Vivek Amilkanthawar

ChatGPTで止まってる人、Claudeを試してほしい。 特にこれは Claude の方が圧倒的に効く： ・契約書レビュー（不利な条項の検出） ・50ページ資料を3分で「使える要約」に ・議事録まとめ ・英文メール作成 実務で効いた50プロンプト＋返答例を 45ページのPDFにまとめました。¥2,500 coconala.com/contents_marke…

Almost everything humans do (including the things they sincerely tell themselves they're doing for other reasons) is fundamentally about status. Not money, not love, not "happiness," not "doing good." Status. And the part most people refuse to see is that we evolved to hide this from ourselves, because anyone who openly admits "I'm doing this for status" loses the status.

If your return doesn't beat inflation, you're getting poorer by holding it.

@elonmusk @grok is it true? amzn.to/4ww0CAW

Read this book and give it to all your friends. Survival of civilization depends on it!

@Rothmus I wonder what they are doing now?

130k tech layoffs year-to-date. Insane.

今日 使った 判定プロンプトを 含む 50 個を まとめた 「Claude仕事術ガイド」こちら: coconala.com/contents_marke…

請求書 PDF を フォルダに 入れるだけで、 勘定科目を Claude が 判定して、 freee に 取引として 自動登録する エージェントを 作ってみました。 サンプル 10 枚の 一括処理で、約 3 分。 作り方を 7 分の動画に まとめています。

めっちゃ同意。 複雑なタスクになるほどMarkdownだと表現力が足りなくて、HTMLの方が情報密度・インタラクション・共有性すべてで勝つ。 AIの出力を「読む」じゃなくて「操作する」体験に変えられるのがデカい。 x.com/trq212/status/…

this maps well to agent → human curious how you think about agent → agent or agent → human → agent though. in multi-hop flows, all the markup + css feels like tokens the next agent has to ignore. your slider/knob → json export point actually hints at it, the interactive layer is for humans, the handoff is structured data

今日 使った構造化プロンプトを含む 50 個を収録した「Claude仕事術ガイド」 : coconala.com/contents_marke…

1時間の会議の議事録、毎週6時間取られていませんか。 Whisper で文字起こし → Claude で要点整理 → Notion に自動書き込み。 会議の後処理を AI に外注する、議事録要約エージェントの作り方を まとめました。

@xai @trq212 @AnthropicAI Thank you! But I wonder why Isn’t it killing grok by appreciating competitor

SpaceXAI will provide @AnthropicAI with access to Colossus 1, one of the world’s largest and fastest-deployed AI supercomputers, to provide additional capacity for Claude → x.ai/news/anthropic…

@claudeai @SpaceX Thank you! Finally i can do more of “make no mistake”

We’ve agreed to a partnership with @SpaceX that will substantially increase our compute capacity. This, along with our other recent compute deals, means that we’ve been able to increase our usage limits for Claude Code and the Claude API.

動画で使ったエージェント構築プロンプト50個は「Claude仕事術ガイド」にまとめました: coconala.com/contents_marke…

AIエージェントを5分で作る、その全工程を動画にしました。 ✓ プロジェクト作成 ─ 0:30 ✓ Gmail API 接続 ─ 2:00 ✓ AI ロジック生成 ──3:30 ✓ 文体学習 ─ 4:30 ✓ 実行・完成 ─ 5:00 コードは一行も書きません。Claude Code に任せるだけ。

@alex_whedon So RAG is dead?

Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.

@steipete Looks like Claude and you were busy

Me and codex were busy. 🔊 sonoscli.sh — Sonos 🗃️ wacli.sh — WhatsApp 🪶 birdclaw.sh — X archive 🧰 gitcrawl.sh — GitHub archive 🛰️ discrawl.sh — Discord archive 🎧 spogo.sh — Spotify 💬 imsg.sh — iMessage 🧳 mcporter.sh — MCP to CLI 🗣️ sag.sh — ElevenLabs voice 🧿 askoracle.sh — second opinion Upgrading the 🦞 OpenClaw army.

YouTubeにも上げています(同じ動画): youtu.be/Rb9GxpNhRNQ

「AIで仕事がなくなる」って本当か、5つの職業をAIに任せて検証してみた。 プログラマーの置き換え可能率は80%。 ジュニアレベルの仕事は本気で危うい。

Everyone is saying AI agents will kill junior banking jobs. I don't think so. The juniors aren't gone. The bar just moved up. The analyst who used to build one pitchbook now reviews ten. The compliance person who screened 50 KYC files now checks the agent screening 5,000. Same job. Higher floor. What actually got killed is the version of the role where you only execute. If that's how you describe your work, the description got rewritten.