Kirtaner / Milady Maker for 💞Game Boy!!💫 retweeted
Qwen 3.8 27B is actually the most brilliant retarded model that I have ever used. 😅
It spends about 2/4 of its reasoning trace coming up with the most retarded ideas known to man, and then an additional 1/4 convincing itself that its ideas are completely retarded.
It then manages, from the ashes of retardation, to arise towards the cleanest, most efficient, and bug-free implementation of any open-weight model of this size.
It kinda works like American politics, in a way?!
[pats hypurr]
BREAKING: $HYPE BREAKS $95 AS @elonmusk REACTS "INTERESTING" TO LEGENDARY INVESTOR STANLEY DRUCKENMILLER (@standuquesne) BACKING @tradexyz, WHICH IS BRINGING 24/7 @HyperliquidX PERPS TO THE U.S. WITH @NYSE PARENT @ICE_Markets AND @okx
One inmate I did time with, was a kid running LOIC in Operation Payback in 2010. Snuck up on me and started speaking in memes and my first thought was "oh god what is happening". Spent the next two days telling the range I was a legend. Also, only inmate that gave me his LinkedIn
>when you feel the urge to buy a printer in 2026 just to print out research papers from arXiv
I had a co-worker during my time at Ericsson back in 2016-2017 that was extolling to me the work happening at Google Deepmind. I should have paid closer attention myself.
There has never been a better time to be an autodidactic polymath in human history, and that's beautiful.
K-shaped economy but it's actually independent capability vs the neurotypical normies.
AI might take *your* job, but some of us will thrive like we never have before.
Some of this is fearmongering that leads into the interests of Dario directly as they pull the ladder up behind them to try and protect market dominance. Do not let them win.
Shoutout to the whitehat hackers that kept me sane in prison as support structures via my limited phone access. You know who you are
Kirtaner / Milady Maker for 💞Game Boy!!💫 retweeted
终于有人把“本地大模型打真实电话”做出来了,而且完全开源!
以往体验市面上的各类 AI 电话助手,最让人顾虑的就是隐私安全:通话录音、转写文本、沟通对象全都会留在第三方云端。开发者刚刚开源的 ai-phone 则彻底打通了本地闭环——让你的 Mac 直接接管真实电话线路,所有推理均在本地完成,通话数据完全不出机器。
🛠️ 核心技术栈(基于 Apple Silicon & MLX):
语音识别(STT): Parakeet TDT 0.6B 本地运行,极速语音转文字
语言模型(LLM): 默认本地运行 Qwen3.5 4B(32GB 内存支持 9B 4-bit),同时也支持接入自己的 Claude 或 ChatGPT 端点
语音合成(TTS): Kokoro 82M / Phonon,生成自然流畅的声音
通信接入: 通过 SIP 协议直连 Telnyx 或 Twilio,直接购买或绑定真实电话号码(月租约 1 美元)
✨ 核心亮点:
低延迟与实时打断: 在 Apple Silicon 芯片上实测首字响应延迟约 1~1.5 秒;支持实时插话打断(Interruption),一旦你开口,AI 会立刻停下倾听。
支持外部 Agent 联动: 提供了 call、listen、say、hangup 等标准化 CLI 接口,可以直接让 Claude Code、Codex 等编码智能体通过命令编排通话,自动完成预约餐厅、电话咨询等任务。
本地守护进程: 提供向导式 CLI 配置,一行命令拉起本地守护进程,静默接听来电或随时发起外呼。
⚠️ 当前要求与限制:
仅支持搭载 Apple Silicon 的 macOS 系统
目前为单路通话,优先支持英文环境
项目刚刚上线,对本地 AI 语音助手与 SIP 通信落地感兴趣的朋友值得关注:github.com/oxnw/ai-phone
It isn’t vibe coding when you can define an entire production-grade spec. This is what separates amateurs from professionals and is your moat. This will never change.
KV-cache transfer: cat is out of the bag. My predictions are that, by 2028Q1, (a) CoT monitoring will be unambiguously over, (b) inference-time scaling curves will get better slopes and higher plateaus, and (c) context-window size will matter a lot less.
MITの研究者らが、AI同士が言葉を使わず頭の中の状態を直接渡して通信する仕組みを作り「テレパシー」と名付けました。
文章でやり取りするより平均40%正確で4倍速かったといいます。
また「画像を見られるAI」が写真の内容を「テキストしか読めないAI」にテレパシーできるそうです。
写真の内容をテレパシーを受信した「テキストしか読めないAI」は、テキストで説明を渡されるよりも写真についての質問に正しく答えられたとのこと。
なお、別の会社のAI同士でもテレパシーできるようにしたのが今回の研究の一番の新規性のようです。
例:
Meta(Llama 3)→ Alibaba(Qwen 3)
Alibaba(Qwen 3)→ Google(Gemma 2)
Google(Gemma 3)→ IBM(Granite 3)
Google(Gemma 3)→ Meta(Llama 3)
Mistral AI(Ministral 3)→ Meta(Llama 3.2)
Google(Gemma 2)→ Mistral AI(Ministral 3)
IBM(Granite 3)→ Hugging Face(SmolLM3)


















