替你踩 AI 订阅的坑 额度 封号 涨价 显卡内存 先挨刀的是我 | 中/EN/ES | https://xitter.cf/t.co/ZGNmVLje8T
Joined September 2012
- Tweets3.2K
- Following549
- Followers536
- Likes1.2K
Pinned Tweet
PHP 这次两个洞 我把来龙去脉从头扒了一遍..
谁挖出来的 哪些版本有补丁 老版本各平台补到哪 关 upload_progress 为啥只挡一半 连日本首相官方账号都出来提醒打补丁 全写进去了
手上有 PHP 站的 今晚真别拖了啊
全文放楼下👇
#1 on Artificial Analysis and you can't touch it
- AA v5 index: 59, top spot
- who gets it right now: "a subset of cyber defenders". devs get "as soon as possible", no date
- the $2/$10 everyone's quoting is the launch price. standard is $4/$20, and AA says the discount has no confirmed end date
best model you can't use, at a price that won't last. very Google
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!
所以说这个野榜谷歌充钱了？
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!
523 粉 离 1000 还差 477 个陪我一起骂额度的人
我天天在这蹲 Claude Codex 封号 降智 偷偷改额度这些破事
玩 agent 写代码 折腾本地模型的 留个言 互关 我挨个回
只回关 AI 圈真人 卖号中转 币圈喊单就别来了
Tibo 真把 poteto.com 买下来了 😂
扒了下网页源码 20 个土豆随便点 每个一句冷笑话 点中那个幸运土豆弹 "Hooray, you found me!" 撒完彩带 5 秒后直接给你跳 chatgpt.com
poteto 回了句 "i cant afford tibo.com" 一查 tibo.com 是家德国深孔钻机公司 钻头赢麻了
这是回敬她把 bot 邮箱起名叫 tibo 还是白捡一个广告位？？
出处 Tibo 原帖 "poteto.com has landed" xitter.cf/thsottiaux/status/2115…
poteto 回 "i cant afford tibo.com" xitter.cf/poteto/status/21153050…
美国银行刚给客户群发了封邮件 标题叫 Protecting your finances in the AI World 意思很直白 网银账号密码别交给任何外部服务 "including an AI-enabled tool or agent that offers to act on your behalf" 狠的是后面这句 "If you do share your credentials, keep in mind that you're responsible for activity resulting from that access." 翻译一下 你非要把密码喂给 agent 它乱下单 账户被当成盗号锁掉 或者 agent 自己有漏洞把密码漏出去 锅全是你的 😅 美银发言人已经跟 TechCrunch 确认邮件是真的 这边各家天天喊 agent 替你付款 那边银行免责条款先写好了 夹在中间的只有用户 你们敢把网银交给 agent 吗？？
出处 TechCrunch 原文（邮件原话和美银确认都在里面） techcrunch.com/2026/10/27/ba… 想让 AI 碰钱又不想背锅的 这份清单可以对一下 binmage.com/zh/ai-agent-bank…
Tibo Day 23 先拿 iCloud 邮箱和日历插件逗你一下 然后 just kidding 今天是重置😂
楼中楼补了一句 Lands by midnight PST
换成北京时间 最晚今天下午 4 点前到账
手里 Codex 额度还剩一堆的 现在就去狠狠用 到点反正直接回满
见底的先别急着升套餐 下午就回血
Codex 额度现在跟天气预报一个看法 盯 Tibo 发帖就行
出处 Tibo 楼中楼原话 xitter.cf/thsottiaux/status/2115…
重置时间记录 aihot.news/codex-reset
中国 AI 泡沫 路透直接写「已经破了」 我看完截图第三个红框直接蚌埠住
科创50 不到 4 个月跌 40% 蒸发约 4800 亿美元
宇树上市以来腰斩 摩尔线程 沐曦 7 月以来也腰斩 中际旭创 新易盛离高点跌了六成左右
「泡沫在破」是路透标题和北京一位基金经理 Dong Baozhen 的原话 图里圈了
最离谱的是跌完这样 科创50 还挂着约 100 倍市盈率 高点那会儿快 250 倍 纳指100 才 30 倍
同期纳指也就从高点回撤 6%
美国还在吵 AI 到底有没有泡沫 上海先替全世界跑了一遍 demo 🙃
出处 路透 10-26 分析稿 In China, the AI bubble has already burst（US News 转载全文）money.usnews.com/investing/n…
谷歌搜「GPT-6 Astra 桌面版下载」的 手先停一下
Check Point 刚挖出个叫 Yoba 的木马 在 GitHub 上建假项目 冒充 GPT-5.6 Sol、GPT-6 Astra、Claude Opus 桌面版 谷歌搜出来还排挺前
最阴的是压缩包里那个 Setup.exe 是真的 带 OpenAI / Anthropic 官方签名 旁边塞了个恶意 dll 一运行就被顺手加载 签名验了也白验
报告里抓到的最终载荷是 XMRig 挖矿 但加载器还能再拉别的程序和 PowerShell 脚本
OpenAI 和 Anthropic 本身没被黑 纯蹭新模型名字骗下载
防法就一句 只从 openai.com 和 claude.ai 官网下 解压出来 exe 旁边多个来路不明的 dll 整包删
新模型客户端还没出 骗子的客户端先出了 交付速度吊打大厂😅
出处 Check Point Research 原报告 research.checkpoint.com/2026…
musk's "new features in grok bot" post pulled millions of views. then a zynga founder tried to pay for it and couldn't
- Mark Pincus: so many steps to pay he failed 2x
- he asked for some limited free use daily
- musk: "Major fail, will get fixed pronto"
shipped 1Password and Apple integrations before a checkout that works. very on brand
谷歌半夜扔了个 Genie 4 世界模型 一句话改天气 加角色 挪镜头 里面的人还会开口跟你对话
看完演示已经想进去逛一圈了
结果翻官方博客 现在只给一小撮受邀测试者 Ultra 会员和 API 都还写着 coming soon
两百刀一个月的会员都在排队 我们先云玩吧
出处 官方博客原文 blog.google/innovation-and-a…
Pro 和 Ultra 到底差啥 要不要为 Genie 4 上 Ultra 整理在这 binmage.com/zh/gemini-member…
wow
@thsottiaux My usage randomly went down from 93% in the morning to 33%, even though I had hardly used Codex. It then disappeared to 0% after I sent in one prompt.
I'm a $500/month user, and Codex usage limits have been very frustrating for me. I've been burning through the weekly limit in one day, and the usage limit has dropped 10x compared to when I was leveraging it while just using the $200 plan last month.
After I used my full reset, it went all the way up to 100%, then all the way down to 0%. After I reset it and had my reset removed, this is extremely frustrating and a terrible experience. I wasted credits as well. Although I was given additional credits because of the mishap and was at 70,000 last week, I now have 30,000 credits left.
This has been extremely frustrating. Please get this resolved ASAP.