Gemini 4 Carbon 对长任务执行和网络安全能力的强调,让我更关注 AI 在实际工作中的可靠性,模型能力的进展,仍需要放进真实任务中检验。
拿 Web3 项目研究来说,资料整理得再完整,也需要分清:哪些是可核实的事实,哪些是项目方的表述,哪些暂时缺少证据。
使用非小号AI时,可以先从这样一个具体任务开始:“整理这个项目的公开资料,区分已确认事实与待核实信息,并列出需要继续查证的问题,”再结合原始来源逐项核对。
我认为,评价 AI 是否实用,可以看它减少了多少重复整理、需要多少人工纠错,以及交付的结果是否方便复核。这些细节,直接影响它能否融入日常工作。
I did say we'd be releasing our models periodically...introducing Gemini 4 Carbon, our next frontier model.
Following Argon's release just two weeks ago, Carbon shows the pace at which we’re pushing the frontier, it ranks #1 on the recently updated v5.0 of @ArtificialAnlys Intelligence Index at 59. It also leads across key cybersecurity evaluations scoring 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for exploitation, and 71.0% on CWE-bench for patching.
We’re seeing significant improvements in long-horizon tasks even compared to Argon, and we’ve strengthened frontier safeguards.
As we did with Argon, we’re going to roll it out gradually + responsibly. Some trusted cyber defenders will get access starting this week. We'll be expanding it to more partners through our Fairwind Program, ahead of wider availability soon, so hang tight!
Saw Binance Day 33's three-layer stack. The replies keep asking which layer is "stronger." Cheaper first step: align where each layer stops—and which one can touch your funds or permissions.
① Where do Binance AI / AI Pro / Agent OS officially draw the line?
② Do you want in-exchange "approve-then-run" automation, or cross-source comparison to form a judgment?
③ After reading, which lines are announced capabilities—and which did you fill in as "it makes you money"?
Stuck on how exchange AI differs from market-intel AI? Drop it to Feixiaohao AI—it'll help you find sources and break it down by what happened / why / risks. Not investment advice.
#AI
Seeing Ondo × Robinhood × Uniswap?
Don't read it as "US stocks are now freely tradable 24/7."
Three questions against the public docs usually clear it up:
Same ticker, two tokens — same issuer?
On weekends, is mint/redeem closed, or secondary trading?
When they describe swapping at a "big enough discount," is that a path — or a return promise?
Want those three checked against the blog? Drop them to Feixiaohao AI — find the source, unpack the terms, verify today's page. Not investment advice.
Ondo Stocks, now on Robinhood Chain.
Live today, 15 Ondo tokenized stocks and ETFs can be bridged from Ethereum and BNB Chain to Robinhood Chain.
11 new @Uniswap StablePair pools connect Ondo Stocks and Robinhood Stock Tokens, bringing 24/7 liquidity to the chain’s DeFi ecosystem.
What this unlocks:
→ New potential arbitrage opportunities
→ More ways to trade and earn LP fees
→ Connectivity to Ondo Stocks’ 24/7 liquidity
With this expansion, more traders can access Ondo Stocks and more developers can build financial applications around them.
807 亿美元净亏,是现金烧掉了吗?
不全是。大半来自可转债(能转成股票的债)的公允价值(账面按市价重估)变动——公司估值涨了,会计上会记一笔亏。
上半年营收约 163 亿美元,经营亏损从约 21 亿收到约 6.9 亿。
下次看到吓人标题,先问一句:哪些是现金亏,哪些是估值记账?
Anthropic has filed an S-1 registration statement with the SEC related to its proposed IPO.
Read more: anthropic.com/news/s1-sec
Legal disclaimer: A registration statement relating to these securities has been filed with the SEC but has not yet become effective. These securities may not be sold, nor may offers to buy be accepted, prior to the time the registration statement becomes effective. This post does not constitute an offer to sell or the solicitation of an offer to buy these securities, nor shall there be any sale of these securities in any state or jurisdiction in which such offer, solicitation or sale would be unlawful prior to registration or qualification under the securities laws of any such state or jurisdiction.
非小号AI快讯|GPT-6.1 Terra 发布:更强操作能力,更低调用成本
10月27日消息,OpenAI 发布 GPT-6.1 Terra,强化视觉推理、屏幕理解与电脑操作能力,可用于技术图纸解读、表单填写、视频编辑等工作场景。其标准 API 输入、输出价格均较上一代 GPT-5.6 Terra 降低50%。
据 OpenAI @OpenAI 介绍,GPT-6.1 Terra 强化了视觉理解与电脑操作能力,可用于图纸解读、表单填写、视频编辑及发票信息录入等场景。在官方公布的 Agents’ Last Exam 评测中,新模型在中等推理强度下,得分较上一代提高15个百分点,单次任务成本约为其三分之一。
同时,其标准 API 输入、输出价格均较 GPT-5.6 Terra 降低50%,分别为每百万 Token 1美元、5美元,目前已通过 OpenAI API、ChatGPT Work 和 Codex 提供。
模型持续升级,行业应用也在持续进化。
非小号AI持续跟进底层模型能力更新,将不断升级的理解、分析与工具使用能力,与非小号九年的 Web3 数据、行业知识及专家工作流结合,服务项目研究、信息整理与持续跟踪等具体需求。
从看懂信息,到分析问题,再到协助完成工作——模型提供持续进化的 AI 能力,非小号的行业积累让这些能力更贴近 Web3 用户。
了解 Web3,就问非小号AI。
feixiaohao.ai/fclaw
#非小号AI #OpenAI














