Ken Fukuyama (@kenfdev) header
Ken Fukuyama (@kenfdev) profile picture

Software Engineer. CEO of Craftell Co., Ltd. Loves Continuous Improvement. I tweet both in Japanese and English. https://teapaw.org/t.co/XpCiiuAo4X

Osaka, Japan
Joined September 2012
Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)Photo by Ken Fukuyama (@kenfdev)
Ken Fukuyama retweeted
We've decided to sunset Arc on December 22, 2026. That's not an easy sentence to write. We spent years building Arc and celebrating each and every release along the way: we have so much pride and love for what we made alongside the Arc community. Our priority is to deliver best-in-class security for our members. Keeping a browser secure takes constant, dedicated work that is only increasing in the age of AI. Dia was built from the ground up to meet today’s security standards, so that's where our engineering and security teams will focus moving forward. We've worked to make switching to @diabrowser as easy as possible for Arc members. Here's how: · Your passwords, bookmarks, extensions and history come over to Dia seamlessly. · Spaces, Folders, and Tab Groups are in Dia and import over with your data intact. · Dia now surpasses Arc benchmarks in stability, speed, and security. · The browser features you use in Arc are free in Dia, and the AI is optional. Thank you for using Arc, for pushing us to make it better, and for building alongside us. The team that made Arc is the same team behind Dia, and some of our most ambitious work is just ahead. We'd love for you to be there with us when it lands.
2,262
965
2,697
8,151
3,382,896
Ken Fukuyama retweeted
Good afternoon from Vienna People of Pi ☀️ Some slightly different Sunday meditations today in which @mitsuhiko talks about some shared frustrations with agentic software engineering. Remember that you're not alone in finding it harder than it looks!
31
45
6
492
55,861
Ken Fukuyama retweeted
We also pushed out a new release so that the library can use the latest models. pi.dev/news/releases/0.87.1
7
8
255
18,995
Ken Fukuyama retweeted
We tested Jev against LLM judges on accuracy, repeatability, latency, and cost to see whether System One models could offer a new approach to agent evaluation.
213
380
136
3,000
639,194
Ken Fukuyama retweeted
In just over an hour, I'll be live with @unclebobmartin We'll be figuring out if software fundamentals still matter in the age of AI. Don't miss it: youtube.com/live/zcLPGC-tvgk…
34
59
20
693
141,141
Ken Fukuyama retweeted
書きました。AIで開発量増えてPRレビューがボトルネックになる問題の対策。 58% の Pull Request を AI が承認するようになった|qsona zenn.dev/she_techblog/articl… #zenn
30
7
280
47,914
Ken Fukuyama retweeted
【再告知】 TAKT が 8/9 に Product Hunt にローンチされます! 当日、みなさんのユースケースなどを Product Hunt に投稿いただけると嬉しいです!! 大変恐縮ですが応援していただきたく、アカウント作成等含めてご準備よろしくお願いします!
TAKT を 8/9(日) 16時に Product Hunt でローンチします🚀 日本で育ててもらった TAKT を世界に向けて発表する日です ページで Follow いただけると心強いです 応援よろしくお願いします! producthunt.com/products/tak…
7
2
26
5,170
Replying to @DmytroKrasun
Here are a few repositories that you can look at. As for customer satisfaction, I think you’re going to have to wait a couple of years for all the AI code to actually get out to customers who can then evaluate them. GitHub.com/unclebob swarm-forge, crap4java, Acceptance-Pipeline-Specification, empire-2025, arch-view, crap4clj, AIR-J, dry4clj, clj-mutate, mutate4java, Pharaoh, dry4go, scrap, mutate4go, gospringies, crap4go, experiment-htw-go-swarm, dry4java, dependency-checker, sf-orbit-simulator, speclj-structure-check, htw-6-clj-vid, deintroverter4clj, Pharaoh-js, htw-clj-six-pack, experiment-htw-clj-swarm.
4
6
3
136
14,279
Ken Fukuyama retweeted
業務委託で手伝ってくれる人を探しています。 ※ DELTAの話でもSpeech Linkの話でもないです 興味ある人DMください ▼【開発内容】 ・生成AIを活用したWebアプリの設計・開発 ▼【求めるスキル・経験】 ・TypeScriptおよびNext.jsでの開発経験 ・AWSを用いた開発経験 ・アプリケーション開発の実務経験がある方 ・抽象的な要件からやるべきことに分解し実装まで自走できる方 ▼【その他】 ・週3日月に100時間程度の稼働が可能な方 ・地方在住の方も大歓迎!
3
18
1
85
21,823
Ken Fukuyama retweeted
I had early access to 5.6/Sol for ~month. Sol is my default. It is faster, plans/judges just as good as Fable, and I think produces better overall work. I’ll reach for Fable still for highly targeted debug or performance work with clear reward functions. A cheeky way I describe Sol vs Fable to my friends is that Sol is a charismatic, efficient, talented coworker you’re jealous of. Fable is a genius recluse that is brilliant at its fixations but doesn’t go out, doesn’t date, and you don’t want to hang out with them much lol. Fable is undefeated at highly targeted debug/security/performance goals. It’s a sight to behold and I was never able to get Sol to push as hard in this category. I’ll keep using it for this. Sol is better or comparable at everything else, in my experience. Give it a shot, it’s hard to describe but it’s just more enjoyable to work with. (Disclaimer I have no financial ties to either lab, wasn’t paid for any of this.)
225
301
102
6,757
1,216,528
Ken Fukuyama retweeted
Introducing cmux Finder cmux now has a file explorer that can preview videos, images, PDFs, and markdown files I like using it to preview demo videos and screenshots that codex films for me Can be toggled with Cmd+Shift+P -> Enable/Disable Cmd-click File Previews Requires v0.64+
18
13
2
228
21,229
Ken Fukuyama retweeted
I strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
500
1,885
394
15,213
1,625,569
Ken Fukuyama retweeted
Introducing cmux Vault: cmux now has a right sidebar with a vault pane where you can see every single Codex, Claude Code, OpenCode, and Pi session You can search through all sessions with full-text search, and drag them directly into your workspace Out in v0.64+
48
27
7
373
32,522
Ken Fukuyama retweeted
In case you've wondered how much usage you'll get on these $20-$200 credits. Just ran my Token Optimizer to calculate my CC usage in the last 30 days, with a breakdown of -p sessions in API costs. Yes, the subs are heavily subsidized. Yes, Anthropic can change this whenever they want. But please, just be straight with your user base. This is a downgrade. A severe one. In every sense of the word. Trying to spin it as a win is poor marketing.
This means you'll be able to build on Claude without needing an API key! Everything running on the Agent SDK will instead draw from a new monthly credit, separate from your subscription limits.
8
6
4
94
15,355
Ken Fukuyama retweeted
it's official. Anthropic pulled the plug on ALL programmatic use of claude subscription i've found myself increasingly bullish about OpenAI a few key reasons - 1. Anthropic's only lead was on coding, and gpt 5.5 has flipped that already. coding with gpt 5.5 fast mode is an experience nothing else can match right now 2. OpenAI has voice, image, video - when building my AI tutor app, it became very clear OpenAI was the ONLY viable platform choice 3. it took an Anthropic to see the "Open"-ness of OpenAI. Anthropic is destroying its developer ecosystem with changes like this which tells their subscribers they can't use the best apps out there on Anthropic's models - they can only use Anthropic's apps. Codex on the other hand remained an open platform for developers to build on not just trying to dunk on Anthropic - i genuinely hope they change their strategy and recognize that the world needs them to focus on making better, faster, and cheaper models, not locking everyone into their apps
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
107
80
8
1,469
612,626
Ken Fukuyama retweeted
Ghostty is leaving GitHub. I'm GitHub user 1299, joined Feb 2008. I've visited GitHub almost every single day for over 18 years. It's never been a question for me where I'd put my projects: always GitHub. I'm super sad to say this, but its time to go. mitchellh.com/writing/ghostt…
541
1,565
510
16,474
2,935,246
Ken Fukuyama retweeted
Replying to @divyaanshgoel
Yeah, I have no illusions that design will change or that these tools will not be useful. Small changes I'm already doing directly with the Linear cloud agent to the codebase and can run preview url to see the change. What I am still unsure about is how important a tight connection to production code is overall in the design process itself, especially once agents can implement things quickly from the right direction and constraints. A lot of my design process is still broad, exploratory work. I use it to find the shape of the product, communicate the idea, and create something we can react to together. Then we usually have to redo parts of it anyway once we start prototyping and understand the real tradeoffs, or changes we need to make to the broader product. I do not rely much on Figma as a system of components, auto layout, or dev mode. For me, design is mainly a tool for exploration and communication, not an overly considered handoff artifact. To me what would more useful for someone to build is a a way for designers, or anyone, to have lightweight frontend environment where you can quickly modify the UI, fake features or API calls, and work without taking on the full complexity of the production codebase. Then, once the prototype feels right, you could hand that to an agent+engineer and have it implement the real thing in the actual codebase with the correct way and right tradeoffs.
3
2
2
56
3,193
Morning bathrobe rant: Rule files.
109
143
79
2,585
710,818