JOSEPH DESOZE

@Walrus 🦭/acc $WAL #Walrus
Quand les personnes au sein d'une entreprise parlent honnêtement des sauvegardes et de la reprise après sinistre, cela ressemble rarement à une discussion technique claire. Cela a un aspect émotionnel, même si personne ne le dit explicitement. Il y a toujours une peur silencieuse sous les schémas et les politiques, la peur que, lorsque quelque chose de vraiment grave se produise, le plan de récupération paraisse bon sur papier mais s'effondre en réalité. J'ai vu cette peur apparaître après des incidents de ransomware, des pannes régionales des clouds, et des erreurs humaines simples qui se sont propagées bien au-delà de ce que l'on aurait pu imaginer. Walrus entre dans cette conversation non pas comme un remplacement spectaculaire de tout ce que les équipes utilisent déjà, mais comme une réponse à cette peur. Il a été conçu en supposant que les systèmes échoueront de manière désordonnée, que tout ne sera pas disponible en même temps, et que la récupération doit fonctionner même lorsque les conditions sont très éloignées de l'idéal.

@Walrus 🦭/acc $WAL #Walrus
Je vais expliquer Walrus comme je l'expliquerais à quelqu'un qui est curieux mais las de la hype, car le stockage n'est pas un sujet flashy jusqu'à ce que vous réalisiez que le stockage est là où le contrôle vit discrètement. La majeure partie de ce que nous faisons en ligne dépend du fait que les données soient stockées quelque part, et pendant des années, « quelque part » a généralement signifié une poignée de sociétés cloud et de leurs règles, de leurs pannes, de leurs changements de prix, et de leur capacité à supprimer ou restreindre l'accès quand cela leur convient. Walrus a été conçu autour d'un instinct différent, et cet instinct semble personnel parce qu'il repose sur la confiance : ils cherchent à garantir que de grands fichiers puissent exister sur un réseau qui ne dépend pas d'un seul intermédiaire. Walrus fonctionne dans l'écosystème Sui et se concentre sur le stockage de grandes quantités de données, ce qui signifie simplement de grands fichiers qui n'ont pas leur place dans un registre blockchain classique, et WAL est le jeton qui alimente l'économie et la sécurité de ce réseau de stockage. Si son adoption devient massive, le changement important ne sera pas seulement technique, il sera aussi culturel, car nous voyons de plus en plus de personnes et d'applications demander un stockage pouvant être vérifié, partagé et gouverné sans remettre tout le contrôle entre les mains d'une seule entreprise.

@Dusk $DUSK #Dusk
Dusk a été fondé en 2018 avec une compréhension calme mais sérieuse de la manière dont l'argent fonctionne dans le monde réel, car la plupart des premières blockchains ont été conçues comme des tableaux d'affichage publics où chaque mouvement est visible pour toujours, et cela semble noble jusqu'à ce que vous imaginiez une entreprise versant des salaires, un fonds rééquilibrant un portefeuille ou un lieu réglant des instruments réglementés tandis que l'intégralité d'internet observe la stratégie se dérouler en temps réel. Je ne dis pas que la transparence est fausse, car la transparence peut protéger les gens, mais la finance comporte également responsabilité, confidentialité et obligations légales, et lorsque ces éléments sont ignorés, l'adoption ne échoue pas bruyamment, elle échoue silencieusement, car les institutions ne se présentent tout simplement pas. Dusk a été conçu pour combler ce vide silencieux entre ce que promet la blockchain et ce que les marchés réglementés exigent, en essayant de créer une couche 1 capable de soutenir des activités financières sérieuses où la confidentialité est normale, la conformité est attendue et la traçabilité est possible sans transformer la vie financière de chacun en divertissement public.

@Walrus 🦭/acc $WAL #Walrus
Vous vous êtes déjà arrêté pour réfléchir à l’endroit où se trouvent réellement tous ces jeux de données d’entraînement, ces images NFT et ces modèles d’IA ? Je ne parle pas du joli dossier sur votre ordinateur, mais de l’endroit réel où vos données reposent lorsqu’elles sont « dans le cloud ». La plupart d’entre nous imaginent des rangées de machines dans de gigantesques entrepôts détenus par quelques entreprises puissantes, et même si ces entreprises n’ont pas l’intention de nuire à personne, la vérité est que nous leur laissons le dernier mot sur l’accès, les prix et la visibilité. Ce sentiment peut lentement s’installer au fond de votre esprit jusqu’au jour où quelque chose est bloqué, supprimé, censuré ou simplement rendu inaccessible en raison de son prix, et alors il devient impossible de l’ignorer. Walrus a été conçu pour apaiser cette peur de manière concrète, non pas avec des slogans, mais avec un système qui rend le stockage de grands fichiers plus facile dans un réseau décentralisé tout en permettant toujours aux utilisateurs de prouver que les données sont bien présentes. Créé par l’équipe Mysten Labs et conçu pour fonctionner avec la blockchain Sui comme couche de coordination, Walrus se concentre sur les blobs, ce qui signifie essentiellement de grands fichiers non structurés comme des vidéos, des jeux de données, des poids de modèles, des ressources de jeux, des archives et tout autre élément qui ne s’inscrit pas facilement dans un stockage on-chain de petite taille. Ce qui le rend différent, c’est qu’il ne prétend pas que tout doit être sur la chaîne ; au contraire, il utilise la chaîne là où elle excelle : pour la coordination, les règles, les paiements et la preuve, tout en gardant les données lourdes hors chaîne, où elles peuvent être gérées de manière efficace.

@Dusk $DUSK #Dusk
I still remember the first time I truly “got” what Dusk Foundation is trying to do, because it didn’t feel like another blockchain chasing speed for bragging rights, it felt like a real attempt to make on chain finance behave like the real world, where privacy is normal and rules are not optional. Founded in 2018, Dusk was shaped around a simple tension that keeps coming up in finance: institutions need confidentiality to protect clients and strategies, but regulators need visibility to enforce laws, and most blockchains pick one side and ignore the other. On fully transparent networks, every balance and transfer can be watched by anyone, which is basically asking a serious financial firm to broadcast its playbook to the entire market, and on fully private systems, it becomes hard to prove anything to an auditor without breaking the privacy promise. Dusk tries to live in the middle, and I think that’s why people describe it as regulated and privacy focused infrastructure, because it’s not just about hiding data, it’s about hiding data in a way that can still be explained and verified when the right party asks for it.

To understand why they built it this way, it helps to look at what “tokenization” usually means and why it often disappoints once you move past the marketing. Tokenization can create a digital token that represents an asset, but very often the real asset still sits in a traditional custody stack, which means someone still has to reconcile records, settle trades through old pipes, and handle compliance off chain, and the blockchain becomes a fancy front end rather than a true market infrastructure. Dusk leans harder into the idea of native issuance, where the asset is created and managed directly on chain, because if the asset itself is native, then settlement and ownership updates can happen in one place, with less reconciliation, fewer middlemen, and fewer points where things can break. If this approach works at scale, it becomes more than a crypto product, it becomes a new kind of financial plumbing, and that is exactly the kind of ambition that feels risky but meaningful.

Now let me walk you through how the system works in a step by step way, because the design choices matter here. Dusk uses a modular architecture, which means they separate the layer that finalizes and secures transactions from the layer that executes different kinds of smart contracts, and that separation is not just technical style, it’s a practical decision for regulated markets. At the base you have the settlement and consensus layer, the part that makes sure blocks get produced, agreed upon, and finalized with strong guarantees, and above it you have execution environments that can be optimized for different needs, including a privacy friendly environment and an EVM compatible environment so developers can use familiar tooling. I’m describing it like this because it’s easier to see the point: they want one reliable settlement foundation, then multiple ways to build on top of it without compromising the settlement rules, and that’s how you reduce integration friction while still keeping the core compliance and privacy story consistent.

Consensus is where regulated finance usually gets nervous, because probabilistic finality and chain reorganizations create uncertainty, and uncertainty is expensive when real assets and legal obligations are involved. Dusk’s consensus design is proof of stake and committee based, and the flow is basically three phases that repeat: one party proposes a block, a committee validates it, and another committee ratifies it, and when ratification is reached the block is treated as final in normal operation. They call the stakers “provisioners,” and the system uses a deterministic selection method so committees are chosen in a way that aims to be fair and hard to game. If you’re thinking like a market operator, the reason this matters is simple: once a trade is settled you need confidence it won’t be reversed, and they’re trying to make that confidence feel closer to what traditional settlement systems promise, only with an on chain backbone. I also like that they pair this with a network layer designed to move messages efficiently and reduce predictable bottlenecks, because in practice, even the best consensus design suffers if the network can’t propagate blocks and votes quickly and reliably.

Privacy is where Dusk gets interesting, because instead of forcing everything into one model, they support two transaction styles that users and applications can choose from depending on what they need. One model is transparent and account based, similar to what people already understand from common smart contract platforms, and it’s useful when you want open market signals and simple verification. The other model is built for confidentiality and leans on zero knowledge proofs, where the chain can verify that the rules were followed without revealing sensitive details like amounts and counterparties to the public. This is where I start to feel the emotional logic of the project, because they’re basically saying, “We’re not here to hide crime, we’re here to make privacy normal again while still letting lawful oversight exist,” and that’s a stance that fits regulated finance better than extremes. If a transaction needs to be auditable to the right authority, the system is designed so disclosure can happen in a controlled way, and if a transaction needs to be private from the general public, it can stay private while still being valid and enforceable. They’re trying to make privacy a built in feature, not a bolt on trick, and they’re trying to make compliance a native function, not a manual process that someone hopes gets done correctly.

On top of these foundations, the project aims at a specific end goal: institutional grade financial applications, compliant DeFi, and tokenized real world assets that can actually live and move on chain without turning into a compliance nightmare. That’s why you’ll see them talk about things like confidential securities, lifecycle management for regulated instruments, and identity and permissioning primitives that help separate public flows from restricted flows, because regulated markets are full of rules about who can hold what, who can transfer to whom, and what reporting must happen. They also focus on building execution environments that developers can realistically use, because privacy and compliance mean nothing if the tooling is so unfamiliar that nobody builds. We’re seeing a broader industry trend where people want EVM compatibility for speed of development, so Dusk leaning into an EVM equivalent execution layer while keeping a settlement layer with privacy and compliance features is a strategic attempt to meet developers where they already are, without giving up what makes the chain different.

If you want to follow the project like a grown up, not like a gambler, the metrics to watch are not just price and hype, they’re the signals that the network is becoming real infrastructure. Finality time is one of the biggest, because financial markets care about predictable settlement, not theoretical throughput. Staking participation is another, because a proof of stake system needs broad, healthy participation to avoid centralization and to keep governance and security credible. Adoption metrics matter even more than raw transaction counts, like how many serious applications launch, how much real asset issuance happens, how much value settles on chain, and how widely developers actually use the execution layers. Token economics also matter in a quiet, long term way, because emissions, incentives, and staking rewards shape who secures the chain and whether security stays strong when speculation cools down. It becomes obvious over time whether incentives are building a durable validator set or just attracting short term yield seekers.

No honest article is complete without the risks, and Dusk has real ones, even if I personally respect the direction they’re taking. The first risk is complexity risk, because combining modular architecture, zero knowledge privacy, dual transaction models, and regulated asset logic creates a larger surface area for bugs, misunderstandings, and integration mistakes, and in finance, mistakes are expensive. The second risk is regulatory risk, because rules evolve, interpretations shift, and what looks acceptable today might require changes tomorrow, especially when privacy is involved. The third risk is adoption risk, because institutions move slowly, and even when the tech is solid, legal clarity, trusted partners, and operational comfort take time, and sometimes the market moves on before the best design gets its moment. There is also liquidity and market structure risk, and this is one place where mentioning Binance can be relevant, because if a large portion of trading volume concentrates in one venue, it can create fragility if that venue’s access or operations change. Finally, there is competition risk, because other ecosystems are also pushing privacy tech, compliance tooling, and real world asset narratives, so Dusk has to prove its edge through delivery, partnerships, and real usage, not just vision.

When I look at the future, I see two possible paths, and both feel believable. In the stronger path, Dusk becomes a quiet backbone for regulated markets that want on chain efficiency without public exposure, and it grows through real integrations, real issuance, and real settlement, and the token becomes more than a speculative chip because it supports staking, fees, and the security budget that keeps the network honest. In the harder path, the tech remains impressive but adoption stays slow, regulations get messy, or competing standards win mindshare, and the project struggles to reach escape velocity even if the ideas are good. Still, I think the most important thing is that they’re aiming at the right problem, because privacy with auditability is not a luxury in finance, it’s the default expectation, and if Dusk can make that expectation feel natural on chain, we’re seeing a new kind of bridge between traditional systems and decentralized rails. I’m not saying the road is easy, but I do think the direction is worth watching, and if the team keeps building with patience and clarity, there’s a real chance that what starts as a niche solution becomes a standard way we move value in a world that wants both freedom and responsibility.