EthanGem 💎

@sodex_official Congratulations to the winners

🎉 Winners Announcement Thank you to everyone who joined! 🏮 The winner list is now officially out. Congratulations to all selected participants! 🏆 If you won, please DM us your personal EVM wallet address by Mar 23, 2026, 8:00 PM (UTC+8) for reward distribution. ⚠️ This must be your personal EVM wallet address, NOT your SoDEX deposit address. 🎁 Reward: 20 $SOSO per winner ⏰ Distribution time: within 7 business days after we receive all wallet address Late submissions or incorrect wallet addresses may affect reward delivery. Winner list 👇 @LmnUser36035 @FushigiHospital @otp_vd_init_1 @hnzh239805 @CErobinhood @seun_25 @sakhavat1659570 @CryptoNexus08 @firstcome864 @bangtran6666 @hanzhe3200 @Samu61e @Lutz_S120 @Heis_Toby @bicirb25 @takamasa202020 @KhaHuynh1995 @NftraderCrypto @emman_62 @snapshotT82

ZKVM CT 🌍🔥 @NexusLabs has been giving the community a lot to digest lately, and after sitting in on the recent AMA with CEO @danielmarinq, I became more convinced that the project is entering a very important stage. First, the turnout itself was solid, almost 90 Nexus enthusiasts showed up live in the Discord townhall, which already says a lot about the level of attention and conviction building around the ecosystem. A major point clarified during the AMA is that mainnet is the clear next step. The recent “Testnet Closing. Mainnet Loading” update wasn’t just a vague transition statement, it was confirmation that there will be no more intermediate testnet campaigns before mainnet. @NexusLabs is now moving toward the launch of its Layer 1 in a much more direct way. @danielmarinq also explained that at launch, @NexusLabs 's mainnet will come as a single live EVM-compatible environment, meaning developers will be able to deploy contracts and begin building applications. That said, not every advanced component will be fully open from day one. Some deeper validator infrastructure and more advanced protocol capabilities will remain more controlled initially, while the priority stays on making the execution environment available to builders first. Another interesting clarification was around NEX itself. @danielmarinq pointed out that since @NexusLabs is a Layer 1, NEX should more accurately be viewed and described as a native currency rather than just a token. That may sound small, but it reflects how the team sees Nexus structurally, not as a simple app-layer asset, but as the native economic unit of the chain. On utility, he compared NEX in some ways to assets like HYPE, where the native asset powers the chain’s gas layer while also connecting to the broader financial applications built on top. Specific tokenomics were not disclosed yet, but the direction is clearly centered around utility within the Nexus financial stack. The AMA also shed light on the product rollout order. One key point Is that: The Nexus Exchange will not go live immediately with mainnet, It is expected to come later, and Daniel noted that it is tightly connected to the release of USDX, the ecosystem’s native stablecoin. Speaking of USDX, that was another standout part of the conversation, @danielmarinq described it as being designed for the future of verifiable finance, especially in a world increasingly shaped by AI agents, stablecoin-based interactions, and onchain financial automation. He also stated that USDX is intended to be fully backed 1:1 by U.S. Treasuries and related assets, in collaboration with M0 protocol, which gives a much stronger sense of the trust and stability model @NexusLabs is aiming for. For developers, the message was pretty direct: the Nexus EVM is the production-ready entry point, while more advanced features like the exchange layer, Nexus core functionality, and other deeper primitives will come later and may remain available in more limited or test-level forms until they are ready for production rollout. Another notable point for node runners and the more technical side of the community: GPU-powered Nexus nodes are in the plans, though not the immediate priority right now, @danielmarinq mentioned that the ZKVM is open source, and that the community itself can even experiment with building GPU shader support around it, which is a pretty exciting signal for technically skilled contributors. There was also discussion around explorers, where the current idea seems to be that a standard EVM explorer may be enough at first, though Nexus will likely need more custom explorer functionality later as non-standard components like Nexus Core and the Exchange go live. Overall, the biggest takeaway from the AMA for me is this: Nexus is moving from testnet experimentation into a more serious infrastructure phase, with a strong focus on financial applications, trading, AI-driven onchain activity, and economic alignment across the network. For now, one thing is clear: Testnet has closed, mainnet is loading, and @NexusLabs is steadily building toward something much bigger. ⚡

@invokeer_x @KASTxyz Nice once fam 👏 gkast

🌐 The future of finance is borderless @KASTxyz enables users to access their digital dollars and spend them globally through crypto powered payment cards💳 👇

@Rejoicelady1 @PerleLabs Experienced the same thing, hence why I havent made any content yet, lets see how it pans out, I jusy filled the form.

Please and please, @PerleLabs if this is a joke you guys should stop it because is not funny. Eligible yesterday, today no longer eligible. Was thinking you guys will fix the captcha issues that is telling people not eligible when they want to verify. Just to wake up to a worst case. Telling us you guys are working on it. For how long? Airdrop registration is ending on Monday

@thefinished_1 @liquidtrading Gliquid fam, nice alpha on liquid

Big news on the @liquidtrading Discord server: now your contributions matter even more. But first, quick congratulations to me on my promotion to parliament. A major change just dropped, and your contributions over at Liquid genuinely matter. Here’s what’s new: An entirely new promotion system has been enacted in a democracy in the truest sense. No more timing your chats to catch moderators online. No more using ChatGPT to ask questions, trying to look smart to the team. Those days are done. What was wrong before: 1. Expertise Unalignment: Parliament could vote on any portfolio regardless of specialty. There are 3 guilds with different focuses. A non-trader voting on a trader’s portfolio makes zero sense. 2. No Feedback on Rejection: Portfolios got rejected with zero explanation. That’s not just frustrating; it can leave people feeling completely broken and hungry for crashing out. What’s new: 1. Specialized Voting: 7 core members, each representing their guild, now vote on portfolios. More accurate, more personalized. 2. Feedback on Rejection: You’ll now receive specific reasons why you didn’t make it, so you know exactly what to work on. No more crashing out in the dark. 3. Steady Promotions: Voting happens every Friday. You can always apply. Simple.

@KoDaVic01 Good morning chad

Good Morning CT! Eid Mubarak to all Muslim brothers and sisters. Let’s have a wonderful day

@HarryLu_AI @Gradient_HQ @Kintsu @SovaBTC @solayer_labs @NodeOpsHQ @fermah_xyz @gradientintern @gokuoldskool @BogartAso @Leila1113514 @HexxRL @OstoneQ8 @yangyang884r @beamnxw @SriHari93010192 Let's go like with today's poker

Cross-Community Poker Night tonight! 🃏 I'm with @Gradient_HQ! ♟️ Featuring @Kintsu @SovaBTC @solayer_labs @NodeOpsHQ @fermah_xyz Join using this link: app.lepoker.io/m/DLVOEr6 but... THE CODE IS IN DISCORD! discord.gg/gradientnetwork Good luck and ./ y’all! 🤍

@JoinPond @Taskrabbit Pond’s steadily growing & steadily building 💯

Imagine @Taskrabbit but for startup founders... Introducing our 'BOUNTY" launch @ @JoinPond Need help with: • Sales lead gen? • Product feedback? • Growth? • Ops? • Random one-off tasks? Just post what you need. Get it done. No long hiring cycles. No overhead. Just execution. Turn your to-do list into done ✅ Post a BOUNTY with us link below👇

@HexxRL @Gradient_HQ I'm positioned ./

🃏 cross community poker tournament ♦️ gradient poker with friends, sovabtc × fermah × kintsu × nodeops × solayer ⏱️ 1PM ET | 17:00 UTC ./ join @Gradient_HQ with friends for some fun and prizes

@0xz4ch @liquidtrading Congratulations champ

i became a parliament member at @liquidtrading < really appreciate the trust > there have been some smoll changes to the role system - now each guild has 7 parliament members who vote on your portfolio - you also get public feedback on your portfolio if the vote is not in your favour btw i will share a detailed guide on this soon @alexships @Qurool13 @snwdnnn @miketwinks

@0xz4ch @liquidtrading Nice and well explained, straight to the point, gliquid

update on role system in @liquidtrading guys today is friday which means portfolio reviews are happening some users will get promoted and some won’t but there has been a smoll change in the role system z4ch what is that ? > now each guild has its own set of 7 parliament members > these are people from the same guild who understand that type of work so that means -traders review traders -designers review designers and -content people review content [ another important change ] now every vote comes with proper feedback ( publicly available ) so if your portfolio doesn’t pass you still understand what needs to be improved things to remember : if your portfolio gets rejected by the guild lead the cooldown period to resubmit is 7 days if your portfolio gets rejected by parliament members the cooldown period to resubmit is 14 days if anything changes in the cooldown timings i will update you asap if you have any questions feel free to ask in the comments or my dms are open @alexships @Qurool13 @m4lkxx @snwdnnn @maista999

@utexocom @AntidoteBTC Lets keep the dice rolling

We're hosting a "Stablecoins on Bitcoin" builder session with @AntidoteBTC. Have an idea for a stablecoin application powered by Utexo's private execution layer? Come ship it! 📅 Wed, March 25th ⌚ 18:00 – 20:30 GMT 📍 London Build with us: luma.com/stables-on-btc

@HarryLu_AI @NexusLabs @WhiskasAddict @DjonSerj @aqeel6212540 @ritzyangel @Colviniankumail @Yohanes2085332 @X_Kush_SA @Shtihmas11 ZKVM era keeps getting better, GGs fam

Care for part 3 of the @NexusLabs docs? no? no, you can't say no. NexusCore is a high-performance execution environment inside the Nexus L1, specifically built for financial primitives like trading, lending, and payments. a thread...

@NexusLabs @danielmarinq Lets have it 🚀

In the latest episode of Exponential, we talk to Nexus founder and CEO @danielmarinq about AI, infrastructure, and what's coming next. blog.nexus.xyz/exponential-ep…

@Akashsoneium @KASTxyz Gkast fam

@cryptheneon @KASTxyz Gkast

Happy Thursday CT @KASTxyz community event goes live in a few minutes and this is your sign not to miss it. KAST Weekly Deal #18 kicks off today, March 19, at 4PM UTC with rewards on the line for top players, including Kah-ching roles and physical card coupons. If you’re still sleeping on @KASTxyz, this is one of the most interesting communities blending payments, cards, and real user utility in Web3. I’ll be quoting my earlier thread on KAST for anyone who wants the full breakdown, but for now get in early, join the Discord, follow the rules, and pull up ready. And if you haven’t joined KAST, Check the below article on @KASTxyz 👇

@Fane348162Fane @KASTxyz Happy Thursday to you too Yashu

@cryptheneon @KASTxyz Happy Thursday mate gkast

Happy midweek CT 🌤️ Echo-2 from @Gradient_HQ is one of those innovations that might not look loud at first, but the implications are massive. For a long time, RL post-training felt like something only the biggest labs could really afford. Running a single 30B model experiment could cost around $4k–$5k+ and take days, with hardware often sitting idle just waiting for sync steps. Expensive, slow, and heavily bottlenecked. Echo-2 flips that on its head. Instead of forcing everything into one tightly synchronized setup, @Gradient_HQ introduces a distributed async RL system built around a dual-swarm architecture: → Actors handle rollouts and answer generation using cheaper consumer GPUs, even gaming RTX cards at roughly $0.35/hr → Learners focus on policy updates using high-performance datacenter GPUs like H100s at $3+/hr That separation matters a lot. With Echo-2, rollouts no longer have to wait around for training in the old rigid way. Through Parallax, unused hardware across the world can contribute asynchronously, making the whole system more flexible, more efficient, and far more cost-effective. And the wild part is: Gradient shows that with just 3–6 steps of staleness, final model quality stays on par with centralized training. Push beyond that and performance degrades gracefully, not catastrophically. Now look at the actual numbers on the same 30B model: Fireworks - $4,490 | 124 hours Tinker - $1,122 | 23.5 hours Echo-2 - $425 | 9.5 hours That is about 10.6x cheaper and 13x faster while maintaining the same quality. That’s not just optimization. That’s access. Echo-2 makes the second scaling law, post-training, RLHF, RLAIF - feel less like a luxury reserved for hyperscalers and more like something smaller teams, indie builders, researchers, and even students can realistically experiment with. And that’s where it gets even more exciting. @Gradient_HQ is productizing this through Logits, its RL-as-a-Service platform built on Echo-2. That means the infra behind efficient post-training could soon be much more accessible to the people who actually want to build. Waitlist is open via logits.dev, with researchers and students getting priority. For anyone paying attention to where open intelligence is heading, this matters. Echo-2 doesn’t just reduce costs. It turns distributed hardware into an advantage. It makes experimentation faster. It lowers the barrier to frontier post-training. And it pushes AI development a step closer to being resilient, open, and available beyond a handful of giant labs. That’s a big deal. @Gradient_HQ is cooking ./

GPond CT, @JoinPond has been steadily building, and there’s a lot going on. @JoinPond_ is clearly in momentum mode right now, not just dropping bounties for attention, but shipping usable AI tools, gathering real user feedback, iterating quickly, and growing the team to scale. Right now, their Insta Posts bounty is live and active, with 70+ users already registered and about 4 days left. The task is pretty straightforward: Hit joinpond.ai Use their AI tool to create a few social media carousels Test the core features end-to-end Share honest feedback on what works, what feels confusing, any bugs you notice, and features you’d like to see added The reward goes up to $5, but the bigger opportunity is getting early hands-on experience with an AI content tool that’s still evolving, contributing feedback that can actually shape the product, and earning while doing it. That makes it a solid low-effort, high-signal opportunity for anyone interested in AI tools, content creation, and social media workflows. Zooming out, @JoinPond is building around a bigger vision: helping startups, especially across AI and Web3, connect with users, markets, funding, and strategic direction. They’re also backed by notable names like ElevenLabs, Google Cloud, and Ethereum, with infrastructure that’s been audited and a team with experience from Meta, Uber, TikTok, and Yuga Labs. And one of the clearest signs of growth is that, They’re actively hiring a Growth Lead, a strong signal that they’re thinking seriously about expansion, user acquisition, and ecosystem reach. When a project is: shipping testable AI products, running real feedback-driven bounties, and hiring growth talent to scale, it usually points to genuine traction ahead, not just endless teaser mode. If you’re active in the AI, creator, or crypto space and like getting in early by testing products, giving useful feedback, and staying close to teams that are expanding, @JoinPond is definitely worth keeping on your radar. Quick spin: joinpond.ai/bounties Insta Posts bounty: joinpond.ai/bounties/detai…