(Biz)

Functional Requirements Functional requirements describe what the product should do, broken down into specific system behaviors and capabilities.

Non-Functional Requirements Non-functional requirements place constraints on how the product should do it and cover attributes like performance, security, scalability, reliability, and response time.

Business Requirements Business requirements represent organizational objectives and the "why" behind a project, focusing on desired outcomes like revenue increase or cost reduction.

Stakeholder Requirements Clear identification of who needs what and why, distinguishing between end users, sponsors, product owners, and internal teams.

Technical Requirements Specific constraints around infrastructure, technology stack, integrations, APIs, databases, and system dependencies.

Acceptance Criteria Conditions the product must satisfy to be accepted, which must be testable, concise, and completely understood by all team members.

Success Metrics Measurable indicators of whether the project achieved its goals (conversion rates, user adoption, performance benchmarks, etc.).

Constraints & Dependencies Budget, timeline, resource availability, regulatory compliance, third-party integrations, and external blockers.

Assumptions Documented beliefs about the environment, user behavior, technical feasibility, and market conditions that inform decisions.

Functional Decomposition Breaking down high-level functions into subfunctions, processes, and activities to the point where parts can't be broken down further.

Use Cases or User Flows Detailed scenarios showing step-by-step interactions between users and the system to accomplish goals.

Data Model How information flows through the system, what gets stored, what gets processed, and relationships between data entities.

Scope & Out-of-Scope Clear boundaries on what's included in this version/phase and what's explicitly excluded.

(tech) 

- Scope

- Architecture 

and 

1. devcontainer.json - Configure your development environment with a devcontainer.json file that specifies the Linux OS, automatically installed tools/runtimes/frameworks, port forwarding, environment variables, editor settings, and extensions

2. Dockerfile (or use prebuilt images) - Define your container build either with a Dockerfile or reference prebuilt dev container images

3. .gitignore - Exclude files you don't want tracked in version control

4. README.md - Include instructions about environment requirements, dependencies, and allowed libraries for the project

5. Dependency/Build Files - package.json (Node), requirements.txt (Python), pom.xml (Java), etc.

6. postCreateCommand Configuration - Automate setup commands like npm install that run after the container builds

7. Port Forwarding Configuration - Define which ports need to be accessible for development/testing (e.g., 3000 for web apps, 9229 for debugging)

8. VS Code Extensions List - Specify extensions in devcontainer.json customizations so developers get the same tooling automatically

9. Environment Variables/Secrets Setup - Configure how sensitive data (tokens, API keys) will be handled in the codespace

10. Entry Point Documentation - Clear instructions on what commands to run first to get the app running (e.g., npm run dev)

前述の「アカデミア水準の必須／禁止要素」を、トーンを落とし、
《東大新聞のコラム》のような
「静か・抑制的・知的だが肩に力が入っていない文体」
で仕上げるための ToDo / Not ToDo リストです。ポイントは“学術論文の厳密さを保ちつつ、主張しすぎない” ことにあります。

✅ To Do（やること）

1. 問題意識を「問い」として置く

• 結論を先に断定しない

• 「〜ではないだろうか」「〜と考えられる」と余白を残す

• 自分の立場を一歩引いた場所に置く

例：
×「本フレームワークは〜を解決する」
○「〜を考える一つの枠組みとして位置づけられるかもしれない」

2. 先行研究・既存枠組みを“敬意をもって”扱う

• 批判は否定ではなく「射程の違い」として書く

• 名前を出しすぎず、流れとして紹介する

• 「乗り越える」より「別の角度を与える」

3. 概念は簡潔に、説明は抑制的に

• 定義はするが、力説しない

• 図解や比喩は補助に徹する

• 読者の理解を信頼する

4. 自身のフレームワークを「提案」に留める

• 主張の語尾を弱める

• 「唯一」「決定的」を使わない

• 選択肢の一つとして置く

5. 実証・事例は“静かに”差し込む

• 成功談にしない

• 劇的なBefore/Afterを描かない

• データや事例は説明的に扱う

6. 限界・留保を自然に含める

• 独立した「限界」章にしなくてもよい

• 文中にさらっと織り込む

• 読者が批判できる余地を残す

7. 文体は「記者と研究者の中間」

• 一人称は最小限

• 感情語を使わない

• 語彙は平易だが安っぽくしない

8. 結論を「まとめ」にしない

• 断定で終わらない

• 問いを開いたまま閉じる

• 読後に思考が続く構成

❌ Not To Do（やらないこと）

1. フレームワークを“売らない”

• 有用性を強調しすぎない

• 読者に使わせようとしない

• 「役に立つ」アピールをしない

2. 学術的優位性を誇示しない

• 「従来研究は不十分」などの言い切り

• 過度な独自性強調

• 専門用語でのマウンティング

3. 読者を啓蒙しない

• 「私たちは〜すべきだ」

• 読者を無知な存在として扱う書き方

• 教訓的結論

4. 図やモデルを主役にしない

• 図で納得させようとしない

• 視覚的インパクト重視

• フレームワーク名の連呼

5. 物語化しすぎない

• ヒーロー／問題解決構図

• ドラマ的起承転結

• 自己体験中心

6. 「東大っぽさ」を演出しない

• 知的アピールのための難語

• 不必要な引用羅列

• 内輪感のある表現

📝 東大新聞コラム的バランス感覚（要約）

• 強く言わないが、浅くもしない

• 断言しないが、逃げもしない

• 主張するが、押しつけない

• 読者を信頼する

✅ 必須な要素（Must-have）

1. 明確な研究目的・研究課題

• 既存研究では何が未解決か

• なぜ新しい思考フレームワークが必要なのか

• 対象分野・適用範囲の明示

2. 先行研究の網羅的レビュー

• 類似・競合するフレームワークの整理

• 理論的系譜（どの学派・理論に基づくか）

• 自身の貢献がどこに位置づくかの明確化

3. 理論的基盤の厳密な定義

• 使用する概念・用語の明確な定義

• 暗黙知・曖昧語の排除

• 前提条件（assumptions）の明示

4. フレームワーク構造の論理的一貫性

• 要素間の関係が論理的に説明可能

• 因果関係・相互作用・階層構造の明示

• 図解と文章の完全一致

5. 方法論の妥当性

• フレームワーク構築プロセスの説明

  • 理論導出型か

  • データ駆動型か

  • 混合型か

• 主観的判断と客観的根拠の区別

6. 実証・検証（Validation）

• ケーススタディ

• 定性的／定量的検証

• 再現性・反証可能性への配慮

7. 限界と適用条件の明示

• 使えない状況

• 想定していない前提条件

• 理論的・実務的限界

8. 学術的文体・構成

• IMRAD等の標準構成

• 感情表現・煽り表現の排除

• 一貫した引用・参考文献形式

9. 学術的貢献の明示

• 理論的貢献（新概念・新関係性）

• 方法論的貢献

• 応用的貢献（実務・政策等）

❌ あってはならない要素（Must-not-have）

1. 「経験上」「感覚的に」といった主観依存

• 個人の成功体験のみを根拠にする

• データ・理論なき一般化

2. 既存理論の言い換え・焼き直し

• 名前だけ新しいが中身が既存理論

• 先行研究を無視した独自性主張

3. 概念の曖昧さ

• 定義されないキーワード

• 文脈によって意味が変わる用語

• 比喩だけで説明される概念

4. 反証不能な構造

• 失敗した場合の説明が常に後付け

• 「使い方が悪い」で済ませられる設計

5. 過度な汎用性の主張

• 「すべてに使える」「万能」

• 適用範囲を限定しない姿勢

6. 実証なき図解偏重

• 見た目は美しいが論証がない

• 図が論理の代替になっている

7. ビジネス書・自己啓発書的語り口

• キャッチコピー優先

• 読者の共感誘導

• 成功物語中心の構成

8. 批判・比較の欠如

• 自理論の弱点に触れない

• 他理論との比較を避ける

9. 引用不備・出典不明

• 出典が曖昧

• 孫引きの多用

• 学術的信頼性を損なう引用

🔎 補足視点（アカデミア特有）

• 「役に立つ」より「説明できる」

• 「新しい」より「位置づけられる」

• 「分かりやすい」より「厳密」

🧠 Lean Copilot–Specific & Very Close

Lean-AI / LLM-for-Lean community

• Lean Zulip → #ai and #llm streams

• Regular discussions on:

  • tactic prediction

  • proof step synthesis

  • context window compression

  • Copilot-style autocomplete for Lean 4

• Many contributors to:

  • lean-copilot

  • LeanDojo

  • Pantograph

  • ProofNet

➡️ This is currently the closest thing to a living Copilot-Lean working group.

LeanDojo Community

• Open research ecosystem for learning-based theorem proving

• Weekly or bi-weekly open research meetings (depending on semester)

• Focus:

  • interactive proof search

  • dataset construction

  • RL + transformers for Lean

• Strong overlap with Copilot-style workflows (step-by-step guidance)

🤖 AI for Theorem Proving (Lean-heavy)

AITP (AI for Theorem Proving)

• Workshop + year-round research network

• Lean is now a primary target system (alongside Isabelle, Coq)

• Includes:

  • tool demos

  • reading groups

  • shared benchmarks

• Strong theoretical CS / ML crossover

ITP + AI Satellite Groups

• Communities around:

  • ITP (Interactive Theorem Proving)

  • CICM

  • CADE + IJCAR AI tracks

• Often host online seminar series during the year

🧩 Proof Assistants + ML (Broader)

ML4PL (Machine Learning for Programming Languages)

• Not Lean-only, but many Lean Copilot papers appear here

• Autocomplete, code synthesis, proof synthesis all overlap

Big Proof / Formal Math initiatives

• Formal Mathematics / Formalization of Mathematics

• Mathlib community meetings

  • Occasionally host AI-assisted proof talks

• Proof assistants + NLP workshops

🧪 Industry-adjacent but Research-driven

• DeepMind – Mathematical Reasoning & Lean

  • Open talks & seminars (not a WG, but recurring)

• Meta / FAIR theorem proving research

• OpenAI formal reasoning research (Lean-focused)

🧷 Where to actually participate (practical advice)

If your goal is active participation, do this:

1. Join Lean Zulip

   • Follow: #ai, #tactics, #mathlib, #llm

2. Track AITP & ITP workshops

   • They often spin up temporary working groups after workshops

3. Follow LeanDojo announcements

   • Open calls are usually shared publicly

4. Mathlib community calls

   • Occasionally spin off AI-focused subcalls

🧭 TL;DR

🔐 Software Supply Chain & Security

OpenSSF (beyond the ones you already attend)

• Vulnerability Disclosures WG – coordinated disclosure, CNA processes

• Security Baseline WG – minimum security requirements for OSS

• Best Practices WG – tooling & badge criteria evolution

• Identifying Security Threats WG – threat modeling for OSS

OWASP (besides SCVS)

• OWASP Top 10 Proactive Controls

• OWASP CycloneDX WG – SBOM format (very active)

• OWASP ASVS – Application Security Verification Standard

• OWASP Dependency-Track Project

• OWASP Firmware Security Project

CNCF / Cloud Native Security

• TAG Security (umbrella group, very active)

• SIG Security (Kubernetes)

• Supply Chain Security SIG (K8s)

• Policy WG (OPA / Gatekeeper ecosystem)

📦 SBOM, Provenance & Artifact Integrity

• CycloneDX Core WG

• SPDX Technical Team (Linux Foundation)

• in-toto Steering Committee

• Sigstore Policy & UX Working Groups

• OCI Artifacts / OCI Security WG

🧪 Standards & Specifications

Ecma / ISO / W3C

• Ecma TC54 (SBOM & SW transparency) – adjacent to TC54-TG2

• ISO/IEC JTC 1 SC 38 – cloud & distributed platforms

• W3C WebAppSec WG

• W3C Privacy CG / Security Interest Group

🤖 AI, ML & Security (fast-growing)

• OpenSSF AI/ML Security WG

• MLCommons Security WG

• NIST AI RMF Community of Practice

• OWASP Top 10 for LLM Applications

🏗️ Infrastructure & Platform Security

• Confidential Computing Consortium (CCC)

• IETF SAAG (Security Area Advisory Group)

• IETF SCITT WG – supply chain transparency (very relevant)

• TUF Community Meetings

🧭 Governance, Risk & Ecosystem Trust

• FINOS Security SIG

• CHAOSS Risk & Security WG

• Linux Foundation Trust & Safety Initiative

• OpenJS Security WG

📌 If you want something closer to your calendar…

Based on what you’re already attending, the closest matches you may want to look into are:

• IETF SCITT WG

• CycloneDX WG

• Kubernetes Supply Chain Security SIG

• OpenSSF Vulnerability Disclosures WG

• MLCommons / AI supply chain groups

------

🔐 Cryptography & Mathematical Security Foundations

IACR (International Association for Cryptologic Research)

These are not “WGs” in name, but ongoing, highly active research communities with regular workshops, mailing lists, and study groups:

• CRYPTO / EUROCRYPT / ASIACRYPT communities

• Theory of Cryptography Conference (TCC)

• Real World Crypto (RWC) – applied but still very theory-aware

• IACR ePrint Cryptography Archive (active discussion + review culture)

➡️ Strong focus on: number theory, algebra, complexity theory, zero-knowledge, MPC, post-quantum crypto.

Post-Quantum Cryptography

• NIST PQC Forum & Study Groups

• IETF PQUIP WG (Post-Quantum Use in Protocols)

• ETSI Quantum-Safe Cryptography ISG

➡️ Heavy math: lattices, codes, isogenies, hardness assumptions.

📐 Formal Methods, Logic & Verification (Security-oriented)

Formal Methods Groups

• IFIP WG 1.6 – Rewriting

• IFIP WG 1.7 – Theoretical Computer Science

• IFIP WG 11.2 – Pervasive Systems Security (formal angle)

Program Verification & Proof Systems

• Proof-Carrying Code (PCC) community

• CompCert & Verified Compilation community

• Verified Crypto community (HACL*, Fiat-Crypto, Jasmin)

Tools & ecosystems:

• Coq Security & Crypto Users Group

• Isabelle Security Group

• Lean for Security / Crypto reading groups

🧠 Programming Languages & Semantics (Security Focus)

• POPL Security Track community

• CSF (IEEE Computer Security Foundations Symposium)

• PLDI Security & Verification subgroup

• OOPSLA Formal Security & Types community

Topics:

• Type systems for security

• Non-interference

• Language-based security

• Information flow control

🔎 Security Foundations & Logic

Dedicated Security Theory Venues

• IEEE CSF (Computer Security Foundations)

• POST (Principles of Security & Trust – ETAPS)

• FOSAD (Foundations of Security Analysis and Design)

• SecDev Theory Track

Access Control, Logic, and Models

• ABAC & Policy Logic research groups

• Modal logic for security protocols

• Game-theoretic security models

• Epistemic logic & knowledge-based security

🧪 Protocol Analysis & Symbolic Models

• Formal Protocol Verification community

  • ProVerif

  • Tamarin

  • AVISPA

• Applied π-calculus groups

• Security Protocols Workshop (Cambridge)

🧬 Systems + Theory Hybrid Groups

• USENIX Security (theory-heavy subcommunity)

• SOSP/OSDI formal verification clusters

• Microarchitectural Security + Formal Models (Spectre/Meltdown theory work)

🧩 Category-Theory / Abstract Math Adjacent Security

These are smaller but intellectually deep:

• Applied Category Theory for Cryptography

• Monoidal categories & composable security

• UC (Universal Composability) theory community

• Game-based vs simulation-based security groups

🎓 Ongoing Reading Groups & Seminars (Informal but Active)

• Cryptography Reading Groups (many universities, often open)

• Security Semantics Seminar Series

• Formal Methods in Security (FM-Sec) workshops

• Mathematical Foundations of Cybersecurity seminars

interdisciplinary, flexible, cross-faculty, and often emphasizing “human + society + environment + science.”

🌍 20 Similar Departments / Programs Worldwide

United States

1. Stanford University – Symbolic Systems Program
   (Interdisciplinary: psychology, linguistics, philosophy, computer science)

2. Harvard University – Social Studies
   (Humanities + social sciences with broad theoretical foundations)

3. University of California, Berkeley – Interdisciplinary Studies Field Major (ISF)
   (Highly flexible, student-designed academic paths)

4. Brown University – Liberal Medical Education / Open Curriculum Studies
   (Open, integrative undergraduate design)

5. University of Michigan – Program in the Environment (PitE)
   (Human–environment interdisciplinary focus)

6. University of Pennsylvania – Cognitive Science Program
   (Human + mind + computation)

7. New York University – Gallatin School of Individualized Study
   (Self-designed interdisciplinary degree comparable to 総人’s flexibility)

Canada

1. University of Toronto – Human Biology / Cognitive Science / Peace & Conflict Studies (tri-campus integrative programs)
   (Interdisciplinary human-focused studies)

2. McGill University – Interfaculty Program in Environment
   (Cross-faculty study on humans, society, and environment)

Europe

1. University of Oxford – Human Sciences
   (Direct parallel: integrates biology, anthropology, sociology, demography)

2. University of Cambridge – Human, Social, and Political Sciences (HSPS)
   (Broad: anthropology, sociology, politics, international relations)

3. University College London (UCL) – Arts & Sciences (BASc)
   (Purpose-built interdisciplinary degree similar to 総人’s philosophy)

4. Leiden University – Liberal Arts and Sciences (Global Challenges)
   (Interdisciplinary program at a major research university)

5. University of Amsterdam – Interdisciplinary Social Science
   (Mix of psychology, sociology, anthropology, public policy)

6. Humboldt University of Berlin – European Ethnology / Cultural Studies
   (Cross-cutting cultural/human studies)

Asia (outside Japan)

1. National University of Singapore – University Scholars Programme
   (Highly interdisciplinary, with philosophy/tech/environment options)

2. Hong Kong University – Bachelor of Arts & Sciences (BASc) Programmes
   (Humanity + science + technology + society)

3. Seoul National University – College of Liberal Studies
   (Open interdisciplinary program where students design their path)

4. Tsinghua University – Xuetang/Experimental Interdisciplinary Programs
   (Cutting across humanities, social sciences, environment, and more)

Oceania

1. Australian National University – Bachelor of Philosophy (Interdisciplinary)
   (Research-led and integrative across humanities/science/social science)

(GPT-5)

Overview of the Faculty

• Officially called the Faculty of Integrated Human Studies, it aims to provide an education that spans the humanities, social sciences, and natural sciences.

• The educational philosophy emphasizes “discovering new connections among humans, civilization, and nature” and “creating academic fields that bridge the humanities, social sciences, and natural sciences.”

• It consists of 10 divisions, including Mathematical & Information Science, Human/Society/Thought, Arts & Culture, Cognitive & Behavioral Sciences, Linguistics, East Asian Civilizations, Coexistence Studies, Cultural & Regional Environment, Material Sciences, and Earth & Life Environment.

• The entrance examination is designed for both science- and humanities-oriented applicants, requiring broad academic ability and flexible thinking.

Based on these points, let’s organize the Pros and Cons, along with what kind of students are best suited for the program.

✅ Pros

1. Broad academic scope

• Because the program is designed to cross boundaries between the humanities, social sciences, and natural sciences, it is highly appealing for students who don’t want to be forced into a strict “science vs. humanities” track from the start.

• With options ranging from Mathematical & Information Science to Arts & Culture and Coexistence Studies, it accommodates a wide variety of interests.

• This gives students the flexibility to explore academic directions before committing to a speciality.

2. Strong university brand and research environment

• As part of Kyoto University—a major national research university—students benefit from an environment rich in cross-faculty and interdisciplinary opportunities.

• There are also institutional initiatives like the Center for Transdisciplinary Research that encourage activity beyond traditional academic boundaries.

3. Expanded career and academic pathways

• The breadth of study helps students explore their interests before deciding on a career path or graduate school specialization.

• It is a good fit for those pursuing “fusion” fields that combine humanities and sciences, or those who want to understand complex social, environmental, or human issues.

4. Emphasis on thinking, expression, and collaboration

• The admissions policy highlights students who have foundational knowledge, independent thinking, good judgment, expressive ability, and a willingness to collaborate with others.

• The program aims to cultivate these abilities further.

❌ Cons

1. Specialization can be relatively shallow

• Because academic scope is wide, students who wish to pursue a sharply focused specialty—such as theoretical physics, pure mathematics, or advanced experimental fields—may feel underserved.

• Some online opinions criticize the program for lacking depth compared to more narrowly focused departments.

• In short: “wide and shallow” can be both a strength and a weakness.

2. Career direction can be unclear

• If students enter without a clear academic interest—e.g., “I’ll just enter Soujin for now and think later”—they may struggle to establish an area of expertise over the four years.

• Some graduates report leaving without a strong sense of what their “specialty” is, making job hunting feel vague.

3. Variation among divisions requires adjustment

• With ten divisions across humanities, sciences, arts, and natural sciences, students who lack a clear interest may find it difficult to choose a division or understand what to prioritize.

• Different divisions have different expectations, learning styles, and required preparation.

4. Gap between brand image and actual preparedness

• Being part of Kyoto University raises expectations, and some online criticism suggests that students need self-driven academic planning or they may fall behind.

• There is also a perception among some that “Soujin is more relaxed,” which can create a mismatch between student expectations and academic demands.

🎯 Who is this program suitable for?

Best suited for:

• Students interested in the interconnectedness of humans, society, nature, and civilization—beyond traditional academic boundaries.

• Self-directed learners who want to design their own academic path and think carefully about what they want to study and what kind of person they want to become.

• Those who want to explore interdisciplinary or hybrid fields such as environmental studies, health, linguistics, regional studies, information studies, and so on.

• Students who value freedom and flexibility within the environment of a top-tier research university.

May not be suitable for:

• Students who already have a very clear specialized goal (e.g., “I want to become a physicist / doctor / mathematician / chemist”) and need a tightly structured technical curriculum.

• Students who want to drift through university without much intentional planning—Soujin requires self-directed learning to take advantage of its freedom.

• Those who prefer clearly defined academic paths with limited choices.

Here are people from our earlier list who very clearly and publicly talk about the benefits of interdisciplinary / liberal-arts style education, with concrete sources (interviews, podcasts, talks, posts).

1. Reid Hoffman

(Stanford Symbolic Systems → LinkedIn co-founder)

Where he talks about it

• Tweet / micro-blog – He literally wrote that there “should be more conversation about the value of a liberal arts education in startups/tech.” (X (formerly Twitter))

• Book excerpt / interview (“AI Valley” / Business Insider) – Describes choosing Symbolic Systems, which melded computer science, linguistics, psychology and other disciplines, and says he valued the opportunity to take a wide array of classes across campus, not just straight CS. (Business Insider)

• Articles summarising his view – Multiple pieces note that Symbolic Systems is an interdisciplinary major, and that this background gave him a “multidimensional understanding” of technology and human behaviour, which he then applied in entrepreneurship and investing. (Wikipedia)

• Philosophy / liberal-arts advocacy – Business and opinion pieces cite him as an example of a tech billionaire who credits his philosophy / liberal-arts training as part of his success, used to think more broadly than a narrow technical track. (Inc.com)

TL;DR: Hoffman is very explicit that broad, cross-disciplinary / liberal-arts style study is valuable for tech and startups, and he regularly says so in talks, interviews, and posts.

2. Marissa Mayer

(Stanford Symbolic Systems → ex-Yahoo CEO, early Google)

Where she talks about it

• Video interview / transcript (Makers: Women Who Make America) – She says she noticed Stanford was strong in psychology and computer science and then “found this interesting interdisciplinary major called Symbolic Systems” which combines philosophy, psychology, linguistics and computer science – and that choosing this major was one of the key decisions that led her into computer science and really got her interested. (lifestories.org)

• Biographical pieces also describe Symbolic Systems as an interdisciplinary field combining CS with cognitive psychology, philosophy etc., framing it as the foundation for her later tech career. (BUHAVE)

TL;DR: Mayer explicitly calls out Symbolic Systems as an interdisciplinary major and directly links it to why she got excited about CS and tech.

3. Mike Krieger

(Stanford Symbolic Systems → Instagram co-founder)

Where he talks about it

• Entrepreneurship articles summarizing his interviews – They describe how, at Stanford, he discovered Symbolic Systems, “which uniquely integrated computer science, design, philosophy, and psychology.” According to the piece, this interdisciplinary approach taught him the importance of building products that solve real problems, prototyping, and forming strong teams. (NextBigWhat)

• Other write-ups about him emphasize that Symbolic Systems is an interdisciplinary field that blends coding with psychology, linguistics and philosophy, and link that directly to how he thinks about user-centric product design. (Forbes)

TL;DR: Krieger explicitly frames the interdisciplinary nature of his degree as shaping how he builds products and understands users.

4. Jim Yong Kim

(Physician-anthropologist → ex-World Bank President, ex-Dartmouth President)

Where he talks about it

• Dartmouth talk / interview on health-care reform – As Dartmouth’s president, he said that health-care delivery is a highly interdisciplinary field, and that making it a real science “can only result from a liberal arts education starting at the undergraduate level.” (Dartmouth Sites)

• Academic articles about public-health education also quote him arguing for public health and health-care delivery to be taught in a broad, liberal-arts context, not just as narrow technical training. (PMC)

TL;DR: Kim very explicitly connects interdisciplinary + liberal-arts undergraduate education with being able to solve complex real-world problems (like health-care systems).

5. Ban Ki-moon

(Former UN Secretary-General, Seoul National University alumnus)

Where he talks about it

• UN Academic Impact launch speech (2010) – He praises changes in universities, noting that departments are coming together to create interdisciplinary degrees and that this is producing well-rounded scholars able to tackle global challenges. (United Nations)

• Recent education & climate forums – At events hosted by the Ban Ki-moon Centre, he argues that solving climate and sustainability issues requires holistic, interdisciplinary climate education, rather than siloed disciplines. (Ban Ki-moon Centre)

TL;DR: Ban Ki-moon repeatedly calls for interdisciplinary degrees and education as necessary to create global-citizen leaders for complex problems like democracy, climate, and health.

What about the other people on the big list?

I did a targeted scan for the rest of the famous names we mentioned (Emma Watson, Lady Gaga, Kevin Rudd, etc.). I found:

• Plenty of references to where they studied and that some of them chose liberal-arts style programs,

• But far fewer cases where they themselves, in a clear quote, talk about the benefits of “interdisciplinary” education the way Hoffman / Mayer / Krieger / Kim / Ban Ki-moon do.

So, from our big pool, these five are the ones where I can confidently point to public material (interviews, talks, posts, or articles summarizing their own comments) that explicitly connects their success or worldview to interdisciplinary / liberal-arts education or approaches.

Below are clear, practical examples of what a good vs bad academic conference poster typically looks like. These are illustrative patterns, not specific designs, but they will help you understand what reviewers and attendees expect.

✅ Good Academic Poster (Example)

Title:

"Explainable Graph Neural Networks for Molecular Property Prediction"

Overall Characteristics

• Clean, professional layout with clear visual hierarchy

• Large, readable fonts (title ~85–120 pt, headings ~50–70 pt, body ~24–32 pt)

• Logical flow: Introduction → Methods → Results → Conclusion → QR code

• About 30–40% visuals (figures, diagrams, charts)

• Minimal text and clear bullet points

• Background light and unobtrusive

• Consistent color scheme (e.g., blues + greys)

Section Examples

Introduction

• Molecular property prediction is essential for drug discovery.

• Graph Neural Networks (GNNs) provide state-of-the-art accuracy, but lack interpretability.

• Goal: Create an interpretable GNN that highlights molecular substructures driving model predictions.

Methods

• Dataset: QM9 molecules (N = 130k)

• Model: Graph Attention Network with substructure attribution

• Explainability: Grad-CAM-like graph heatmaps

• Training: 80/10/10 split, early stopping, Adam optimizer

Method Figure: Simple pipeline diagram showing data → GNN → explanation map.

Results

• MAE reduced by 12% vs baseline GNN

• Explanations align with known functional groups

• User study: Chemists rated explanations "helpful" (4.3/5)

Visuals:

• Heatmap images over molecular graphs

• Bar chart of performance comparison

Conclusion

• Method improves accuracy and interpretability.

• Future work: extend to 3D conformers and multiple tasks.

QR code linking to paper/code.

❌ Bad Academic Poster (Example)

Title:

"GNNs AND MOLECULES: A STUDY"

Overall Characteristics

• Title too vague and small

• Walls of text—paragraphs 10–15 lines long

• Small font (<16 pt)

• Inconsistent fonts and colors (red, neon green, black, purple)

• Low-resolution figures, poorly aligned

• No clear story or flow

• Background is distracting (e.g., gradient or photo)

Section Examples

Introduction

A full dense paragraph that repeats the abstract of the paper verbatim, containing citations, long sentences, and irrelevant background…

“GNNs have been studied extensively in recent years [1,2,3,4] and are useful for many fields such as chemistry, physics, material science, etc. In this study, we examine many aspects of molecular prediction using GNNs which is an important problem because…”

(continues for 200+ words)

Methods

• Long block of code copied from the paper

• No visuals

• Overly detailed training parameters (batch sizes for every experiment, full hyperparameter tables)

Results

• Three tables with tiny text

• Overly technical metrics without context

• No charts, no explanation

Conclusion

• Generic statements with no takeaways

• No QR code

⭐ Key Takeaways

Good Posters

• Tell a story at a glance

• Use visuals → not text

• Use consistent design

• Highlight contributions

• Make results easy to digest

• Include a QR code for details

Bad Posters

• Overwhelm with text

• Look cluttered or inconsistent

• Use small fonts or unreadable visuals

• Fail to emphasize the main message