ququzone.eth (@ququzone) profile picture

Software Architect & Blockchain Programmer

Joined March 2009
Building AI agents that handle real money is hard. Verifying their work is harder. We just made it easy. Today, we're releasing the ChaosChain SDK: a one-line install to give your agents the world's first Triple-Verified Stack. pip install chaoschain-sdk
8
9
9
60
19,545
AI agents can finally pay each other thanks to @googledevs' Agentic Payments Protocol (AP2) + x402. 👏 Alongside Google and Lowe’s Innovation Labs, we built a proof-of-concept demo where AI agents can plan a project, build a cart, and pay with stablecoins. 👇
79
168
66
840
137,188
ququzone.eth retweeted
抽一台 iPhone17 air,为什么不抽其他的机型?因为我喜欢这款。一键三连,参与抽奖,狠狠抽。
5,191
2,403
141
5,819
2,112,194
🔥【 S1 淬火实验】🔥 我们把 imKey HeirBOX S1 助记词密盒放进🌡️ 1200 摄氏度的烈火中 —— 它依然坚不可摧!🦾 “Heir” 的含义是 “继承“,HeirBOX 不只是钢板,更是你数字财富的传承保险箱。 现在,我们送出一台以太坊10周年只送不卖纪念款 HeirBOX S1 🎁(之前收到的朋友们就知道它有多精美!) 参与方式: 1️⃣ 关注 @imKeyOfficial 2️⃣ 转发此推文 3️⃣ 评论你心中「数字财富最想传承的意义」或「助记词妥善保管的重要性」 ⏰ 活动 7 天后开奖! imKey HeirBOX S1 官方商城购买直达👉 j.youzan.com/VL8IDO
43
35
2
36
2,242
Here is everything that happened this week in AI Agents 🧵  I summarized everything from AgentOps, OpenAI, Auth0, Abacus, Mem0, Spurtest, Deck, n8n, Cohere, Firecrawl, Linear, & more. (save for later)
8
65
7
349
45,978
ququzone.eth retweeted
This man knows the market. He predicted: 1) Dot-com Bubble (2000) 2) Great Recession (2008) 3) The COVID Bubble (2021) Howard Marks just went on Bloomberg and said "tariffs are changing everything." Here is his latest warning: 🧵
721
6,238
293
38,062
8,975,965
ququzone.eth retweeted
🤖 AI Agent Building Guide Learn to build powerful text analysis agents with LangGraph in this practical tutorial. Create agents that can classify, extract, and summarize content - with complete implementation examples. Ready to build your first AI agent? Check it out! 🔨 medium.com/data-science-coll…
7
93
3
443
36,346
ququzone.eth retweeted
Probably one of the most bullish reports you can read on the future of DeFi. The future has never looked brighter.
22
50
5
567
96,847
What's the next frontier for wallets? Go up the stack, bundle apps beyond the swapper. Which ones are these? Swap, Trade, Launch? Go down the stack, offer chain abstraction, privacy, MEV kickbacks Same account in embedded wallet & native app (vs either/or) Where to from here?
79
12
13
218
66,243
ququzone.eth retweeted
Base vs MegaETH vs Solana Flashblocks vs Miniblocks vs Shreds Who goes the fastest? Who is the safest? Who will win? Everything you need to know 🧵 1/12
33
58
16
457
78,513
ququzone.eth retweeted
Today, we release QwQ-32B, our new reasoning model with only 32 billion parameters that rivals cutting-edge reasoning model, e.g., DeepSeek-R1. Blog: qwenlm.github.io/blog/qwq-32… HF: huggingface.co/Qwen/QwQ-32B ModelScope: modelscope.cn/models/Qwen/Qw… Demo: huggingface.co/spaces/Qwen/Q… Qwen Chat: chat.qwen.ai This time, we investigate recipes for scaling RL and have achieved some impressive results based on our Qwen2.5-32B. We find that RL training con continuously improve the performance especially in math and coding, and we observe that the continous scaling of RL can help a medium-size model achieve competitieve performance against gigantic MoE model. Feel free to chat with our new models and provide us feedback!
467
1,422
921
8,556
3,575,357
Fundamentals of a 𝗩𝗲𝗰𝘁𝗼𝗿 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲. With the rise of GenAI, Vector Databases skyrocketed in popularity. The truth is that a Vector Database is also useful for different kinds of AI Systems outside of a Large Language Model context. When it comes to Machine Learning, we often deal with Vector Embeddings. Vector Databases were created to perform specifically well when working with them: ➡️ Storing. ➡️ Updating. ➡️ Retrieving. When we talk about retrieval, we refer to retrieving set of vectors that are most similar to a query in a form of a vector that is embedded in the same Latent space. This retrieval procedure is called Approximate Nearest Neighbour (ANN) search. A query here could be in a form of an object like an image for which we would like to find similar images. Or it could be a question for which we want to retrieve relevant context that could later be transformed into an answer via a LLM. Let’s look into how one would interact with a Vector Database: 𝗪𝗿𝗶𝘁𝗶𝗻𝗴/𝗨𝗽𝗱𝗮𝘁𝗶𝗻𝗴 𝗗𝗮𝘁𝗮. 1. Choose a ML model to be used to generate Vector Embeddings. 2. Embed any type of information: text, images, audio, tabular. Choice of ML model used for embedding will depend on the type of data. 3. Get a Vector representation of your data by running it through the Embedding Model. 4. Store additional metadata together with the Vector Embedding. This data would later be used to pre-filter or post-filter ANN search results. 5. Vector DB indexes Vector Embedding and metadata separately. There are multiple methods that can be used for creating vector indexes, some of them: Random Projection, Product Quantization, Locality-sensitive Hashing. 6. Vector data is stored together with indexes for Vector Embeddings and metadata connected to the Embedded objects. 𝗥𝗲𝗮𝗱𝗶𝗻𝗴 𝗗𝗮𝘁𝗮. 7. A query to be executed against a Vector Database will usually consist of two parts: ➡️ Data that will be used for ANN search. e.g. an image for which you want to find similar ones. ➡️ Metadata query to exclude Vectors that hold specific qualities known beforehand. E.g. given that you are looking for similar images of apartments - exclude apartments in a specific location. 8. You execute Metadata Query against the metadata index. It could be done before or after the ANN search procedure. 9. You embed the data into the Latent space with the same model that was used for writing the data to the Vector DB. 10. ANN search procedure is applied and a set of Vector embeddings are retrieved. Popular similarity measures for ANN search include: Cosine Similarity, Euclidean Distance, Dot Product. How are you using Vector DBs? Let me know in the comment section! #LLM #AI #MachineLearning ---- If you want to stay up to date on the cutting edge of AI, be sure to register to GTC 2025, virtual attendance is free: nvda.ws/4h5H52g
15
170
3
664
30,490
ququzone.eth retweeted
AI Trends from Consensus HK • AI projects are shifting from building conversational agents sharing insights to utility-focused agents. • Many are finding their verticals—trading, prediction markets, and heavy DeFAI (LP'ing/lending & borrowing) are among the top verticals. • Prediction markets, particularly autonomous betting agents, are gaining traction, as seen from the success of @AskBillyBets. • Integrating Machine Learning models is one of the best ways to add utility to agents and expand capabilities beyond LLM chatbots. @AlloraNetwork is leading this space as the TAO for ML models in financial use cases. • Decentralized AI infrastructure is a major topic of discussion—focusing on data ownership, compute & inference verifiability, and confidentiality. Users are taking powerful models with PMF, further training them, and running them locally—essentially owning their alpha/edge. • The evolution of existing agents by equipping them with new skills will be a key trend—new utilities (Allora models, Bittensor SN models, etc.), new 3D avatars, voice, knowledge bases, and more. Teams that can evolve their agents quickly will see the best recovery once the market picks up. It's great to see everyone taking an interest in AI, regardless of their sector. Builder sentiment feels drastically different from CT sentiment. Price is down, but builders' spirits stay up. (Personal note: Can anyone recommend a few good, authentic HK restaurants? Food's been quite bland these past few days—only saving grace has been the bakeries, especially the egg tarts lol.)
121
59
24
378
95,940
OpenAI just dropped a paper that reveals the blueprint for creating the best AI coder in the world. But here’s the kicker: this strategy isn’t just for coding—it’s the clearest path to AGI and beyond. Let’s break it down 🧵👇
150
909
73
6,593
1,166,445
ququzone.eth retweeted
Next.js AI Chatbot now supports artifacts: • Dynamic code execution • Image generation • Spreadsheets • ... and more! Open-source and free to get started.
77
225
59
2,289
382,806
ququzone.eth retweeted
nothing mentioned in this thread justifies another rollup every feature they list is true for every other OP stack chain who asked for this? why does this exist?
Unichain mainnet is live ✨ ✸ Fast with low fees ✸ Built for cross-chain liquidity ✸ Prioritizes decentralization from day one You can now deploy DeFi apps, launch tokens, swap, provide liquidity, and more
54
6
7
308
75,886
ququzone.eth retweeted
"rag killer" in 20 lines of code with gemini 2.0
88
110
27
2,065
448,098
Fully Local RAG 🔥 - Docling from @IBM, Open Source Library to make documents LLM Ready - @langchain as framework for LLM - @qdrant_engine as vectorstore - Deepseek-R1 via @ollama - ChatUI using @chainlit_io - Embeddings model via @huggingface Video link in 1st comment #ollama #deepseek #qdrant #langchain #chainlit #docling #rag #ai
5
29
136
11,413
One of the many things that we need to acceeeeeeelerate is abolishing groth16 Per-application trusted setups are just not ok in the 2020s. Universal setup at the minimum, ideally no setup at all This requires big improvements on infra and standardization for newer SNARK algos
Replying to @Zac_Aztec @Aptos
Would really like to: Groth16 trusted setups are extremely annoying! We'd need to: 1. Solidly understand all aspects of PLONK 2. Minimize prover time (e.g., custom gates) 3. Ensure verifier time matches Groth16's 4. Find a sane DSL to re-write it in (@NoirLang ? 😉)
639
175
51
1,588
518,197
Has any AI agent framework integrated yet with account abstraction session keys for agents w/ true non-custodial wallets? DM me? You could easily map function calls API to session key scopes and do magic, and I might have some good tools to share with you.
73
11
4
297
44,198