Back to All Events

#11 David Dalrymple (davidad) on Mathematics for Safeguarded AI program at ARIA

About the event

​In the eleventh episode of the ALIGN webinar series, we are pleased to invite David A. Dalrymple, a distinguished research scientist with extensive experience in AI safety and computational neuroscience. David Dalrymple currently leads the Safeguarded AI program within the broader Mathematics for Safe AI opportunity space at ARIA, the UK’s Advanced Research and Invention Agency. The Safeguarded AI program focuses on developing mathematically robust and human-auditable models to ensure the safe deployment of advanced AI systems, addressing the critical need for reliable and ethical AI technologies.

David Dalrymple holds an S.M. in Media Technology from MIT and has pursued Ph.D. studies in AI under Marvin Minsky and Biophysics at Harvard University. He has made significant contributions to the field of distributed computing as a co-inventor of the Filecoin protocol and has also developed the Hypercerts mechanism for public goods funding at Protocol Labs.

Throughout his career, David Dalrymple has been involved in numerous innovative projects, including the creation of a data-driven simulation of the nematode worm’s nervous system and advancements in AI governance concepts. His work has been widely recognized, and he has been an invited speaker at various prestigious events, such as TEDx, Protocol Labs’ Lab Day, and the Estonian Academy of Sciences.

In this webinar, David Dalrymple will share his insights on ARIA’s Safeguarded AI and his latest views on AI safety.

  • Date: Monday, 2 Sept, 2024 8:00 pm-9:00 pm (JST) = 12:00 pm - 1:00 pm (GMT+1)

  • This recording talk will not appear on the Youtube Channel.

ALIGNウェビナーシリーズの第11回では、ARIAの「Safeguarded AI」プログラムを、より広範な「Mathematics for Safe AI」の機会領域の一環として率いる著名な研究科学者であるDavid A. Dalrymple氏をお招きします。Dalrymple氏は、AI安全性と計算神経科学の分野で卓越した経歴を持ち、Filecoinプロトコルの開発および公共財資金調達のためのHypercertsメカニズムにおいて重要な貢献をしてきました。このウェビナーでは、Dalrymple氏がARIAの「Safeguarded AI」プログラムについて議論し、AIの安全な展開を確保し、信頼性と倫理を兼ね備えたAI技術を実現するための数学的に堅牢で人間が監査可能なモデルの作成に関する洞察についてお話いただきます。

  • 日時:2024年9月2日(月)8:00 pm- 9:00 pm (日本時間)

  • 講演は英語です。

  • 本イベントの事後の動画配信はありません。

Agenda

  • ​20:00-20:05 House keeping from ALIGN

  • ​20:05-20:35 David A. Dalrymple on Mathematics for Safe AI and Safeguarded AI for constructing guaranteed safety

  • ​20:35-20:55 Q&A and discussion with participants

  • ​20:55- closing

Previous
Previous
19 August

#10 LLMのアライメント技術 ゲスト:今井翔太さん(AI研究者)