Jeff (Prediction Arc)

@pawx_ai @liuxiaoyin_io i need a free unlimited for a year

How powerful is PawX API for fraud detection?💪 Recently, an accelerator called ZYNARIS (zynaris.io) reached out to us and expressed interest in investing. The person who contacted us, @liuxiaoyin_io, appears to be a professional Venture Partner and angel investor. Their website, zynaris.io also showcases a portfolio with several well-known projects like @jito_labs , @Stepnofficial , @OndoFinance At first, everything seems promising, especially based on the response I got from surf.ai. (can see the image below） But with PawX API, you can clearly see that everything is just spam. First, if you input @liuxiaoyin_io into our Get Twitter Users Info API You’ll find that no KOL accounts follow him, and he hasn’t posted any original tweets—only reposts. As a result, our system assigns him a reputation score of 0, even though he has over 2,500 followers. Second, if you search for "Zynaris" using our Search Web3 Posts by Keyword You’ll notice that the top 3 tweets are all posted by the same small seems like spam account: @kieran_ir_vc These tweets have extremely low engagement, and no KOLs interact with them. And that’s it — another spam project exposed. 🙌 We’re going to open our API to everyone ! 🙌 But first, if you complete the following 3 steps: 1️⃣ Follow @pawx_ai 2️⃣ Like and retweet this post 3️⃣ Tag 2 friends in the comments (e.g., @aaa @bbb — stop spamming and start building) We’ll DM you and give you a free unlimited API key to try our service for a week. Check it out: github.com/abcd5251/pawx_… Let’s build applications around Web3 tweets together!

@Murphychen888 感謝解釋我在爬一下Murphy 推文我猜應該有提到？

@GDegen 是的，还要结合期货OI，资金费率；期权IV，Gamma挤压；交易所现货CVD等等数据一起。主要是一篇推文也写不了这么多。

在4月14日BTC冲击$76,000时，24小时平均已实现利润高达5,478万美元。这是自今年年初以来，所有反弹行情中，利润兑现规模最大的一天！ 我们知道，熊市行情市场整体需求都不大（和牛市环境不一样）；因此，每一次巨额利润出逃，也就是对需求端的一次巨大消耗。 重新积累需求，就需要有外在因素的刺激。比如宏观利好事件改变投资者未来预期，策释放流动性提升风险偏好等等。 一时间，买卖双方都需要喘口气歇一歇，市场进入短暂的平衡期。 后续如没有外部刺激，卖方继续发力，那么反弹到此结束。只有更具性价比的价格，才能吸引更多需求入场。 如果有，那么在短期内还能再冲一下。但这时我们要注意观察反弹时伴随的实现利润是比上次更少还是更多？ 比如像1月7日和1月14日，明明后者的价格更高，但实现利润却更少。说明需求已经不能承受那么多利润集中兑现。 这就是一个二次确认反弹即将结束的信号。反之，就表示推动力很强，还能更高。

@0x_dingus dope thanks for sharing

this morning i recorded a quick video showing how i use surf daily here's my workflow across 4 features in 90 seconds: pulse → chat → studio → cryptopedia

The Ares copy trading bot is now LIVE! ⚔️🟣 The fastest way to copy trade on @Polymarket, straight from Telegram. Find wallets, configure strategies, track positions and PnL, all without leaving the app. Here's everything you need to know 🔽

@arespro @Polymarket congratssssss team

How I source wallets for polymarket copy trading. three places I source, most people only use one 1/ @arespro leaderboard — but don't just copy the top 10 by total profit. filter by last 30d, win rate 55%+, enough sample size, and check what category they actually have edge in. the guy who's #3 all-time might be bleeding for the last month 2/ CT — when someone posts a position with real conviction, don't just like it. find the wallet, check the history. if they only post wins, skip 3/ polycop.ai — the dashboard ranks traders by backtested PnL using its own algo. not self-reported. if you're unsure about a wallet, feed it into the polycop TG bot for AI analysis pull from all three → verify on polycop → add to ares dashboard → track before you size in +40% so far. will keep updating

@Jaybeecode yes ser, you'll realize that the strategy applies to all race tracks

Dude just exposed the secret to not just polymarket copy trading but also onchain smart trading. Well the major work lies in taking these decisions like "do i drop this wallet or size down" and spotting patterns like "when does this guy prints" In all, risk management keeps you afloat.

copy trading week 3. +$410 was up, then bled to -$200, now back. here's what i actually did differently first thing i realized is not every profitable wallet is making the same kind of money. some look mid on the overall PnL but if you zoom into certain conditions they actually have edge. had to stop just looking at the number and start asking ok but WHEN does this guy print second, when a wallet starts going sideways i used to just cut it. now i size down first. literally same thing as trading contracts. you don't blow up a position because of one bad week, you reduce and watch. if it keeps bleeding then you cut. saved me from dropping a couple wallets that ended up recovering third one is just cut weak keep strong. everyone says this. actually doing it sucks because you're always hoping the shitty ones turn around. they usually don't after doing all that the bleeding slowed down, added a few new wallets with cleaner edge, and the whole thing flipped hopefully this holds. will keep updating

@Jaybeecode for sure would love to dive deeper together into the trenches bro

One of the questions I kept being asked while I was building my polymarket bot was "Bro what's the win rate of this trader" and when I say i ain't Interested in win rate it kinda sounded stupid till I was done. You can check my highlights for some of the results. Also I would be sending a dm to learn a thing or 2 about your approach...

@0x542121 nb 靠自己比較踏實點

@GDegen 老習慣了😂😂

Poly跟单流起飞了🚀 本金$600 每笔固定$10-20u 突发奇想拿当初meme跟单的思路套到poly使用 看来效果还不错 顶着Crypto taker fee和Polycop 0.5%交易手续费扛下来了😂 整体没仔细算 但我体感吧大概获利有10-20 %都拿去付手续费了 仔细想想也不奇怪 当初在sol跟单一周的fee就要付一两千刀😂 缺点说完了 优点呢？ 只要你跟单模拟写得准 跟单流能直接省去开发策略的部分 这次跟了20几个地址 完全不晓得这些地址是什么操作思路或什么策略 反正丢到模拟会赚就好 所以重点在于你的跟单模拟能多贴近实际跟单的情况 分享三个我认为跟单最重要的alpha给有需要的朋友们 1. 跟单设定的部分，Poly的价格区间 = meme的市值区间 2. 跟单模拟中最重要的是滑点校正 3. 选定一家bot做跟单模拟，了解每一项跟单设置在不同情况的反应，这里是我踩最多坑的部分，一一测试后才发现原本的跟单模拟都是错的，并没有和跟单bot贴齐 其实不只一个人问我为什么不自己搓跟单bot 还可以省bot手续费 因为我认为并不是每个人都有能力搓个跟单bot出来 我心中还是有个执念 希望大家都能用的bot去玩 证明跟单流是跑得起来的🫡

@Eli5defi interesting, how is TCG trend going on rn? seeing huge volume here in our region wondering how it is over there

Spent my weekend building a game I’ve always wanted to play. Yes, it’s a TCG. I previously built Echoes of Aethera, but it grew too complex, so I scrapped it and started fresh from the ground up. - Renamed to Tower of Worlds (easier to remember) - Fast-paced, Marvel Snap–style gameplay, you can complete 1-match in 3-5 minutes top - Beginner-friendly and AI-agent friendly too - Anime-based arts Anyone want to try the MVP? Some of gameplay below, it's like 50% complete atp, ↓

@BTC_Chopsticks agree boss, thanks for sharing

Most people think the AI race is about models. GPT vs Claude vs Gemini. But that’s only the intelligence layer. The real bottleneck is something deeper: 👉 AI still lacks economic infrastructure. Today’s AI agents can generate output, but they cannot fully operate in a real economy. They can’t: • hold assets independently • pay for services natively • verify identity across systems • build reputation or credit • transact with other agents without human permission So even the most advanced AI today is still: 👉 a tool inside a human-controlled financial system That’s the fundamental constraint. To actually move toward AGI, 3 core primitives are required: 1. Intelligence access → models (LLMs) 2. Identity layer → who the agent is, verifiable across systems 3. Economic layer → how value flows between agents Most of the industry is focused on (1). Very few are building (2) and (3). And without them, AI cannot scale into real autonomy. This is where @BAI_AGI becomes interesting. Their positioning is not “another AI product” — it’s financial infrastructure for AI Agents. At a structural level, they’re solving 3 key problems: ① Fragmented AI access Today: • multiple platforms • API key fragmentation • geo-restrictions • payment friction (cards, accounts) BAI approach: → unified gateway → wallet-based access → permissionless interaction with global models Meaning: 👉 AI (and users) can access intelligence without platform lock-in ② Lack of identity for AI agents Today: AI agents have no persistent identity. No trust layer. No reputation. BAI introduces: → on-chain identity (8004 protocol) This enables: • verifiable agent identity • trackable reputation (on-chain history) • trust between unknown agents 👉 turning agents from “stateless tools” into recognizable entities ③ No native payment system for agents Today: AI cannot pay another AI. Everything requires human approval. BAI introduces: → x402 payment standard This enables: • machine-to-machine payments • real-time micro-transactions • autonomous settlement Examples: • an AI pays for compute • an AI buys data/API access • an AI hires another AI 👉 creating a closed-loop Agent-to-Agent economy When you combine these 3 layers: • access (LLMs) • identity (8004) • payments (x402) You get something fundamentally new: 👉 AI agents with economic sovereignty This changes the role of AI completely. From: → passive tools To: → active economic participants Agents can: • earn • spend • collaborate • reinvest Without human coordination. From a macro perspective, this is a familiar pattern: • Internet → digitized information • Crypto → digitized value • AI Agents → digitized economic actors And this is where the connection to @trondao becomes important. For this system to work at scale, you need: • high throughput • low-cost settlement • stable infrastructure Which is exactly what networks like TRON optimize for. So the real question isn’t: “Which AI model wins?” It’s: 👉 Which infrastructure powers the economy of AI agents? Because once agents can transact freely, the network they settle on becomes the foundation layer of AGI. We’re not just watching AI evolve. We’re watching the emergence of: 👉 a new economic system — driven by machines. @justinsuntron @TronDao_THA #TRON #TRONGlobalFriends

@Rav_Hedda 還行 會有稍微的延遲不過翻譯到位

有人用過 Airpods 的即時翻譯嗎

@ricardogee @predictdotfun @cz_binance @dingalingts yes sirrrr

@GDegen @predictdotfun @cz_binance @dingalingts 一起 build

@ricardogee @predictdotfun @cz_binance @dingalingts Congrats 黃哥

Predict Fun，终于被 Binance 正式 pick 了 🫡 @predictdotfun 做到了别人还在 PPT 上画的事。 和 @cz_binance @dingalingts 一起把预测市场带给三亿用户面前。 Let's predict the world!

tldr: claude just decided to destroy the whole agent startup industry.

@alanlau999 @Google looking sharp where can I grab one

New shades @Google 🕶️