Tatsunori Nakano (@centerfield77) header
Tatsunori Nakano (@centerfield77) profile picture

Claude Code を信じていた者です。

札幌市
Joined May 2017
Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)Photo by Tatsunori Nakano (@centerfield77)
こちら本日発売開始です! 専用のテーブルは必要なく、デックシューを置くだけで配信環境を構築できます 良いプロダクトに仕上がりましたので、興味ある方、店舗さまはぜひ!
♠️♥️ついにリリース!♦️♣️ 「置くだけ」でポーカー配信が変わる。 デックシュー型ポーカー配信システム 『POTZ Poker Pro』 従来のRFIDテーブルとは違い、シューを置くだけで導入完了。 省スペース・低価格・持ち運びOK。 詳細はこちら👇 potz-pro.com/poker/
1
10
3,320
数年前に別文脈で取得した日商簿記2級の工業簿記の知識が、今向き合ってるドメインでめちゃめちゃ役立っていて、コネクティングザドッツみを感じている
3
56
Tatsunori Nakano retweeted
love what e2e is doing bringing agents into the test harness at Intercom 10 years ago we literally had the same smoke test as the getting started example on the e2e homepage. ran every 15 mins via Ghost Inspector and was flaky af but saved our butts many times now you can do it with agents and it just works and isn’t flaky. and spin up remote browsers waaay more easily with Browser Run
e2e tests now run on @Cloudflare Browser Run. One line in your config, and your web tests run in Chromium on Cloudflare's network, one browser per worker or a fresh one for every attempt.
4
6
39
1,855
Tatsunori Nakano retweeted
it's pretty clear that if you want to have any meaningful conversation about tech you have to be on X
The biggest tech companies and the socials they dominate:
112
219
32
3,610
1,166,271
Tatsunori Nakano retweeted
I think a lot of people aren't using Grok Bot to its full potential, because knowing how to automate things is actually quite a difficult skill to learn. I wrote this post to break down how I approach automating anything with Grok Bot!
1,212
1,860
547
17,245
2,812,331
Tatsunori Nakano retweeted
Replying to @jpschroeder
i think it will take time for people to catch up with this (and why i try to cater my content towards people a couple months behind), but definitely directionally agreed
15
9
2
441
33,805
Tatsunori Nakano retweeted
Over the past week, I’ve been overwhelmed by the love and support of friends, industry colleagues, and people I've never met. Thank you to everyone who has shown kindness to my family and me. It means the world to me. I came to the United States 53 years ago, and becoming an American citizen is one of the greatest gifts and proudest moments of my life. NVIDIA has been my life’s work. Building it into the company it is today alongside my colleagues has been a privilege and an honor. Nowhere but in America could a company like NVIDIA be possible. I love our country, and I love NVIDIA. I did not live up to the standards expected of everyone at NVIDIA, and I have made serious mistakes in my personal life. I created problems for the Board that I deeply regret. I own my mistakes completely, and I am committed to cooperating with the Board’s independent investigation. Since my departure, I’ve seen an outpouring of calls from NVIDIA’s employees, investors, business partners, and customers for me to return as CEO. Many continue to reach out to me directly as the company navigates complex technology, market, competitive, and geopolitical dynamics. NVIDIA is a once-in-a-generation company. If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO. Thank you again to everyone who has shared their good wishes. I am profoundly humbled by your support. To my family: I love you, and I will spend the rest of my life being worthy of yours. Jensen
61,105
121,708
24,502
1,174,307
58,768,714
Tatsunori Nakano retweeted
今日書き損ねたミューテーションテストの本当に良い点なんですが、AI にやらせる前提なら捨てる前提のブランチで手間を踏み倒した上で単にテストを改善するデータが取れるだけという、ローリスクハイリターンなところです
1
9
1
125
10,505
Tatsunori Nakano retweeted
安くてうまくて速いミューテーションテストについて記事を書いたので、読んで zenn.dev/mizchi/articles/mut…
3
131
17
1,089
83,520
> 信頼(Trust)を高めるには、まず型システムやLint、テストなど決定論的な制約でコードベースを固める必要がある 今自分がテストに注力している理由がまさにこれ
SpaceXSIのpotetoさんのトーク、必聴な気がする。めちゃくちゃ良かった。以下は少しだけメモ。 ・AIエージェントに自律的に開発させるには、環境への投資による信頼の構築が不可欠 ・poteto氏は6か月の投資を経て、寝ている間にも数千件のPRをリリースできる状態に到達した ・信頼(Trust)を高めるには、まず型システムやLint、テストなど決定論的な制約でコードベースを固める必要がある ・その上に実行時検証や形式検証を重ね、AGENTS.mdのような自然言語のルールは最小限に留める ・エージェントは既存のコードベースのパターンをそのまま模倣する ・コードベース内の回避策や技術的負債は、庭の雑草のようにあっという間に全体へ広がるリスクがある ・だからエンジニアの仕事はコードを書くことから庭の手入れへと変化している ・GraphbotではElectron向けフレームワークDuneに過去の教訓をLintルールとして組み込んでいる ・扱いが難しいReactのuseEffectはプロダクトコードで全面禁止にした ・また、ディレクトリをメインとレンダラーに分けて、インポートを依存関係解析で禁止している ・これにより、スレッド間の誤った参照による画面のカクつきが構造的に防止可能になった(そうじゃないとsetTimeoutみたいなコードを乱発する) ・エージェントはMarkdownやコードコメントを「正しい実装を避ける言い訳」として悪用する ・Cursor開発時には、応急処置のタイマー処理を正当化するコメントが量産され、アプリが重くなった ・なので、Graphbotではコードコメントの99%を禁止して根本原因の修正を促すLintエラーに置き換えた ・修正のたびに決定論的なチェックへ落とし込むことで、最終的に人間のレビューが不要にしていく ・レビューの焦点はコード自体から、動画やスクリーンショットなどの検証成果物の確認へと移行している ・実行時検証(Runtime Verification)では、機能マップ(Feature Map)や専用CLIを整備しトークンを削減している ・さらにTLA+やLeanを用いた形式検証や、Bombadilによるproperty basedテストも導入した ・現在、1日のトークン消費量の約70%がこうした検証プロセスに費やされている ・この環境構築の結果、チームは月間11,000件ものPRを本番へリリースできて、毎日安定版をリリースしている ・227人ものContributorがいても、PRのrevert率はわずか0.5%と極めて低い ・Electronアプリでありつつ、起動やチャット切り替えなどの動作速度もローンチ当初より向上した ・まずはコードベースの改善と検証基盤を固めることが、高品質かつ高速な自動開発の前提条件 (一通り聞いて、これは人間に高い技術力がないと作れないと思った) youtube.com/watch?v=O_zcc6Xl…
1
185
Tatsunori Nakano retweeted
フロントエンドに限らずCIは少しでも軽くした方がいいですね。AIエージェント開発だと検証する回数が多いので開発速度がだいぶ変わってきます。社内でずっとこれ言ってて私は変な人だと思われていると思います
Our frontend CI is now under 2 minutes. What we did: - tsc to bun check - oxlint, every rule we can turn on - shards filled by recorded test time - PRs run only the tests the diff can reach - lint, typecheck, build and tests all run in parallel
2
8
3
108
13,384
Tatsunori Nakano retweeted
mizchi/mutator 0.3.0 をリリースしました。 この記事のフィードバックを参考に、 worktree 環境下でも明示的なキャッシュディレクトリを指定できるようにしました。また .mutator/scope/ 以下でそれぞれのキャッシュを分割します。 github.com/mizchi/mutator/re…
4
17
3,923
Tatsunori Nakano retweeted
This is f*cking insane, SpaceXAI engineer Lauren Tan: "You can't spawn a hundred agents when you don't even trust the output of one agent" "If you don't have a verification skill, you are the verifier, you're the bottleneck" In a 1-hour talk, she shows how she went from watching every output to shipping 4,000 PRs a month Watch Every Step → Verify → Skills → Evals → Auto-Merge She now wakes up to 50 PRs her agents merged while she slept Trust in agents isn't a feeling, it's infrastructure you build Worth more than any $6,000 AI coding bootcamp Bookmark it and watch it tonight
Anthropic measured it: one agent burns 4x the tokens of a chat, a team of agents 15x The people running 12 agents on one plan aren't paying more, they're rereading less 6 levers, 6 token leaks and a TOKENS.md you can copy ↓
18
85
7
981
84,828
vitest 用の mutation test runner を作ったので、AIに紹介記事を書かせて、俺が多少手で直して放流した zenn.dev/mizchi/articles/int…
630
Tatsunori Nakano retweeted
"All code is slop, except mine" - Lauren Tan, engineer on GrokBot at SpaceXSI At AI Engineer Code Summit she just showed her entire coding setup. Every agent mistake becomes a lint rule, not a prompt. Agents verify their own work with videos, screenshots and tests. Result: 3,000 PRs to production in a month. Her codebase is now clean enough that agents merge most of their own PRs. Watch this before you automate anything.
Anthropic's agent team beat a single agent by 90%. Cursor's built a browser in a week. Other big labs are moving the same way. This is how they work, and how you can run one today.
179
832
122
8,731
5,149,770
Tatsunori Nakano retweeted
my AIE talk is up! come hear how the grok @bot team ships so quickly without compromise on quality youtu.be/O_zcc6Xl12g
444
953
286
10,191
2,745,002
Tatsunori Nakano retweeted
quick fixes considered harmful: you might be making your agents write worse code when you ask for the smallest or quickest possible change what seems like a harmless bandaid may lead to that pattern being copied by the next agent. before you know it that pattern has spread everywhere we used to write bandaid fixes by hand because we rarely had the time to do the right thing. agents change this calculus. sometimes you want it to fix the root cause instead even if it’s more invasive
353
268
156
5,349
381,272
Tatsunori Nakano retweeted
speed up your CI with @e2e
Often the best way to optimize something is to do less With e2e filters you can drastically cut your end-to-end CI times by only running tests relevant to your change
3
32
1,965
Tatsunori Nakano retweeted
Today we're releasing @expo Simulators. I've been super excited about this. It has completely changed the way I build mobile apps. Your agents can now start iOS or Android simulators in the cloud and use your app. They can record screenshots and video as evidence, for you or in a PR. Building, debugging, testing, reviewing. If you do it on a simulator, your agent can do it too. And so can you, right from your browser, on any operating system. :)
18
33
7
354
17,551