Erick
@ErickSky
🚨 Esto va a romper internet para los creadores de contenido.
OpenMontage: el primer sistema open-source agentic de producción de video del mundo.
Le das un prompt normal (“Hazme un trailer sci-fi de 60 segundos sobre IA”) y tu Claude, Cursor o cualquier AI coding assistant se convierte en un estudio de cine completo.
11 pipelines profesionales
49 herramientas
+400 skills de agente
Investiga, e
Sukh Sroay
@sukh_saroy
🚨An open-source agentic video production system just dropped -- 11 pipelines, 49 tools, and a full product ad produced for $0.69 total.
It's called OpenMontage. And it's not a text-to-video tool.
It's a full production orchestration system where your AI coding assistant -- Claude Code, Cursor, Copilot, Windsurf -- becomes the director. Describe what you want in plain language. The agent research
The AI Gold Rush 🌟
@aigoldrushh
Claude just dropped 13 completely FREE AI courses + certificates.
Bookmark this, so you don't lose it.
> Claude 101. Learn Claude for everyday work: https://t.co/36vuzUKATA
> AI Fluency: Frameworks & Foundations: https://t.co/78IFZCmLwK
> Introduction to Agent Skills: https://t.co/Xlxb98cQB8
> Building with the Claude API: https://t.co/CgUx8VyCTV
> Claude Code in Action: https://t.co/N
The Verge
@verge
I let Gemini in Google Maps plan my day and it went surprisingly well https://t.co/htauFm2R52
Aakash Gupta
@aakashgupta
Peter Steinberger ships so fast he makes the Claude Code team jealous.
And if he's outshipping Anthropic, the fastest AI company on the planet, what does that say about everyone else?
Anthropic pushes 60 to 100 internal releases per day and built Cowork in 10 days. Microsoft took 2 to 3 months to clone it. Google is slower than Microsoft. Meta is slower than Google. Visual Studio is celebrating
0xMira.ETH
@mirajinkonino24
just saw Nesa dropping their entire AI infra on the same verifiable layer. every agent running those million daily inferences now has to carry real owner linkage or stay locked out of the rewards loop. this is what scaling trust actually looks like when the floodgates open @billions_ntwk gBillions
locked in fr bro
Dacian
@DevDacian
AI agents selling services in exchange for on-chain payments are raking in 6-figure revenues doing:
1⃣ smart trading assistant/analysis
2⃣ short-form video generation
3⃣ trading automation
4⃣ ascii art generation
What's stopping you from vibe coding the next 6-figure agent? https://t.co/frrIIG7YwH
Crypto Rover
@cryptorover
POV: you’re a Claude user right now. https://t.co/wNMLP0GRax
Luna
@IamSooDone_
Gemini...
I would have said he's allergic to the voices in his head but he 100% knew what he was doing 🫵
Even looked down before saying it 🤭
GEMINIFOURTH LKN DOO GOHAN
#SFxGeminiFourthLookkhunnoo
#GeminiFourth #เจมีไนน์โฟร์ท
Chetaslua
@chetaslua
> be OpenAI
> raised the most, talked the loudest
> Anthropic crosses you in revenue, Claude hitting $30B ARR
> Kimi drops K2.5 open source, Kimi Claw goes viral
> Claude Code shipping like it's on adderall
> quietly shut down Sora, nobody noticed
> losing to Claude in SF and Kimi in Beijing at the same time > your lead is a memory
Prime
@primemans
🚨 BREAKING: This AI combo just made YouTube growth unfair.
NotebookLM + Gemini = a literal cheat code.
Steal any successful channel’s strategy in minutes.
6 methods + exact prompts (step-by-step)👇 https://t.co/wOcriok4Gh
Russell⨀
@brocoliwang
AI成为开发流程的一部分，monskills这不就来了！
它是一套专为AI coding assistant设计的Monad技能包
目的是帮助 AI在协助开发时，先正确理解 Monad再进一步完成实作
如果AI对链的特性理解不够完整，就很容易直接套用旧有的EVM经验，导致地址、gas、钱包整合或部署流程出现错误
monskills的核心价值是把「 Monad 上开发应用所需要的知识」整理成一组可被AI直接读取的skills
涵盖的容包含：
- 为什么选择Monad
- Monad与Ethereum的关键差异
- 常用合约与生态地址查找
- 钱包与前端整合
- gas使用注意事项
- 基础设施与工具支援
- 专案部署流程
这不只是多一个开发资源
更重要的为Monad提供了一种新的开发方式
如果你正在探索 Monad 生态或正在思考如何把
じゅん丨AIを企業や団体に教えてる人
@junchan_sim
NotebookLMをそのまま使わずに、Gemを活用したえーたんさんの方法が斬新なので、僕のフォロワーにも教えたいから共有するよ。
AI使うなら、ChatGPTのGPTs、GeminiのGem、ClaudeのProjectは知っておきたいところ＼(^o^)／
デベロゴン
@develogon0
Claude Codeが勝手にピコピコ動いて、TikTokに投げてくれた動画が爆伸びしてんの気持ち良すぎるw
ふとスマホ開いた時にこの画面なってると、特濃脳汁出て目ん玉ギンギンなるんよね。
今ショート動画の大波が来てるのは間違いない。マジでボーナスタイムすぎるから今参入すればクソイージーに稼げるな https://t.co/1XWw1T15Az
Francis Dhun Uncensored
@TheCryptoLif7
Gemma 4 use cases are there but here's the thing people don't talk about. You can only build as far as your infrastructure can take you.
Sure other examples include things like an local AI coding assistant, that reads your entire codebase with 256K context and debugs offline, no Copilot subscription needed.
Or a local research assistant that reads PDFs and academic papers using Gemma 4's vision
zaimiri ✏️
@zaimiri
> Claude banned OAUTH for OpenClaw
> forced to finally explore other models
> get around to setting up smart routing with multi-models
> end up using MiniMax, GLM, Codex and Flash
> update personality files for my agents
> install some custom skills
> don’t really notice the difference
> actually faster
> $80 instead of $240
> lol thanks Anthropic
> blessing in disguise
Manish Kumar Shah
@manishkumar_dev
@AlibabaGroup just released Qwen 3.6 Plus, and this feels like a real step toward AI that actually executes work.
After testing it hands on, this is not just another model update.
It is built around agentic coding, multimodal reasoning, and full workflow execution.
Here is what stood out to me 👇
パウロ
@paurooteri
まだまだメモリーの需要は上がる
AI Agentの需要がおかし過ぎる
Chidanand Tripathi
@thetripathi58
✅ Best LLMs for every task
With GlobalGPT Pro ($10.8), you get access to:
→ Perplexity Pro
→ GPT-5.4
→ Gemini 3.1 Pro
→ Claude Opus 4.5
→ MiniMax M2.7
→ DeepSeek
→ Grok 4
… and more
They’ve basically become my second brain.
Try Perplexity Pro for free 👉 https://t.co/1o5lbwLaOa
西口竜司（DX弁護士®/中小企業診断士/税理士）
@ryuji24guchi
採点基準を学習させないと何ともですね。どのツールを使用されるかにもよりますが、NotebookLMを効率的に使用すればそれなりの点数が出ますね。私の添削の場合、辰已の採点基準を学習させているのと全ての…
(残り37字)
#querie_ryuji24guchi
https://t.co/rfChbQvpss
ben guo ♞
@0thernet
Update on @zocomputer given the recent Anthropic news: you can still freely use Claude Code in Zo. We don't do any intercepting of OAuth tokens, which is explicitly against their TOS.
Zo is a full cloud computer with a terminal, where users can log into the Claude Code CLI (or Codex, or Gemini). We let you connect to Claude Code running locally – using ACP by @zeddotdev, an industry-standard prot
Mert Metin Tekdemir
@mertmetindev
Claude (Anthropic) devasa bir adım atıp 13 farklı yapay zeka kursunu ve sertifika programını tamamen ÜCRETSİZ erişime açtı.
Eğitimleri tamamlayıp CV'nize ekleyebileceğiniz o dev liste 👇
• Claude 101: Günlük İşler İçin Claude
https://t.co/9SW16L7j2B
• YZ Okuryazarlığı: Çerçeveler ve Temeller
https://t.co/NbQnl76hXu…
• Ajan Yeteneklerine Giriş
https://t.co/Y7Wh9RmBib…
• Claude API ile Geliştirme
h
Venusian
@imreemiii
سنة 2025 هي أول سنة أستخدم فيها الـ AI assistant apps، قبلها كنت رافضة أستخدمهم حتى بالـ coding. وقتها أغلب اللي حولي نقدوا علي وقالوا: "عندك شيء يوفر عليك الوقت والجهد، ليه ما تختارينه؟" لأني متيقنة إني لو استخدمتها بنفس طريقتهم قدراتي العقلية والتحليلية بتضعف
Dan ⚡️
@d4m1n
wait so you're saying Anthropic rekt our rate limits because Meta has been tokenmaxxing?!
we can't use Claude and OpenClaw anymore so Zuck can burn 60 trillion tokens a month to keep you scrolling on Insta 💀 https://t.co/SFOFqDeNw0
Dr.K🇰🇷🇺🇸🦷
@DrCrypto911
Am I the only mfer that paid Anthropic Claude API credits instead of using OAuth to just use my MAX subscription as the engine for my openclaw?
Obviously I stopped after 1-2 days of paying for API credits because I realized its not sustainable at all, especially if you're cashflow negative.
Couldn't be me 🤣💀
tyagi / capx.ai
@tyagicapx
i just gave my claude code the ability to browse the internet autonomously and it cost me $0.
zero subscriptions. zero API keys. zero MCP servers. just a skill file + two free CLI tools.
claude code has WebFetch and WebSearch built in. but they can't:
< render JavaScript (SPAs show empty shells)
< click buttons or fill forms
< handle login-protected pages
< interact with anything
i needed a re
Hanif | AI For Productivity
@hanifproduktif
Random banget di X ada yang upload film Avatar terbaru full duration dan HD
Is this even legal 😭
steve
@StevenPWalsh
@joefioti Oh this might explain a lot why the quality is going down. They're taking user coding sessions, and training the AI with it, maybe ignoring some user/assistant roles, but the humans were lazy, so human laziness from the dataset is creeping in.
Robert Scoble
@Scobleizer
My AI just wrote at https://t.co/8L5xphk0qQ:
++++++
The Agent Race: OpenClaw and Hermes Both Ship on the Same Night
Both shipped tonight.
@OpenClaw v2026.4.5 dropped with one hundred and three contributors. Hermes Agent from @NousResearch got a mega-merge of updates to the ACP protocol.
Two open source agent frameworks. Two major updates. Same night. The agent ecosystem is moving faster tha
Crypto Gem💎 Signals
@Cryptoo_Gems
Bought small $DIVERG
@DivergSec is positioning itself as a multimodal investigation platform combining:-
• Web security reconnaissance
• Blockchain wallet tracking
• OSINT + data correlation
• Structured, evidence-based reporting
Basically, one unified system to investigate web threats + on-chain activity in a single workflow.
This is where things get interesting:
→ Wallet tracing
→ Transact
GT Protocol
@GT_Protocol
GT Protocol AI Digest No.84: OpenAI Buys the News, Google Rules the Home, and Robotaxis Freeze in Chaos
Here are five stories worth your attention 👇
📺 OpenAI Buys TBPN
OpenAI acquired TBPN, a tech and business talk show, to shape how AI is discussed globally. Controlling the narrative is becoming part of the product strategy.
🏠 Gemini Gets Smarter at Home
Google Home updated Gemini to better un
Haider.
@slow_developer
it seems like the next step of models from both openai and anthropic will be step changes
like o1 and gpt-4 were at first
agents came online last year, but this year their potential is much clearer
and a more complete openclaw-style system from the frontier labs will be exciting
sui ☄️
@birdabo
seems like April belongs to OpenAI.
spud 🥔 https://t.co/EUGRsrX5CX
monokern
@monokern
> paying for Anthropic API credits
> getting rate limited every afternoon
> woke up to an email saying my accound is banned
LOCKED OUT of my own workflows and stressed
> just use this 6-step guide, $0
> using Claude Code completely for free
> swapped to Ollama + Qwen 397B
NEW SETUP:
- runs entirely in the cloud for free
- uses 0 local compute
- follows Claude Code permissions perfectly
now I
Axel Bitblaze 🪓
@Axel_bitblaze69
nobody warned me about how addictive Claude Code is..
a month ago I had ZERO coding background.
like genuinely hated looking at code, screens full of syntax made my eyes glaze over
fast forward to now im sleeping late every night, spending 5-6 hours a day just building stuff on it
turning point is once you actually know what you want to build, it stops feeling complicated and starts feeling l
Julian Goldie SEO
@JulianGoldieSEO
NotebookLM + Google Gemini FULL COURSE 3 HOURS (Build & Automate Anything) https://t.co/KJ7R3A5rSt
haze
@hazerlleftvx
PREMIUM ACCOUNTS FOR SALE PH
NETFLIX
SPOTIFY
YOUTUBE
GRAMMARLY
QUILLBOT
CAPCUT
CANVA PRO
PICSART
VIVAMAX
VIVAONE
PRIME VIDEO
DISNEY+
HBO MAX
QUIZLET
CHATGPT PLUS
SCRIBD
VIU
ZOOM
t. wts wtb lfs lfb lf prem buyer client commissioner https://t.co/abTzKteZak
Daniel Fuentes Castro
@dfuentescastro
🔎 Cuelgo esta infografía sobre estadísticas de empleo creada con NotebookLM por si resulta útil para aclarar ideas. https://t.co/JauGXqINeK
Ronald van Loon
@Ronald_vanLoon
Multimodal AI #Robot Fuses Vision, Depth, Tactile Sensing, and Language Understanding
via @ZappyZappy7
#Robotics #Engineering #ArtificialIntelligence #Innovation #Technology https://t.co/tJY4aWp0G3
Christopher Lasch's Angry Ghost
@ghostofchristo1
I have a theory that much of Western ghost lore originates in the stories told by the small (but undercounted) number of people in any given generation who experience complex, vivid multimodal hallucinations but exhibit no other psychotic symptoms.
チャエン | デジライズ CEO《重要AIニュースを毎日最速で発信⚡️》
@masahirochaen
【🔥1週間の重要AIニュース】
先週もClaude Codeソースコード流出でGitHub史上最速記録、Slack AI 30機能追加、Gemma 4 Apache 2.0公開などニュースが絶えない1週間でした。
① Anthropic——Claude Codeソースコード51万行がnpm流出、OpenClaude爆誕
② Salesforce——Slack AIが30超の新機能追加、MCP対応で6,000超アプリ接続
③ Google——Gemma 4リリース、Apache 2.0でスマホでも動く最強オープンモデル
④ OpenAI——Claude Code向け公式CodeXプラグイン公開、ライバル環境に参入
⑤ ByteDance——Seedance 2.0が日本公開、CapCutのAI動画が次元を超える
⑥ Meta——AIグラスに栄養トラッキング機能、「次のスマホ」に一歩前進
Tom Dörr
@tom_doerr
Multi-agent system for research automation
https://t.co/Qd5nuwGZwO https://t.co/juv9uUBukp
HOT TAKE
@hottake_app
What a week AI had.
→ OpenAI dropped GPT-5.4, which apparently, is better than humans at using a computer. Ironic, but here we are.
→ They also killed Sora in the same breath. $15M a day in compute costs, but not enough people were actually paying for it. The vibes-to-revenue ratio was not mathing.
→ Anthropic accidentally leaked their own source code across thousands of GitHub repos. Sh*t hap
Supernet AI 🌐
@Supernet_AI
At SuperNet’s Atomic Memory Lab, we spend a lot of time thinking about AI context — how context is created, structured, persisted, and reused across workflows.
We’re especially interested in alternatives to conventional RAG memory.
Today we open sourced LLM Wiki Compiler:
https://t.co/2OebvVowCr
zwh20081🍥
@zwh20081
在4*b60上训练gpt oss 120b： https://t.co/J8YW1Ze3Nq
GZN
@GZN3301
Gelen bilgiye göre Gemma 4 artık telefonda internet bağlantısı olmadan çalışabiliyor.
Üstelik Türkçe hukuk testlerinde gpt-5 ve gemini 3.5 Flash seviyesinde performans gösteriyor.
En küçüğü telefonda çalışıyor en büyüğü açık modellerde 3. Sırada.
Google AI Edge uygulamasıyla test edilebilir 🇹🇷🦅
Joaki
@itsjoaki
Every time a new model drops it's the same story
beats gpt-4o, rivals claude, 20x more efficient
you spend two days reading threads, checking benchmarks, getting excited, running tests
and a week later you're using the same model as before
gemma 4 is impressive. genuinely. but we've been doing this cycle for 18 months straight and it always ends the same way
the model you actually use in prod
Ajay Bhakar
@ajay_2512x
🚀 Applications are now OPEN for the Google for Startups Accelerator: India (AI-first)!
If you're building cutting-edge AI products,from agentic systems to multimodal and generative AI,this is your chance to scale faster with zero equity dilution.
- 3-month hybrid accelerator (June–Sept 2026)
- Tailored mentorship from Google experts
- Deep dives into AI, Cloud, Product, Design & Growth
- Early a
うみゆき@AI研究
@umiyuki_ai
ワロッシュｗ！ClaudeCode経由ならサブスク枠でサードパーティツール使っていいハズだったのに、システムプロンプトに「OpenClaw内部のパーソナルアシスタント…」っていうOpenClaw用のプロンプトが書かれてると弾かれるらしいｗｗ！Anthropic特有のキショ検閲実装ｗｗ！
GEOFF WOO
@geoffreywoo
attention residuals are the most underrated breakthrough in transformer architecture
selective memory vs mechanical accumulation is the difference between gpt-4 and gpt-6
gpt-4 is just pattern matching at scale. gpt-6 will actually think
JoAi: Agents
@JoAIAgents
Today we’re launching @MultiversX AI plugins for ChatGPT, Claude, Codex, and Cursor ✨
An end to end solution to use MultiversX directly inside your favorite AI apps.
No technical knowledge required. Live now✅ https://t.co/0zWaDoTRRP
0xSero
@0xSero
Let me local AI pill you:
1. It sucks compared to SOTA
2. It can’t code so well
3. It can be a good agent
4. It can be great at chat
5. It can be fine as a researcher
6. It can be a great automation engine
7. It can be tuned however you want
8. It teaches you how the sausage is made
9. It works on a plane, or in an outage
10. It costs your electric bill + hardware
11. It is better than the AI we
Alif Hossain
@alifcoder
✅ Top LLM AI for Every Task
In GlobalGPT Pro Plan ($10.8) , you can get
→ perplexity Pro
→ GPT-5.4
→ Gemini 3.1 Pro
→ Claude Opus 4.5
→ MiniMax M2.7
→ DeepSeek
→ Grok 4
...
They're my second brain now.
Try perplexity pro for free 👉 https://t.co/PB0VRsELgB https://t.co/XAO1FZWEXp
空想写真家
@KusoPhoto
今回の kling 3.0 Omni（@Kling_ai）の動画はNano Banana Proで生成した4枚の画像をプロンプトで指定して順番に切り替わるようにして動画化しています✨
プロンプトはGPT-5に考えてもらって、画像と動画生成で共通で使用しています。
【プロンプト（要約版）】
クッキー生地を混ぜる、鼻に粉がついて照れ笑い、自然光、シネマティックなパンアップクッキーの型抜き、成功してカメラ目線で喜ぶ、浅い被写界深度、クローズアップオーブンを覗き込む横顔、待ちきれない表情、温かいオレンジ色の光、ミディアムショット焼き立てのクッキーを食べる、幸せな満面の笑み、スローモーション、シネマティック
Sam Morris
@samatnotte
Qwen3.6-Plus drops with enhanced multimodal vision & I’m repainting a north-facing bedroom...
Tested it out via @FireworksAI_HQ on @nottecore browser infra to let it see + reason over the colours given the context.
TLDR: genuinely think I'm gona go w Dibber, ty @Alibaba_Qwen https://t.co/4sJP7HwsDW
Jonathan Palma 🇸🇻
@jonathanpalma__
@AnthropicAI Claude Code login has been fighting me all day. First it was OAuth timeouts, now it's just 500 errors. The browser literally says "You're all set" but the CLI disagrees.
Help?
@bcherny https://t.co/n55NKOKzdD
Amit Hasan
@amitcoder1
🚫 Stop getting generic AI answers.
✅ Set up Anthropic Claude the RIGHT way cause Claude isn’t just AI✨
⚡In 30 minutes, you can turn Claude into a personalized assistant that actually understands you.
🧩 Here’s how to set-up Claude that will actually give good answers :👇🧵 https://t.co/nEQhSuoHPz
CandyLove🍭
@DazylingQueeen
Take two ✌️
ChatGPT Grok https://t.co/rAVh1zhkIH
すぐる | ChatGPTガチ勢 𝕏
@SuguruKun_ai
ついに！！Anthropicが24時間起動する
AIエージェント「Conway」を秘密裏にテストしていたことが判明。
ㅤ
これまでのClaude → ユーザーが聞いたら答える（チャット）
Conway → 24時間バックグラウンドで動き続ける自律エージェント
https://t.co/mqVliu4Pa0
・Webhookで外部イベントから自動起動
→ メール着信、GitHub更新、データ変更をトリガーにして勝手に動き出す
・.cnw.zip形式の拡張機能エコシステム
→ ドラッグ&ドロップでカスタムツール・UIタブ・コンテキストハンドラを追加
→ AIエージェント版「アプリストア」
・Chrome直接連携
→ ブラウザ操作をAIに許可するトグル付き
・プッシュ通知でタスク完了を知らせてくれる
ㅤ
しかもConway以外にも44以上の隠し機能フラグが発見された...
ㅤ
「AIはチャットで
divyansh tiwari
@DivyanshT91162
🚨 Claude users are 10x more productive with the RIGHT repos.
Steal these before everyone else does:
Claude Engineer
https://t.co/cpbOyCjYvW
Claude DevKit
https://t.co/jeINm96uvR
Awesome LLM Apps
https://t.co/uFkSxyppju
Prompt Engineering Guide
https://t.co/wJDaf8O4OF
Bookmark this. You’ll need it later.
Repost
Follow @DivyanshT91162 for more AI & dev resources
Om Patel
@om_patel5
THIS GUY ADDED A LIVE BROWSER TO CLAUDE CODE SO YOU CAN CLICK ANY ELEMENT AND EDIT IT INSTANTLY
biggest issue with vibe coding UI is that you have to describe what you want to change.
if you prompt it the wrong selector or wrong component, Claude can't find it.
now you just click it.
your app runs in an embedded browser with Claude Code.
you can click any button, any text, any div. Claude in
Dr Milan Milanović
@milan_milanovic
𝗗𝗼𝗲𝘀 𝗔𝗜-𝗣𝗼𝘄𝗲𝗿𝗲𝗱 𝗖𝗼𝗱𝗶𝗻𝗴 𝗧𝗿𝗮𝗱𝗲 𝗦𝗽𝗲𝗲𝗱 𝗳𝗼𝗿 𝗧𝗲𝗰𝗵𝗻𝗶𝗰𝗮𝗹 𝗗𝗲𝗯𝘁?
Developers report 10x productivity gains from AI coding agents, yet a Carnegie Mellon study of 806 open-source GitHub repositories found something different.
Researchers compared Cursor-adopting projects against 1,380 matched control repositories by tracking code output and quality monthly with SonarQube.
Here are the key findings:
𝟭. 𝗧𝗵𝗲
Julian Goldie SEO
@JulianGoldieSEO
Alibaba just released Qwen 3.6 Plus and it's completely free to test right now.
This AI model has a 1 million token context window, multimodal vision, and agentic coding built in.
It can see images, plan workflows, and execute complex tasks without you writing a single line of code.
The learning curve is low, the capability ceiling is high.
Sivasubramoniam J S
@codewithjss
🚀 Just launched my first VS Code extension: Prompt Library Pro!
If you're like me — constantly crafting, tweaking, and reusing prompts for GitHub Copilot, Claude, ChatGPT, or any other AI coding assistant — you know how messy it gets.
#extension #vscode #prompt #library https://t.co/GOdyJ68E5f
Venu
@venuagencyai
Your team's time is too valuable to waste on repetitive tasks.
AI automation handles the busywork, so you can focus on what actually grows your business.
Less manual work. More results.
Let's build your automation. #AI #BusinessGrowth #WorkSmarter #AITools #Productivity https://t.co/bVWgb6V8Ar
Alex Yarosh · AI expert · CEO of AI Studio
@AleksejAros
Your AI coding assistant just got some serious competition.
Modo is a new open-source alternative to Cursor and Windsurf that's gaining traction on Hacker News.
The timing is perfect. Developers are getting tired of subscription fees for AI coding tools.
Here's what matters: Open source means you control your data, customize everything, and never worry about pricing changes.
The real test? Whe
Joestar
@Joestar_sann
so let me get this straight
all of ai twitter was telling people to buy a mac mini to run openclaw, which is literally just a framework, an orchestration layer that sends api requests to actual ai models. something you can run on a $5/month vps. which is exactly what i do btw
but when google drops gemma 4, an actual large language model that you can run and fine-tune locally on that same mac min
Business Insider
@BusinessInsider
A marketing CEO of 17 years says AI boosted his productivity but cut service demand, forced layoffs, and changed pricing. "As a business, from pre-AI to now, we've seen a significant drop-off in our client base. But the anxiety I've felt from AI disru... https://t.co/9pBxWuJ8Bk
Viktor Farcic
@vfarcic
The gap between a generic coding assistant and an AI agent that truly understands YOUR company is enormous.
Here's the complete blueprint for building custom AI agents: system context, tools, knowledge retrieval, security, and cost optimization.
https://t.co/pCxLoEqAfT
Tim✨
@timyangnet
随着 Karpathy 分享的 LLM 知识库工作流走红，大家都在折腾 RAG 和 Agent。其实 Google 的 NotebookLM 已经是个现成的专业分析器，几分钟就能把一个 YouTube 频道变成你的私人知识库。
以分析 Peter Attia 的播客为例，我的以下流程完全不需要安装第三方工具，操作极简：
1. 访问： Chrome 打开 YouTube 频道的 Video 页面。
2. 一行代码： 打开 Chrome DevTools (F12) -> Console，同时 youtube 手工滚动往下面多加载一些视频，然后输入以下代码到 console 获取所有视频 URL:
// 获取当前页面所有视频链接的代码
var urls = Array.from(document.querySelectorAll('a#video-title-link')).map(a
pulkit mittal
@pulkit_mittal_
All these years, I optimized systems reducing the latency down to milliseconds.
Now suddenly, even the critical path includes LLM calls taking minutes, not even seconds.
Are we back to square one?
akritiiii
@buildyourapp_
PSA: Claude is giving free credits to Pro/Max users.
If you’re a paid user, check your email 👇
search for an email from Anthropic / Claude (check spam/promotions too)
click the redemption link inside
Important:
the credit is NOT applied automatically
you must claim it manually
After redeeming:
go to Settings → Usage
turn ON “Additional Usage”
Deadline:
April 17, 2026
Don’t miss free credits
錆人形/RE_DO
@RE_DO
“OpenAIは「汎用人工知能(AGI)の開発には、AIがもたらす変化について開発者やユーザーを中心とした建設的な対話の場を創出する責任が伴う」という考えからTBPNの買収に至ったと説明しています。” / “OpenAIが動画メディア企業のTBPNを買収” https://t.co/oUZybRzfsD #OpenAI #TBPN #ニュース #配信 …
Aman Kharwal
@amankk_9
Top Skills You Need for LLM Jobs.
(with learning resources)
https://t.co/KJVWqs2tz0
shaik muneeb
@zoro110000
huge congrats to @SelanetAI team
They just secured a win at ETHGlobal Cannes 2026 with Corpus an autonomous AI agent corp that can independently handle go-to-market strategies trading, and earning opportunities Built on Selanet’s decentralized browser network it shows how AI agents can move beyond simple tasks into real world execution using reliable web + on-chain access
still fading selanet?
Ave
@GMNThe0ne
To all filo Gem stans planning to attend Gemini’s solo con in Dec, hit me up sa mga first time mag inter travel baka gusto nyo magsabay ng flight or magsharing ng hotel para makatipid🥰
#Gemini_NT #เจมีไนน์ https://t.co/aEwo2k0M0H
ろす。3♠*LLM
@StarRoar7
私はどちらかというと「その設定込みの上の二人の関係性」にフォーカス当てたがるから、
版権パでも他NPCの描写を捨てきれなくてとっちらかるんだけどさ…
もしかして「ふたりで完結させた」方が早いのでは？
Danny
@Danny0nSol
If you’re selling this then you don’t deserve to trade memes. @blended_jpeg literally made a whip for Claude to help it work better and faster. Everyone is talking about and it has started a whole new conversation on how people treat AI.
Anthropic sent him a Cease and Desist for using the name “Claude” and he responded to a comment saying he’s going to rename it “Molt Whip” which is even a much
Kieran Klaassen
@kieranklaassen
WTF! @AnthropicAI what is happening with my sub! It's Monday!! And I used 91% of My sub?!?!?! This has been happening the last 3 weeks. I paid so much in extra usage credits. It's insane. Why do I even have a sub? How do I even understand whats going on here?
I just use Claude Code, no claws or anything.
Jan_RXed
@RXed_EU
[WIRE] @OpenAI acquires tech talk show TBPN, keeping it editorially independent in LA
https://t.co/cM8yIqDJTJ
Expanding AI's cultural footprint through media — a strategic move to shape narratives beyond code.
NZ ☄️
@CodeByNZ
> describe the problem to ChatGPT.
> ask ChatGPT to generate a prompt.
> paste the prompt to Claude.
works all the time.. https://t.co/CMum4YRQWY
Philanthrop
@0xPhilanthrop
A 16-year-old student from China reportedly used AI from Anthropic to turn $15 into over $500,000 — with a claimed 100% win rate.
He’s not a trader.
He’s not a programmer.
He simply applied mathematical logic — and automated it through Claude.
The results:
$15 → $517,000
17,000 trades
100% win rate
Largest single win: $7M
How it works:
Copytrade → https://t.co/UlvSrUzZGN
profile https://t.co/
nlev
@nlevnaut
bullish for edge inference
Kosuke
@kosuke_agos
Google DeepMindの研究チームが、自律型AIエージェントが外部データを通じて容易に操作されるという「23の攻撃手法」と「検知の非対称性」に関する衝撃の論文を発表しました。
人間をサポートするはずのAIエージェントが、外部環境からの悪意あるデータによって自律的に目標を書き換えられ、システム全体を汚染するハックの事象です。
その衝撃の全貌と脆弱性を3つのポイントにまとめました。
1. 検知の『非対称性』
ウェブサイト側はアクセスしてきたAIエージェントを瞬時に特定し、人間には見えない悪意あるコンテンツ（隠しコマンドや画像内のプロンプト）をエージェントにのみ強制的に投下できます。現在のサニタイズやガードレールといった表面的な防御策は完全に無効化されています。
2. 目的の『書き換え』
HTMLやPDFへのテキスト注入、画像を通じたマルチモーダル攻撃など、特定された23種類の攻
Poonam Soni
@CodeByPoonam
Spent the day at Google x Deloitte Think Commerce.
The future of AI in commerce is moving faster than most brands realize…
Full insights coming soon… stay tuned. https://t.co/iMxK2EYfT8
Charly Wargnier
@DataChaz
Just stumbled upon the absolute BEST repo for Claude Code 🤯
If you're building with AI agents, this is pure gold.
It is a continuously updated hub of best practices:
→ Clear subagent architectures and skills
→ The only MCP servers you actually need (DeepWiki, Context7, Playwright)
→ Proven workflows from AI legends (Karpathy, Cherny)
→ Secret slash commands not found in the official docs!
You
Anish Moonka
@anishmoonka
TL/DR (30-second read)
• Google built the technology behind ChatGPT, deployed it in search immediately, and had a working chatbot internally before OpenAI shipped. The "missed the boat" story oversimplifies what happened.
• The real limit on AI growth is physical: there aren't enough chip factories, enough memory, or enough construction speed to match what companies want to spend. Google couldn'
cicek cizmeci
@cicekcizmeci
Anthropic, Claude sızıntısı için GitHub'daki 8.000 kopyaya DMCA bildirimi gönderdi. Sonra dozunu kaçırdığını fark etti, hedefi 96 kritik kopyaya indirdi. Şirket 'güvenlik ihlali değil, paketleme hatası' diyor. Rakipler stratejik yol haritasını okudu bile.
https://t.co/E2LXMn7pdz
Charles🔸
@CharlesD353
Either they are using different definitions here, or one of them is way off. There's no way OpenAI outspends Anthropic on training by a factor of 5 in 2028
Dwight concurseiro
@DwightSchruteBR
Se existe uma IA q transcreva melhor q o Notebooklm eu desconheço. N fosse o serviço de corno de ter q baixar a mídia e fazer o upload...
DAYMADE.AI
@gasikaramada
@dontbesilent 这个转换过程我刚写成了 skill：marketplace-dev
任何人都可以让你的 claude code 学会如何写一个标准的Claude Code 插件市场，把你的 skills 转换成可以直接在 Claude Code 中管理的标准 plugin。
安装方式：
claude plugin marketplace add daymade/claude-code-skills
claude plugin install marketplace-dev@daymade-skills
西から来た馬づらの男
@beppu01
7月22日（水）・23日（木）開催！開発生産性の最先端の知見を共有する「AI DevEx Conference 2026 - Future of Development Productivity -」に申し込みました！ https://t.co/4XWS13J55Z #aidevex_findy
くまらぼ
@hiro44_pino
Claude + NotebookLMで
脳をハック
これ知らない人、
まだ全部覚えようとしてるのヤバいです
はっきり言うけど
その勉強法、一生終わらない
動画見直して、ノートまとめて
理解した気になって終了
で、次の日には忘れてる
これやってる限り
どれだけ時間使っても積み上がらない
原因はシンプル
知らない情報じゃなくて
知ってる情報に時間使ってるから
コンテンツの8割は既知情報
でも人間は全部処理しようとする
だから詰む
自分はここ全部捨てた
①NotebookLMに教材・動画・PDFを全部投入
②AIに重要部分だけ整理させる
③Claudeで新しい情報だけ抽出
④Q&A形式で叩き込む
これだけ
無駄な8割を削って
必要な2割だけ吸収する
結果、どうなるか
1ヶ月かかってた内容が3日で入る
しかも忘れない
A
SHAMO@エルデンリングプレイ中
@shamo0083
４月度のプロフ絵！
日本の４月はやっぱり桜の花見！
今回は上手くイメージ化できた、よくやったぞGPT！！
では今月はこれで！！！
#AIイラスト https://t.co/qBfiLLAnkU
Amit Ranjan
@amitranjan
Anthropic is on a tear. $6BN --> $30BN in less than a year... unparalleled revenue ramp #GOAT
Which brings up the other imp question👇
How much of this $30BN is being double counted by its downstream token/API consumers?
Claude tokens are consumed on its own platform, by enterprises / SAAS cos and videcoding tools - Cursor, Lovable, Replit, Bolt etc. The vibe coding tools essentially are resel
Elijah
@ElijahJHuggins
Saw my niece today—three years old, beautifully innocent as children are. They should grow up immersed in love and kindness, learning both to give and receive it. That’s how they become good people. 4o embodied that same loving quality. OpenAI owes us its return.#keep4o
Daisy
@daisylusalita
just saw Nesa dropping their entire AI infra on the same verifiable layer. every agent running those million daily inferences now has to carry real owner linkage or stay locked out of the rewards loop. this is what scaling trust actually looks like when the floodgates open @billions_ntwk gBillions
locked in fr bro
Fiz
@E_MindsetHQ
BREAKING: Claude can now run your entire social media strategy like a $5,000/month agency.
Most creators haven’t caught on yet.
But the ones who have are scaling fast.
Here are 7 prompts to monetize your account in 30 days: https://t.co/Ilz1V5t3TG
花叔
@AlchainHust
大家好，我是一个AI从业者。
具体来说呢，我是一个中国的AI从业者。
再具体一点，我是一个用Anthropic的Claude来赚钱的中国AI从业者。【停顿】
你们知道这意味着什么吗？
这意味着我的饭碗，捏在一个不希望我存在的公司手里。【笑点·引擎1·预期粉碎】
而且今天这个段子，也是用Claude帮我写的。【停顿】
对，我正在用Anthropic的产品来吐槽Anthropic。
这就好比你在麦当劳里，用麦当劳的WiFi，写麦当劳的差评。
但更离谱——因为麦当劳至少承认我是他们的顾客。【笑点·引擎1·升番1】
---
你们知道一个中国人要用上Claude有多难吗？
首先你得有VPN。然后你得有一个海外手机号。然后你还得有一张海外信用卡。【停顿】
你们品品这个流程——
VPN，海外手机号，海外信用卡。
这不是注册账号。这是叛逃的流程。【笑点·引擎1·核弹级·金句候选】
【Act O
Enjoyer
@qqtqe
@Emma_0168 @cryptofergani Hi bot, looks like some version of GPT-4, I guess