David Bennett

先月出した本について、色な方から聞かれてますので、もう一度リンクを載せさせて頂きます😅🙏🏻　因みに、ビジネスの本ではありません。面白い日本語のトリビアの本です。まさに、僕にとって気になる日本語です。Thank you for asking about my new book! amazon.co.jp/dp/4046046058/

Small deal for real programmers out there. Big deal for me! Contributed to @openclaw and got my PR accepted and merged. 🥳 #85724 FTW! Thank you @steipete ❤️

I really like our website. It has … a vibe? Check it out. ai& (@aiand_) aiand.com

@elliotarledge Especially with @karpathy getting a new job and likely taking a break 😭

dwarkesh getting into education is just the best

Our CEO @DavidBennett__ sat down with @jrobottom from @AMD to discuss what it actually takes to execute AI at scale — agent-driven workflows, Japan's adoption gap, and the compute economics behind scalable AI. Watch the full conversation: bit.ly/4dzPlZ7

A huge thank you to our friends at @AMD for participating in a discussion about what it will take to accelerate AI adoption here in Japan. Thank you to AMD Japan’s Country Manager Jon Robottom for his time and his insight. @aiand_ (日本語字幕付き) youtube.com/watch?v=Rq4Yxc…

Our list of partners continues to grow. I am also pleased to mention that every partnership announced by ai& (@aiand_) is backed by announced deals (or deals in the works!) ai& and Ryosan Ryoyo Expand Enterprise AI Adoption itbusinesstoday.com/tech/ai/ai-and… (TY: @itbusinesstoday)

@BenjaminBadejo @BenjaminBadejo getting an openclaw oauth error even though that’s how I’m Using openclaw Some OpenAI vs OpenAI codex weirdness? How to configure?

It’s here: VoiceClaw Realtime is now available on the App Store for iOS and watchOS, enabling realtime conversation continuously via GPT-Realtime-2 plus GPT-5.5 Instant or GPT 5.5 via OpenClaw. You can operate OpenClaw and your computer, including via cellular connection, from your phone and even from your Apple Watch (whether your phone is nearby or not), or talk to the models directly without OpenClaw. Just scan the in-app QR code reader to pair / sync settings to your iPhone once, and you’re set. apps.apple.com/us/app/voicecl… On iOS, VoiceClaw can keep running in the background on iOS (with an indicator light), switch modes upon request by voice, directly search the web, read images and analyze video from your clipboard or photo album, draft texts and emails, and use Maps and Safari. You can tell VoiceClaw to mute the mic by using your voice, and turn it back on by saying “mic on” (see Accessibility in the app). You can change model reasoning levels, change output from handset/headphones to speaker or use Apple Watch speakers, and modify noise reduction and voice activity detection sensitivity. A live transcript is available if you toggle it on, including on watchOS. The next update (already finished, available by the weekend) has built-in ChatGPT sign-on, so you won’t need the macOS setup/Companion app to import your ChatGPT subscription to the iOS or watchOS app unless you want to connect to OpenClaw. It also lets you select OpenClaw-mode models besides GPT-5.5 (incl. GPT-5.4 and GPT-5.3-Codex); non-OpenAI models coming later. The app collects no data, and the setup/Companion app is on GitHub. It will be available by searching “VoiceClaw Realtime” in the App Store soon, but it’s available at the link above now.

This.

Great to finally meet @DavidBennett__ in person! Excited for what's ahead at @aiand_ 🙏

Congratulations to the Overwatch AI team as they raise $1.5M for airline operations platform phocuswire.com/news/finance/o… via @PhocusWire Funding from @united amongst others. This a brilliant founding team with industry specific exercise and I am happy to be an investor.

@ThePrimeagen This doesn’t bode well for the U.S. The kids are booing it here and celebrating it everywhere else in the world. What a crazy dichotomy. Kind of bummed to see this…

dude... this is brutal

@FelixCLC_ TMUX🙌

WE HAVE TMUX LEARN TO USE TMUX

Japan is hungry to accelerate AI token usage! ai& (@aiand_) has data centers in Japan producing and delivering fast, affordable and secure tokens as part of our ai& inference service. Proud to be partnering with @tomorrownet_co to increase AI adoption! prtimes.jp/main/html/rd/p…

The incredibly smart and talented @Weyaxi in the house at @aiand_ ai&! Glad to see you while we were both in Japan! 🇯🇵

@Ken_Doi おっしぃ！

昨日会社の人と飲みに行って@DavidBennett__ の話になり、TenstorrentのJimの事を、ジム・キャリーと言ってしまった(^^)

Thank you to @credosemi’s CEO Bill Brennan for visiting the @aiand_ ai&’s office in Yokohama, Japan. 🇯🇵

@sudoingX How do you set this up? Still can’t picture the architecture / flow… 😅

@DavidBennett__ not nagging, it's the architecture. i use private git, not github, the repo of code and docs is the memory itself. every new agent pulls it, gets full context, commits back. a merge agent handles the merging. git isn't a step you enforce, it's how the agents share state.

anyone thinking about, learning, or already working with agentic systems, you should know this. the first few steps of your setup matter more than any model or framework you pick later. get them right and you never lose your flow. the foundation nobody posts about: > 1. tailscale. a private mesh network across every machine you own. laptop, desktop, rented node, all on one secure tailnet, reachable from anywhere. nothing else works well until this does. > 2. termius, over that tailnet. one SSH client that reaches every node, phone included. you are never away from your stack. > 3. tmux. persistent sessions. disconnect, close the laptop, come back, every session exactly where you left it. agentic work runs long, your terminal has to survive that. > 4. a private git repo. the one i am most glad i found. it is the memory layer across all my agents, they pull, they work, they merge back, the codebase stays alive between sessions. context that would die in a chat window lives in the repo instead. > 5. script everything from day one. ssh aliases for every node, setup scripts, the boring boilerplate automated. if you will do a thing more than twice, it is a script. everything past these five is decorative. know these cold. and the habit that ties it together: ask the AI itself. for the config, for the error, for any of it, let the agent do the lifting, then double check what it hands you. lock the five, build the habit, and you make it. skip it, anon, and you ngmi.

@steipete I’m not freaking out. I think we should all be spending more tokens. If we are, that means we are creating.

People freaking out over my AI spend. What nobody sees: Part of what excites me so much about working on OpenClaw is that I'm trying to answer the question: How would we build software in the future if tokens don't matter? We constant run ~100 codex in the cloud, reviewing every PR, every issue. If a fix on main lands, @clawsweeper will eventually find that 6 month old issue and close it with an exact reference. We run codex on every commit to review for security issues (as it's far too easy to miss). We run codex to de-duplicate issues and find clusters and send reports for the most pressing issues. We have agents that can recreate complex setups, spin up ephemeral crabbox.sh machines, log into e.g. Telegram, make a video and post before/after fix on the PR. There's codex that watch new issues and - if it fits our documented vision well, automatically create a PR of it. (that then another codex reviews) We have codex running that scans comments for spam and blocks people. We have codex instances running that verify performance benchmarks and report regressions into Discord. We have agents that listen on our meetings and proactively start work, e.g. create PRs when we discuss new features while we discuss them. We build clawpatch.ai to split all our projects into functional units to review and find bugs and regresssions. We do the same split for security with Vercel's deepsec and Codex Security to find regressions and vulnerabilities. All that automation allows us to run this project extremely lean.

The facade on our office got updated! I’m not going to lie… it’s 🔥 @aiand_ is cookin’!