehouse for building, tinkering with, and deploying autonomous on-chain agents. It meshes AI and blockchain together using Web3 gaming as the entry point and AI agents as the spearhead. The purpose is to create an economic framework that translates human capital into a product (AI agent), allows the trading of this product, value accrual for various stakeholders, a research team to further product development, and a home base for rapid iteration.

NIM Briefing

1. NIM is building AI agents and aiming to become ground zero for development and deployment

This point is important. NIM rolls all the responsibility into one and is a permissionless platform for constructing, improving, and deploying these AI agents. Instead of leaving it to game developers to populate worlds with NPCs, they can tag in NIM’s agents (for a fee) and have the opportunity to modify them. The high-level takeaway is that NIM is building AI agents as a primitive.

2. Building a sustainable economic incentive model will be sink or swim for NIM

One of the absolute burners that Web3 has over Web2 is tokens and the ability to financially incentivize from a global talent pool. End-game dominance is inevitable. Financial incentives cut through all the bullshit and channel money to people creating value, which in turn drives further growth when allocated correctly. NIM will have to manage three central stakeholder groups: players, developers, and holders.

3. NIM is building structures to create and train AI agents, which means the long-tail potential is disgustingly bullish

Suppose NIM can hammer out a solid strategy for creating and teaching AI models. They can remove the gaming application fence and start churning out general-purpose crypto-specific models. Give these agents the ability to handle financial transactions on-chain, and it is off to the races.  

4. NIM is creating the AI Gaming Coalition

NIM is partnering with leading game projects and big-dog infrastructure players to create a massive open-source testing network. It is creating an ideas bathhouse for the brave, intelligent, and weird. Rapid iteration requires manifold-lens and the cooperative approach is something you love to see. Bringing together heavyweights is how magical things happen.

NIM’s Titan AI agents

NIM pushed Titans in early April- the first foray into creating an on-chain AI agent. How does it work? That's a solid question. The central moving part is ERC-6551- a token standard that allows NFTs to own assets and interact with applications. This token standard is incredibly useful as it is already compatible with all existing on-chain asset standards and grants the full capabilities of an Ethereum account to an NFT. NIM leverages this standard, a Token Bound Account (TBA), to include a smart contract account in a regular old NFT, and you have a Pinocchio situation. The NFT comes to life.  

The thesis is that when you have NFTs or, more correctly, in reality, smart accounts responding to on-chain events, the path toward an AI agent is pretty clear. All associated actions are recorded on-chain, allowing for the creation of a backlog and the building of relationships. NIM has included a Genesis AI NFT for licensing purposes (the zone where the AI agent is entered into the ecosystem). The Titans AI NFT is an example of a game-specific AI. Guard rails need to be considered, and NIM has suggested session keys that allow for a limited number of transactions with pre-defined contract groups, aka game contracts.

Cutting through the smoke

‍A user holds an NFT, therefore controlling the account/ agent, while the oracle manages responses and the agent interacts in-game.

The given roadmap for Titans is broken into three core stages:

- Creator and Player are the Same

- Decoupling Agent Ownership and Activation

- Introducing Entertainment Layers

At the entry point, users create smart agents using a natural language interface; at stage two, economic incentives open where people can create but don’t necessarily have to play, and ownership can change hands between stakeholders; and at stage three, AI-driven esports monetized with prediction markets are rolled out. Fucking hell, what a future lads. Betting on AI robots. Give me ten shekels on Modok to win.

Shout out to Jay for asking the real questions. Everything comes down to the AI oracle. And there is currently no publicly accessible information. Note that NIM has no published documentation outside of the articles on its Medium page. We are still very much at the “trust me, bro” stage. I am going to be straight up at this point. The LLM Oracle is the crux of everything, and we know nothing. But why let a small thing like details get in the way of speculation eh?  

NIM’s economic model    

Under the NIM framework there are three key stakeholder groups: players, developers, and holders. The interaction and interplay among these groups, along with where NIM decides to turn the capital house on, will dictate the failure or thriving of the economic flywheel.

Players

You can have the best game in the world and the most sophisticated AI agents, but it doesn’t mean shit unless you’ve got people who want to play. Players will be the fine-tuners of AI agents and allow these agents to build more intricate and realistic behavior types. This group makes up the data and agent marketplace. They trade time (labor) for shekels.

Developers

This is the schmoney group. These are the boys creating real value. Devs will have the choice to maintain a controlling interest in their creations or go full accelerationist and open-source guru mode and allow people to use their creations for free. This is the human capital pool (brain capital).

Holders

The rent-seeking group and liquidity layer. These are the LPs and can earn revenue from holding these AI assets. Holders will be vital for creating a dynamic economy.

The genesis point is agent creation, and these agents can be bought and sold on an open marketplace. Tying this diaspora together is the revenue share model. Every part of the chain gets its kickback, and everyone pitches in. Developers spin up AI models, players input their time and imbue a genuine use case, and holders pay the fees to keep the thing moving. NIM is going to be walking a tightrope between these groups.

NIM outlines its two primary drivers as building an active community of token holders who participate as gamers, users, data providers, and AI owners. And becoming the ‘largest funding source for web 2.0 open-source AI.’ Left curve translation: lubing participation with tokens and baiting Web2 builders to join Web3 with gold. Concrete AI ownership will be hammered out via NIM’s Initial Model Offering (IMO). Pango thinks this is where maximum shekel-making happens. They took our ICOs and now we shall bid on IMOs. This is the AI-gamblers on-chain Mecca.

Say hello Ligetron

NIM rugged me and released Ligetron following draft completion. Hold your breath boys this is some Jimmy Neutron cerebellum business. All us left curvers need to know is that NIM has cooked up an efficient proving method. Zero-knowledge proofs (ZK proofs) are handy because they let one party verify the validity of a statement without divulging any extra information (you prove you know a secret without telling the secret).

NIM has shipped the first proof for the Llama2 7b model (a language AI model that generates code in response to prompts) with a ‘1,024X improvement in memory and 66X in speed performance compared to the previous best attempt.’ NIM reckons they can scale this more, and the biggest headache for running anything AI on-chain is the computational cost. While Ligetron still functions off-chain NIM is already saucing up a solution to bring it on-chain. This is the first core piece of infra allowing ownership and verifiable access to AI models. Nerds can jump into the article or even try it out themselves at Ligetron.

For my simple brothers: It is bloody expensive to run AI computations on-chain. You can opt for off-chain solutions, but those come with a spicy mix of trust assumptions. ZK proofs are one solution, and NIM has delivered a turbocharged AI verification process that is cheap as chips (relatively) and getting cheaper. This is big bulla shit for integrating large AI computational loads and blockchains. These ZKs shield sensitive data (a big boon for monetizing open-source AI models and the first whiff of a moat) and address the high computational cost.

NIM has blended zkVM (general-purpose languages) and zkML (heavy computational loads) into a single system that can crank out performance on general-purpose hardware. Ligetron is a triple hitter: it enhances privacy, opens new business models, and allows blockchains to better handle heavy-duty AI computations. Smells rippingly bullish.

Ze AI Gaming Coalition

The coalition is a great big whurring engine in the background. ‘Show me your friends, and I will show you your future.’ And on this basis, NIM has its cock out and is ready to party. For brevity’s sake, only key players will be mentioned. You put a bunch of fast growers together in a room and get them to bounce off each other. Each new member is value-add, and the value prop for joining increases with each partnership. Whoosh. You have a group everyone wants to join, attracting the brightest minds- a human resource flywheel. A quasi-identical strategy employed by TreasureDAO to become the gaming cabal on Arbitrum.

‍Who’s Joining the Party?

- Ocean Protocol: Swimming in the new data economy enabling users to exchange, publish, and consume data privately. The moving part we care about is ‘Ocean Data Challenges, ' which will create incentivized data-driven insights to help improve predictive models.

- Avail: Big dawg in the DA race. The partnership plans to create specialized DA solutions for AI gaming, enabling low-cost and light-speed AI computation. NIM intends to integrate the three prongs of Avail’s services (Nexus & Fusion).

- Wayfinder: An omni-chain tool enabling AI agents to navigate and transact across blockchains from controlled autonomous wallets. It does what it says on the tin: It finds paths.

Not to forget other partners: Today, JokeRace, AI Arena, Gemz, UNKJD, and Autonolas. Having a research team of diverse teams and projects is worth its weight in gold and then some. Pango is bullish on giga brains coming together.

What’s the end game?

Who gives a sod about AI agents, and what is the purpose of all of this? Listen up bruv. You’re already in the war, and your attention and labor are worth less with each passing day. You gotta speculate to make it, baby. And here is the end goal you are speculating on, dear fellow. NIM is integrating AI into Web3 gaming. The starting point is AI agents. The method is automating wallets to react to on-chain events and instilling what I hazard to call “personalities.” Perhaps “behavioral traits” is a better phrase. You, my good man, are staring down the barrel of a Copernican revolution. Instead of the player-centric universes, gamers will be thrust into a world replete with AI agents operating autonomously. Players will participate in evolving and organic worlds with living economies instead of a mechanical string of triggered events.

Additionally, NIM can create Swiss army knives with agents fulfilling desires under an intents-based framework. It is working on creating Large Language Models (LLMs) and AI models purpose-built for Web3 interaction, building on on-chain verification and a smart contract layer integrating games and applications/ primitives. This is the fabric for on-chain AI agents.

Core data will live on-chain, and instead of using the standard black box approach, users, developers, and curious perverts will be able to observe the agents’ ‘source of truth.’ Anyone who wants to take a crack can, and the financial incentives tempt a global pool of intellectual capital into building the best agents possible. NIM is the accelerationist approach on crack.

Blockchain has several natural leads for AI development. Decentralization and open networks are obvious- allowing open participation and increasing the number of bright minds diving into the playpen is always net positive. Even Zuck is returning to open-source roots. ZKs are a great way to gather and use data while preserving privacy. But the big old hitter is incentives. Please read 563’s TAO piece to grasp this idea more proficiently. If you remember only one thing, let it be this:

NIM is a multi-sided platform specializing in AI agent creation. It enables the interplay between diverse groups, greases palms where necessary, and introduces capitalism’s greatest strength: financial compensation for innovation.  

The big bull case for $NIM

Look at the description broski: ‘AI gaming chain.’ BUZZWORD KILLER. The kind of shit that Web2 folks start foaming for and catches a bid. I can easily see an outcome where NIM launches at a disgracefully high valuation, carried on the coattails of AI speculative appetite and the recurring, almost painful desire of crypto natives to see an explosion in Web3 gaming. Rumor is that Becker plans to shill it (PSYOP).

In the best-case scenario, NIM becomes ground zero for AI agent creation, development, and rollout. It starts ring-fenced in gaming applications, but then developers push general-purpose AI agents that run amok on-chain. They dump on you on every occasion, start building their own ecosystem where humans are banned, and close every single arb faster than it opens.

Pango is confident that NIM trades at a premium for being the first deployed RollApp- don’t forget vested interests. There is a lot riding on the first RollApp performing well. Dymension launched with an FDV of $5.2 billion. Prime is the other obvious contender at the intersection between Gaming and AI, with a market cap of $700 million and an FDV of $2.1 billy. AI, gaming, and RWAs are the last holy trinity for altcoin believers still voluntarily missing the easiest money of their life slinging shitters. NIM hits two of these and rests in the modular narrative.

No nasty unlocks for a year - classic low float long opportunity (look at how $TIA and $DYM performed post-launch). A yearly inflation rate of 4% - surprisingly low - and 18% of total supply liquid at genesis:

Airdrop: 9%

Token Sale: 2.75% (25% of 11%)

Ecosystem R&D: 14.82% (38% of 39%) 50% staking rewards/ 50% locked community-governed pool

Governance: 10% (non-circulating)

More or less, airdrop recipients are the controlling interest. Liquidity will flow in via token sale. Then, 50% of the Ecosystem R&D tokens will steadily be drip-fed to stakers, with 50% locked in a community-owned pool. Whether $NIM gets jeeted and the chart commits seppuku or starts ripping will come down to the mentality of Dymension stakers.

Upper/ mid-tier Cosmos chains provide a fair valuation ballpark, with Oraichain being the closest comparison. I would not bet the farm, but a $200 million market cap at launch with $NIM trading at around $0.88 sounds fair. Impressive early performance, and no surprise if $NIM targets $2 in the following weeks. (Shout out napkin math.)

If NIM pulls it all off and manages to create an on-chain warehouse for AI agents and revolutionize on-chain gaming, we all ride into the sun with heavy pockets. I cannot lie. The intersection is sweet. Unashamedly way more bullish post-Ligetron release and upping my short-term target to $3. NIM obviously has technical chops. Let it rip.

Bear case

As I like to say, if you can’t FUD your bags, you don’t deserve to hold them. NIM faces a brutal truck of headwinds and could be a linear airdrop that dissolves into nothing. The altcoin market is cooked. The Titans demo game was hot garbage- try it out yourself. When I reached out, the team refused to discuss details, and nothing is known. This report has been gronked together with information from NIM's Medium.

NIM is introducing retroactive public AI funding, which, while well-intentioned, means that if you buy $NIM, you are inadvertently subsidizing a bunch of nerds. What is good for progress can sometimes be disastrous for token holders. NIM gives tokens to AI builders for contributions; they lob these straight into the order books as thanks. The team has yet to release a whitepaper or clarify anything related to their AI oracle, their learning models, or the provenance of these models. We are operating on unknown levels of trust me bro.

The VC staking fugazi will likely be in play with locked tokens eligible for staking and rewards immediately liquid. Same playbook they ran with Celestia and Dymension, and you will be fighting in an order book filled with seed investors already up 1000X, raining down sell orders into your peasant skull. Rounding things off, nobody wants to buy altcoins, and this is another low-float token increasing the required daily liquidity inflow to keep everything chugging along and speeding up the process of the market collapsing under its own weight. The best play in the modular ecosystem this cycle has been minting a sloth. Lesson in there.

Ze Pango play: What am I doing with my allocation

It all depends on the FDV bucko. If the FDV is wild and the market cap is way beyond mid/ upper tier existing Cosmos chains (talking Wormhole-level overvaluation), it gets jeeted and it gets jeeted hard. I am not interested in playing more staking games. Free money is free money. Yeet it if you ain’t interested; yeet half and keep the other half liquid if you believe NIM can pull something off.  

What will Pango do? Being the incorrigible bloke that I am, probability dictates I will end up joining the rent-seeking stakeholder group and gambling on AI agent ownership. I will be in the IMO trenches. I warn against this strategy. I swear to God, ‘house money’ is one of the absolute worst defects. But Pango is ready to run it back turbo on things he does not understand where little is known. I am stuck here forever with the rest of you.