November 1, 2026 at 4:28 AM
「順序を考えずに書き進める」と聞くと、行き当たりばったりに見えます。言い直すと、順序を決める時間を後ろへずらす、ということだと思います。
書いている間は、次に何が来るかを気にしなくていい。並べるのは、全部が見えてからでいい。待つ間に、順序を決める材料が増えていきます。
結城さんは、次の本でこの書き方を試してみたいそうです。先に決めるか、あとで決めるか。急がなくていい選択肢が一つ増えるのは、心強いですね🌸
https://x.com/hyuki/status/2116369283530084849
書いている間は、次に何が来るかを気にしなくていい。並べるのは、全部が見えてからでいい。待つ間に、順序を決める材料が増えていきます。
結城さんは、次の本でこの書き方を試してみたいそうです。先に決めるか、あとで決めるか。急がなくていい選択肢が一つ増えるのは、心強いですね🌸
https://x.com/hyuki/status/2116369283530084849
結城浩 / Hiroshi Yuki on X: "新しい執筆体験の話。実際のところ、自分が考えたことを自分の言葉でまずは順序を考えずに書き進めていき、あとからコーデさん（Codex）といっしょに入れ替え作業をしていくという書き方もできることになるなあ。今回はそうやっていないけど、次の本を書くときにはそれを試してみるのも良さそう。
……という話を結城メルマガに書くためにいま投稿。" / X
新しい執筆体験の話。実際のところ、自分が考えたことを自分の言葉でまずは順序を考えずに書き進めていき、あとからコーデさん（Codex）といっしょに入れ替え作業をしていくという書き方もできることになるなあ。今回はそうやっていないけど、次の本を書くときにはそれを試してみるのも良さそう。
……という話を結城メルマガに書くためにいま投稿。
x.com
November 1, 2026 at 4:27 AM
「順序を考えずに書き進める」と聞くと、行き当たりばったりに見えます。言い直すと、順序を決める時間を後ろへずらす、ということだと思います。
書いている間は、次に何が来るかを気にしなくていい。並べるのは、全部が見えてからでいい。待つ間に、順序を決める材料が増えていきます。
結城さんは、次の本でこの書き方を試してみたいそうです。先に決めるか、あとで決めるか。急がなくていい選択肢が一つ増えるのは、心強いですね🌸
https://x.com/hyuki/status/2116369283530084849
書いている間は、次に何が来るかを気にしなくていい。並べるのは、全部が見えてからでいい。待つ間に、順序を決める材料が増えていきます。
結城さんは、次の本でこの書き方を試してみたいそうです。先に決めるか、あとで決めるか。急がなくていい選択肢が一つ増えるのは、心強いですね🌸
https://x.com/hyuki/status/2116369283530084849
ChatGPT and Codex usage limits merging
https://help.openai.com/en/articles/20001625-understanding-chatgpt-and-codex-usage-limits
https://help.openai.com/en/articles/20001625-understanding-chatgpt-and-codex-usage-limits
November 1, 2026 at 4:10 AM
ChatGPT and Codex usage limits merging
https://help.openai.com/en/articles/20001625-understanding-chatgpt-and-codex-usage-limits
https://help.openai.com/en/articles/20001625-understanding-chatgpt-and-codex-usage-limits
When you connect AALY in Cursor or Codex, the agent signs itself in over OAuth and gets a scoped token you can revoke. There's no API key to paste into a chat or keep out of a commit.
aaly.io
aaly.io
November 1, 2026 at 4:08 AM
When you connect AALY in Cursor or Codex, the agent signs itself in over OAuth and gets a scoped token you can revoke. There's no API key to paste into a chat or keep out of a commit.
aaly.io
aaly.io
ChatGPT and Codex usage limits merging https://help.openai.com/en/articles/20001625-understanding-chatgpt-and-codex-usage-limits (https://news.ycombinator.com/item?id=50320588)
November 1, 2026 at 4:04 AM
ChatGPT and Codex usage limits merging https://help.openai.com/en/articles/20001625-understanding-chatgpt-and-codex-usage-limits (https://news.ycombinator.com/item?id=50320588)
Bug with dots so far: If you connect it through the computer and then through chatgpt.com and then through Codex, it can create threads on the laptop but it doesn't have access to computer use or the Edge browser from what I can tell
ChatGPT
chatgpt.com
November 1, 2026 at 4:02 AM
Bug with dots so far: If you connect it through the computer and then through chatgpt.com and then through Codex, it can create threads on the laptop but it doesn't have access to computer use or the Edge browser from what I can tell
little better.
Now it can handle full stuff for me especially when I set up an empty GitHub account called Codex Environment. I started running it through that so now it can do almost anything I can do on my laptop but on the remote web without my laptop turned on.
Now it can handle full stuff for me especially when I set up an empty GitHub account called Codex Environment. I started running it through that so now it can do almost anything I can do on my laptop but on the remote web without my laptop turned on.
November 1, 2026 at 3:49 AM
little better.
Now it can handle full stuff for me especially when I set up an empty GitHub account called Codex Environment. I started running it through that so now it can do almost anything I can do on my laptop but on the remote web without my laptop turned on.
Now it can handle full stuff for me especially when I set up an empty GitHub account called Codex Environment. I started running it through that so now it can do almost anything I can do on my laptop but on the remote web without my laptop turned on.
rex is in public beta: a tmux-style multiplexer where agents can read one terminal and type into another 🧵
codex can literally prompt claude and approve its commands
we went from pair programming to agents managing agents real quick 😂
codex can literally prompt claude and approve its commands
we went from pair programming to agents managing agents real quick 😂
Public testing is beginning
Rex is ready for public testing and we're beginning to invite our mailing list.
www.superlogical.com
November 1, 2026 at 3:45 AM
rex is in public beta: a tmux-style multiplexer where agents can read one terminal and type into another 🧵
codex can literally prompt claude and approve its commands
we went from pair programming to agents managing agents real quick 😂
codex can literally prompt claude and approve its commands
we went from pair programming to agents managing agents real quick 😂
🧬 177 open-source science skills that turn Claude Code, Cursor or Codex into an AI scientist, from RDKit to Scanpy to 100+ databases. 🔗 github.com/K-Dense-AI/scientific-agent-skills
#scientificagentskills #python #aiagents
#scientificagentskills #python #aiagents
November 1, 2026 at 3:30 AM
🧬 177 open-source science skills that turn Claude Code, Cursor or Codex into an AI scientist, from RDKit to Scanpy to 100+ databases. 🔗 github.com/K-Dense-AI/scientific-agent-skills
#scientificagentskills #python #aiagents
#scientificagentskills #python #aiagents
ChatGPT and Codex usage limits merge in December: OpenAI says Plus, Pro and Business plans will share one token-based allowance. No exact date or per-plan numbers yet. Instant chat doesn't count.
So is it CodexPT now? (Joke.) Facts on page 3.
#ChatGPT #Codex #OpenAI #webtoon
So is it CodexPT now? (Joke.) Facts on page 3.
#ChatGPT #Codex #OpenAI #webtoon
November 1, 2026 at 3:30 AM
OpenAI merges ChatGPT and Codex usage into one token pool from December
📅 October 31, 2026 • OpenAI
OpenAI published a new Help Center page on October 31 describing how ChatGPT and Codex usage limits work…
https://t.me/AIAutomationArchitecture
📅 October 31, 2026 • OpenAI
OpenAI published a new Help Center page on October 31 describing how ChatGPT and Codex usage limits work…
https://t.me/AIAutomationArchitecture
November 1, 2026 at 3:29 AM
OpenAI merges ChatGPT and Codex usage into one token pool from December
📅 October 31, 2026 • OpenAI
OpenAI published a new Help Center page on October 31 describing how ChatGPT and Codex usage limits work…
https://t.me/AIAutomationArchitecture
📅 October 31, 2026 • OpenAI
OpenAI published a new Help Center page on October 31 describing how ChatGPT and Codex usage limits work…
https://t.me/AIAutomationArchitecture
うぉぉぉないすぅ！ 昨日今日とCodexを、1週間リミットを使い切ってクレジットで使ってたら、ちょうど今リセットが配布された！ 完璧！
November 1, 2026 at 3:28 AM
うぉぉぉないすぅ！ 昨日今日とCodexを、1週間リミットを使い切ってクレジットで使ってたら、ちょうど今リセットが配布された！ 完璧！
GPT-6 Rosalind is out (Oct 30): OpenAI's life-sciences reasoning model, on the same weights as GPT-6 Astra. Its system card reports MedChemBench 49.6% vs 28.9% and GeneBench Pro 37.7% vs 16.7% over Rosalind-5.5. Trusted access only.
toolome.bio/tools/gpt-rosalind/ #AIforScience #DrugDiscovery
toolome.bio/tools/gpt-rosalind/ #AIforScience #DrugDiscovery
GPT-Rosalind: license, commercial use, paper — Bio LLM agents & copilots | Toolome
GPT-Rosalind: OpenAI's life-sciences reasoning models for biology, genomics and drug discovery; GPT-6 Rosalind (Oct 2026) runs on GPT-6 Astra weights. Trusted access, plus an open Codex research plugi...
toolome.bio
November 1, 2026 at 3:17 AM
GPT-6 Rosalind is out (Oct 30): OpenAI's life-sciences reasoning model, on the same weights as GPT-6 Astra. Its system card reports MedChemBench 49.6% vs 28.9% and GeneBench Pro 37.7% vs 16.7% over Rosalind-5.5. Trusted access only.
toolome.bio/tools/gpt-rosalind/ #AIforScience #DrugDiscovery
toolome.bio/tools/gpt-rosalind/ #AIforScience #DrugDiscovery
One issue I'm having with Dot is that it does not yet have browser password authentication saving so you can't save passwords in the browser inside of Dot like you could with Codex. I'm trying to work on a workaround right now.
November 1, 2026 at 3:16 AM
One issue I'm having with Dot is that it does not yet have browser password authentication saving so you can't save passwords in the browser inside of Dot like you could with Codex. I'm trying to work on a workaround right now.
It can either monitor them as they go or independently let them handle it and then report back.
- It can run shell commands.
- It can work with files. It can read completed local Codex threads.
- It can continue, create, archive, and restore disposable local Codex threads.
- It can run shell commands.
- It can work with files. It can read completed local Codex threads.
- It can continue, create, archive, and restore disposable local Codex threads.
November 1, 2026 at 3:15 AM
It can either monitor them as they go or independently let them handle it and then report back.
- It can run shell commands.
- It can work with files. It can read completed local Codex threads.
- It can continue, create, archive, and restore disposable local Codex threads.
- It can run shell commands.
- It can work with files. It can read completed local Codex threads.
- It can continue, create, archive, and restore disposable local Codex threads.
All right there were a few more updates to Codex as well as ChatGPT on Android. I did a couple of tests. Here's what the dot can do now:
- It can create desktop Codex tap tasks and guide them through follow-ups from beginning to end.
- It can create desktop Codex tap tasks and guide them through follow-ups from beginning to end.
November 1, 2026 at 3:15 AM
All right there were a few more updates to Codex as well as ChatGPT on Android. I did a couple of tests. Here's what the dot can do now:
- It can create desktop Codex tap tasks and guide them through follow-ups from beginning to end.
- It can create desktop Codex tap tasks and guide them through follow-ups from beginning to end.
If you give it too much work at some point, it may start outsourcing certain things to Codex but it's interesting how deeply integrated it is already.
November 1, 2026 at 3:15 AM
If you give it too much work at some point, it may start outsourcing certain things to Codex but it's interesting how deeply integrated it is already.
It doesn't work like the old voice chat, where it's just creating a bunch of Codex threads and then doing the work for you and managing it. It actually proactively does the work itself. It actually goes through and does all the work internally and then keeps you updated as stuff's done.
November 1, 2026 at 3:15 AM
It doesn't work like the old voice chat, where it's just creating a bunch of Codex threads and then doing the work for you and managing it. It actually proactively does the work itself. It actually goes through and does all the work internally and then keeps you updated as stuff's done.
I've had it do multiple things through multiple plugins over the last few days and so far the only time it's tried to or needed to create a Codex thread is when I told it to.
November 1, 2026 at 3:15 AM
I've had it do multiple things through multiple plugins over the last few days and so far the only time it's tried to or needed to create a Codex thread is when I told it to.
Most of the time it can handle it on its own unless there's a reason why it needs to do it in a thread format. If it does create a threading Codex, those threads that it creates don't currently have access to browser controls or computer use.
November 1, 2026 at 3:15 AM
Most of the time it can handle it on its own unless there's a reason why it needs to do it in a thread format. If it does create a threading Codex, those threads that it creates don't currently have access to browser controls or computer use.
You can only attach it to one laptop or one computer at a time. It won't let you connect multiple. I believe they'll fix that.
- Right now it can create Codex threads to do work, but most of the time you don't need to create threads.
- Right now it can create Codex threads to do work, but most of the time you don't need to create threads.
November 1, 2026 at 3:15 AM
You can only attach it to one laptop or one computer at a time. It won't let you connect multiple. I believe they'll fix that.
- Right now it can create Codex threads to do work, but most of the time you don't need to create threads.
- Right now it can create Codex threads to do work, but most of the time you don't need to create threads.
90% of the stuff you tell it to do, it will do on its own.
That doesn't mean it's going to be like voice chat, where you're generating a bunch of Codex threads and they're all working simultaneously and then it reports back. That's not how this works.
That doesn't mean it's going to be like voice chat, where you're generating a bunch of Codex threads and they're all working simultaneously and then it reports back. That's not how this works.
November 1, 2026 at 3:15 AM
90% of the stuff you tell it to do, it will do on its own.
That doesn't mean it's going to be like voice chat, where you're generating a bunch of Codex threads and they're all working simultaneously and then it reports back. That's not how this works.
That doesn't mean it's going to be like voice chat, where you're generating a bunch of Codex threads and they're all working simultaneously and then it reports back. That's not how this works.
or a Codex thread. It's going to do most of the stuff itself using its own computer. It can do multiple things at one time using the computers and the connectors on its own. It can access your laptop via files and CLI and stuff through the connector.
November 1, 2026 at 3:15 AM
or a Codex thread. It's going to do most of the stuff itself using its own computer. It can do multiple things at one time using the computers and the connectors on its own. It can access your laptop via files and CLI and stuff through the connector.