Projects

Dot triages 2,900 emails toward inbox zero

@thsottiaux

added

Down to 6110 unread from the over 9000 unread emails just earlier. Will get to inbox zero with clean filters within the next 48 hours. Dot has accepted the challenge.

Tibo (Thibault Sottiaux, OpenAI — works on Codex and ChatGPT) reports first-hand that his Dot has taken up the inbox-zero challenge: down to 6,110 unread from over 9,000 unread emails, with clean filters targeted within 48 hours. A concrete, measurable email-triage result from a Dot acting autonomously on the user's Gmail.

View on X

More Dots use cases in Projects

See all 201 →
My Dot works overnight — I wake up to finished tasks

My Dot works overnight — I wake up to finished tasks

いや、OpenAIの新しいAIエージェントのDotsエグいわ。文字通り24時間働くAI社員。寝る前にこう書いといて、タスクを投げとくだけで起きたら仕事が終わってる。↓↓↓ これから8時間就寝しますので、各セッションの残タスクを、これまでの依頼・承認の範囲で自律的に進めてください。利用可能なゴール設定やDotsの定期確認で停止した未完了セッションを再開し、行き詰まったら別の解決方法へ切り替え、私の入力待ちは保留して他の作業を進め、重要な変化だけ通知し、8時間後に完了事項・残タスク・保留理由を報告してください。↑↑↑ つい少し前まで「AIには１つずつ命令をしようね」とか言ってた気がするんだけど…。もう完全に俺よりシゴデキ。俺本体のほうは筋トレと睡眠時間がどんどん増える。どんどん健康になってしまう。腕パッツパツ。
@koh_kira
Dot proactively flagged an 8-day-stale O-1 visa application at 8am

Dot proactively flagged an 8-day-stale O-1 visa application at 8am

The AI Advantage — first 24 hours with his ChatGPT Dot. Overnight the Dot used his connected plugins to scan for work to do. At 8:00 a.m. it proactively surfaced two findings: an upcoming US trip, and an O-1 visa application with no status update in 8 days — suggesting a follow-up action to talk to his assistant about it. He calls it 'super proactive.' Also created a 'your personal scratchpad' image overnight. Concrete proactive-morning-briefing result from real first-hand use.
@aiadvantage
Dot named Winston orchestrated a website build via Codex

Dot named Winston orchestrated a website build via Codex

Chase (Chase AI) — early access to ChatGPT Dots. His dot 'Winston': over a voice call he asked it to create a website in a new Codex thread listing the top five Dots use cases and comparing Dots vs Muse and Grokbot. Winston set up the Codex thread, delegated the build, and walked Soul through the implementation — verifying product claims as it went. The website ended up up and running with the full comparison, which Chase verifies on camera. The win wasn't the site itself but the orchestration: one instruction to the Dot produced a delegated, multi-agent build. He also flags a quirk: the task landed in Recents instead of inside the Dots project folder.
@ChaseAI
Dot reviewed my website copy and redid the whole site — vs my own agent

Dot reviewed my website copy and redid the whole site — vs my own agent

自己用心開發的 Agent 跟現成的 ChatGPT dot 差在哪？ 一樣要求看我的 AI Agen 自學網找出文案的矛盾處和給建議，結果天差地別。 我自己開發的 AI Agent，雖然只用 35B 開源模型， 仍能有效找出文案衝突或不好懂的地方。 而 ChatGPT dot 號稱只用最強模型 GPT 6 Astra， 用了超久的時間，洋洋灑灑列一大堆要改的地方， 還重做了整個網站，把我本來花很多時間修改， 強調特色的文案，都換成平庸的 AI 文。 因為自己開發更聰明，超越 Astra 嗎？ 不是。單純是更懂自己而已。 而自己開發，就是想讓 AI Agent 更懂自己。
@darks0603
Dot auto-registers schedules to calendar and pushes notifications

Dot auto-registers schedules to calendar and pushes notifications

자비스처럼 작동하는 앱 같은거 만들었었는데 토큰 문제때매 나중에나 써먹어야지 생각했는데 Dots가 지피티에 나왔다 개인 비서다 완전... 지피티로 내 일정들 옮기고 그랬는데 알아서 캘린더에 등록도 해주고 알림도 준다. 자기 의견도 내고 메일은 어차피 사내메일 써서 지금 안 쓰고 있긴한데 메일을 수시로 체크하니 메일쪽에 자동매매 보고서 수시로 보내거나 하면 또 알아서 볼수도 있고.. 활용은 무궁무진하다 이거 진짜 괜찮고 좋은데?? 오호
@infp_horong