AI agents, tech, and growth | Building at https://xitter.cf/t.co/HRQ96g80hK blogging about my journey @ https://xitter.cf/t.co/mmPgJcIsV4 and pushing to https://xitter.cf/t.co/LAiwbXURmU
Portland, OR
Joined April 2021
- Tweets891
- Following320
- Followers172
- Likes1.5K
tarlow 🌱 retweeted
📢GIVEAWAY📢
As a way of saying thank you for being rad, one lucky winner will get it all! 🔥😱
To enter: FOLLOW, LIKE, & RT✔️
One person takes it all:
🔹️Atari Console with 130+ games
🔹️Tomagotchi in box
🔹️2 Nicktoon toys
🔹️Complete in box Annalog NES
Ends 10/30🔥
tarlow 🌱 retweeted
Not just PHP. Node.js has also announced security releases for October 21, with HIGH-severity vulnerabilities affecting its 26.x, 24.x and 22.x release lines.
If you run either stack in production, keep an eye on official advisories and apply patches promptly.
Security maintenance doesn't stop at your application code. Your runtime matters too.
Node.js advisory: nodejs.org/en/blog/vulnerabi…
tarlow 🌱 retweeted
An urgent notice from the Railway team
blog.railway.com/p/security-…
tarlow 🌱 retweeted
Railway is compromised, rotate all API keys as soon as possible
Deployments are down so don’t wait to redeploy
Attackers are draining keys in real time, we didn’t lose much but can confirm this is still an active attack
Stay safe friends
tarlow 🌱 retweeted
Have you experimented with the newest Nano Banana 2.1 model yet?
It creates more natural-looking imagery, with significant improvements in visual design, editing and subject consistency.
Image generation costs 0 credits in Google Flow. Try it out!
tarlow 🌱 retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
tarlow 🌱 retweeted
It’s always harder, takes longer, and costs more than you think it will.
And it's still worth it.
tarlow 🌱 retweeted
MiniMax M3.1 Flash: Space Performance + Bunny Speed
🚀🐰 Yes, Space Bunny - which has topped global leaderboards with 200T+ tokens - was an earlier version of M3.1 Flash.
💰Redefining the performance economics of agentic AI. M3.1 Flash achieves frontier-level intelligence across key tasks such as Cowork (Automation-Bench: 62%; ALE: 30%), Coding (DeepSWE: 74%) and Visual Reasoning (BabyVision: 91%), at 1/3 the cost of M3 and 3x the speed.
API pricing (flat rate): $0.15 / $0.60 per 1M input/output tokens, with cache reads at $0.015 / 1M tokens.
Model weights will arrive on Oct 23.
API: platform.minimax.io
M Plan: minimax.io/m-plan
MiniMax Code for desktop: agent.minimax.io/download
tarlow 🌱 retweeted
Nano Banana 2.1 (`gemini-nano-banana-2.1`) is now available. Outperforming previous Pro model at 4x cheaper price (now $0.034 per image vs $0.134).
🕒 Improved instruction following
🔤 Better in-image text rendering
🔍 Grounded with Google Image Search
🎭 up to 5 characters consistency + 14 references images
Available in @GoogleAIStudio, the Gemini API, @googlecloud, the @GeminiApp, and @FlowbyGoogle: aistudio.google.com/docs/ima…
tarlow 🌱 retweeted
introducing @NanoBanana 2.1 🍌
our latest image generation model outperforms our previous models across the board:
- better visual design
- mask-based editing
- subject consistency
- more natural-looking images
try it out today in ai.studio
tarlow 🌱 retweeted
THE REBELLION SHIRT
$1 with code REBEL
Limited edition of 2222, 1 per customer
shop.nousresearch.com/produc…
tarlow 🌱 retweeted
Decision models in llama.cpp are now available
The `/v1/systemone` endpoint is available in the latest llama builds. Use it to do Jev-style inference locally, efficiently and privately. Multiple open models are supported with more to come.
huggingface.co/blog/ggml-org…
tarlow 🌱 retweeted
🚀 Gemma 4 Developer Agent Competition is live! Post-train an open model into a coding agent that speeds up dev workflows on everyday hardware. Don't miss the opportunity to compete for a $100K prize pool by November 25, 2026.
Competition: kaggle.com/competitions/gemm…
Paper track: kaggle.com/competitions/gemm…
tarlow 🌱 retweeted
Six weeks. $60,000+ in prizes.
Drum roll please! The PayPal AI Hackathon: Build What's Next with PayPal and AI is live.
Whether you're a developer, an AI builder, a founder, or a student, here's your chance to use your imagination and build with AI, with PayPal has your back as the trusted payment layer.
With free tools, credits, and support from our partners: Postman, Zapier, Elastic, APIMatic, AG Grid, Bryntum, Channel3, Render, , Astropods, Kernel
Submissions open Oct 1 and close Nov 12.
Learn more and register: paypalaihackathon.devpost.co…
tarlow 🌱 retweeted
🚨突发！疑似Anthropic再次钓鱼执法：免费送Claude实体玩偶，中国地址也能填！
表单：国家一选China，立马弹出National ID Card Number。
玩偶：戴着牛仔帽看着你，像在说「先把身份证交出来」。
别人领周边，你领的是一次实名认证。
搜索claude dev之前，先问问自己：一只橙橘子，值不值得18位。
#Claude #Anthropic #claude_dev #钓鱼执法 #身份证号换玩偶 #Dario #A社 #封号
🤣国庆假期可别躺平兄弟们！
Dario可能正在对着你的工作记录做交叉验证：
平时周日还在新加坡/日本/台湾IP敲代码，假期突然消失，直接进待封名单。
所以国庆期间请维持正常Claude使用频率。实在要放假。
建议找个 “节假日代蹬” 帮你每天问两句帮我写个周报。
本地人在喝kopi，你在续命。无语了啊。
#Claude #Anthropic #国庆不休息 #AI打工魂 #代蹬计划 #Dario请睁一只眼闭一只眼 #Dario #Claude封号
tarlow 🌱 retweeted
🔥抢先看！OpenAI DevDay 2026 最全新品清单汇总来啦！
一口气砸了 20+ 项更新，ChatGPT 正在变成「人 + 智能体」的工作操作系统 ！
1️⃣Dots：全天候常驻智能体，自己有云电脑，能跨对话持续干活
2️⃣GPT-6.1 Sol：接近 Astra 的智能，价格只要约 1/5
3️⃣Ultrafast：最高约 8 倍生成速度，300 tokens/秒
4️⃣ Pro 500：$500/月顶配额度，解锁 Astra Ultrafast
5️⃣Codex Cloud：电脑/手机/云端都能接着写代码，环境可复用
6️⃣Codex CLI 2.0：语音指挥、/agents 多任务并行
7️⃣Code Review：自动审 PR/MR，人走开它也先看一遍
8️⃣Codex Security Cloud：扫仓库、去重、直接准备补丁
9️⃣Decisions API：把 Luna 压成实时决策引擎
🔟Agents API + Computer Use：智能体能直接操作软件
1️⃣1️⃣Bedrock Managed Agents：OpenAI 智能体可整段跑在 AWS
1️⃣2️⃣Plugin Extensions：插件能占侧边栏、做交互面板和文件查看器
1️⃣3️⃣Sites 可托管插件 + MCP Events：网页应用能被事件自动唤醒
1️⃣4️⃣ChatGPT Space / Pages / 协作幻灯片：人和 Agent 共编同一份工作
1️⃣5️⃣Slack/Teams 里直接 @ChatGPT，同事不用单独开会员也能一起用
1️⃣6️⃣Meetings 插件：开会自动记笔记、出待办，音频用完即删
1️⃣7️⃣Sign in with ChatGPT：额度能带到 Devin、Notion、Vercel 等 16 家工具
1️⃣8️⃣ Private Intelligence：零数据留存 + 私有推理预览，企业更敢上
ChatGPT 不再只是聊天框，而是 12 亿用户面前的原生应用平台。
#OpenAI #DevDay2026 #ChatGPT #Dots #Codex #GPT6 #AIAgent #开发者大会
tarlow 🌱 retweeted
🚨重磅！DeepSeek V4.1 Pro 已现身内部测试，蓝鲸即将出水！
🔹显示名锁定：DeepSeek V4.1 Pro
🔹标语：Early access，按账号灰度，公共列表里看不到
🔹默认高推理（reasoningEffort = high）
🔹数据用于训练，非多模态、非 premium
🔹权限开了就能一直用，距离正式上线不远
#DeepSeek #DeepSeekV41Pro #AI大模型 #开源模型 #EarlyAccess #梁文峰 #梁圣
🔥速码！DeepSeek Harness桌面端新手指南来了！
40天后DeepSeek Harness已经不是极客玩具，真的能当个人智能工作台，越来越像Codex了
实测最关键的14个注意事项：
1️⃣ 启动极简：双击图标就能开，首次引导选用途，建议办公+代码开发都勾上。
2️⃣ 工作过程可调：三种展示模式本质一样，只是思考/工具调用折叠程度不同，设置里随时改。
3️⃣ 默认值会被用途影响：只选办公则代码工具默认关、过程简洁；勾了代码开发才会问过程、开工具、显示diff和Agent预设。
4️⃣ 账号余额直接打通：用官方DeepSeek账号登录，赠送额度+剩余余额都能在桌面端用，控制台还能直接充值。
5️⃣ 新增Auto Review权限：类似Claude Code/Codex/Workbuddy的实验性自动审查，很多人默认开着。
6️⃣ 插件一键管理：左上角新增插件按钮，可开关Auto Review、Agent Team、语音输入，右上角直接添加。
7️⃣ 侧边栏终于能用：工作区文件、终端、浏览器全有，还能原生预览Word/PPT/Excel。
8️⃣ 语音输入注意：要下200MB+识别模型，当前内测版macOS麦克风授权有坑，Windows正常。
9️⃣ Agent Team会烧Token：开启后有专门视图，默认只有Lead，可切subagent对话栏。
🔟 完整团队提示词示例：researcher查资料、checker核对事实、writer出初稿，最后还能加把关agent。
1️⃣1️⃣ 定时任务可开：改yaml三行或直接跟DSH说，左上角会出现自动化按钮，还能推到手机。
1️⃣2️⃣ 装插件超简单：界面添加，或直接在对话框让Agent帮你装。
1️⃣3️⃣ 创造模式自己做插件：新开会话选创造模式，让它做侧边栏GLB 3D预览插件，做好立刻能用，不用重启。
1️⃣4️⃣ 结论：插件生态、Preset、工作台理念都已经能落地，人人都可以有自己的智能体工作台。
#DeepSeek #DeepSeekHarness #DSH #AIAgent #桌面AI #智能体 #VibeCoding #AgentTeam #梁文峰 #崔添翼
