What does 'open science' entail?

Open science encompasses practices like sharing research data and code, and publishing in journals that are not behind paywalls. The goal is to increase transparency, reproducibility, and accessibility of scientific knowledge.

How does Fermat's Library help with discovering new research?

Currently, Fermat's Library focuses on annotation and community around specific papers. The team plans to develop features like a recommendation engine and improved discoverability tools for its arXiv extension in the future.

Why is publishing negative results often difficult for academics?

Academics are often incentivized to publish positive, high-impact results to secure jobs or tenure. Publishing negative findings, which can be valuable for avoiding redundant research, is less common due to this incentive structure.

What is p-value hacking and why is it a concern?

P-value hacking refers to manipulating data or analysis until a statistically significant result (low p-value) is achieved, potentially leading to unreliable or misleading findings. It's a concern because it undermines the integrity of published research.

Can you explain the Simpson's Paradox?

Simpson's Paradox occurs when a trend appears in different groups of data but disappears or reverses when these groups are combined. It highlights how aggregate data can be misleading compared to disaggregated data.

Why have scientific papers become so much longer over time?

The increase in paper length is likely due to the development of more complex research questions, the need for greater methodological detail for reproducibility, and more formal publishing standards, rather than a lack of confidence.

What makes scientific writing impactful?

Impactful scientific writing is characterized by clarity, accessibility, and the ability to effectively communicate complex ideas. Sometimes, papers that synthesize and explain existing research have a greater impact than the original discovery papers.

How can the scientific publishing system improve?

The current system faces challenges with paywalls, slow peer review, and misaligned incentives. Supporting open-access journals, encouraging well-known scientists to publish in them, and leveraging platforms like Fermat's Library can drive positive change.

Is it feasible to create a tool for annotating books like academic papers?

Yes, annotating books is a logical extension of the concept, especially for introductory texts. Challenges include copyright, but the potential for collaborative, updatable, and open-source book annotations is significant.

How does Fermat's Library balance being a side project with its growing impact?

The founders view it as a passion project, driven by their interest in academic reading. They leverage their existing workflow and find motivation through peer interaction and the desire to contribute to open science.

Key Moments

Fermat's Library Cofounders João Batalha and Luís Batalha

Y Combinator

Science & Technology4 min read74 min video

Jul 6, 2018|4,960 views|101|2

YC Y Combinator Fermat's Library João Batalha Luís Batalha Podcast Interview Craig Cannon arXiv papers librarian research

Save to Pod

Key Moments

TL;DR

Fermat's Library annotates papers to foster open science and collaboration, offering a platform for researchers to discuss and share knowledge.

Key Insights

Fermat's Library provides a platform for annotating academic papers, supporting LaTeX and Markdown, to clarify complex ideas.

The platform aims to promote open science by facilitating open data, open code, and open publishing, removing paywall barriers.

Fermat's Library enables remote scientific collaboration, drawing parallels to successful online collaborative projects like GitHub and Wikipedia.

Discoverability in scientific literature is a major challenge, and Fermat's Library plans to enhance this through features like paper recommendations.

The initiative addresses issues like the publication of negative results and p-value hacking, aiming to make scientific research more transparent and reliable.

Annotating books and educational materials is seen as a valuable extension, making complex subjects more accessible to a wider audience.

THE ORIGINS AND MISSION OF FERMAT'S LIBRARY

Fermat's Library was founded by brothers João and Luís Batalha, along with two other co-founders, who shared a technical background and a passion for reading academic papers. Dissatisfied with the density and occasional unintelligibility of scientific literature, they developed a platform to annotate papers directly in the browser. This tool supports LaTeX and Markdown, allowing users to add explanations and content to difficult sections, effectively bringing the collaborative spirit of offline journal clubs online. Their core mission is to advance open science, making research more accessible and reproducible for everyone.

PROMOTING OPEN SCIENCE AND COLLABORATION

The founders of Fermat's Library are strong advocates for open science, a movement encompassing open data, open code, and open publishing. They believe in sharing research data and algorithms to facilitate replication and further study. By promoting open publishing, they aim to dismantle the paywalls that restrict access to valuable scientific findings. The platform also serves as a catalyst for remote collaboration among scientists, inspired by the success of online communities like GitHub and Wikipedia, fostering a more interconnected and productive research ecosystem.

ADDRESSING CHALLENGES IN SCIENTIFIC DISCOVERY

Fermat's Library tackles several persistent challenges in scientific research. One key issue is discoverability; the platform is developing features to recommend relevant papers to users. They also aim to improve the publication of negative results and combat practices like p-value hacking, which can skew scientific findings. By providing tools for annotation and discussion, they encourage more rigorous evaluation and transparent reporting of research, particularly in fields with high publication rates like machine learning.

ENHANCING PAPER DISCOVERABILITY AND ACCESSIBILITY

A significant focus for Fermat's Library is improving how researchers find and access papers. Their Chrome extension for arXiv-hosted preprints adds annotation capabilities directly to the existing platform, making it easier for users to engage with new research. They recognize that arXiv is a critical hub, especially for fast-moving fields like machine learning, and their extension aims to add value by enabling comments and potentially a rating system. Future plans include a robust recommendation engine to help users navigate the vast scientific literature.

THE POTENTIAL OF ANNOTATING BOOKS AND TEXTBOOKS

The concept of annotating books, similar to academic papers, is a natural extension that Fermat's Library is considering. Patrick Collison's idea of paying for annotated books highlights a potential market for this. The founders believe that annotating introductory science books, like the classic 'Goldstein' for classical mechanics, could significantly improve learning by offering alternative explanations and clarifications. This approach could lead to a more dynamic and updated form of educational content, akin to open-source projects where specific sections can be improved and shared.

THE FUTURE OF SCIENTIFIC PUBLISHING AND PLATFORMS

Fermat's Library envisions a future where scientific publishing is more open, collaborative, and accessible. They actively support the emergence of overlay journals that leverage arXiv for hosting, reducing costs and paywalls. The platform aims to foster a culture where renowned scientists endorse open journals, validating them and encouraging broader adoption. While acknowledging the challenges and the slow pace of change in academia, they remain committed to building tools that empower researchers and accelerate scientific progress.

INSPIRING AND EMPOWERING LEARNERS

Beyond academia, Fermat's Library has found success in making complex knowledge accessible through engaging content, particularly on Twitter. They observed that learning, when presented effectively, can be highly addictive. By transforming intricate scientific concepts into easily digestible tweets and graphics, they provide a low-barrier entry point for public understanding. This approach not only educates but also inspires, as demonstrated by a fourteen-year-old who developed an alternative proof for a mathematical concept after engaging with their annotated paper.

BALANCING SIDE PROJECT WITH AMBITION

Fermat's Library operates as a passion project, with the founders dedicating their free time and resources. They emphasize that while it consumes significant effort, the enjoyment derived from reading, annotating, and sharing knowledge fuels their commitment. The platform started with a rapid prototyping approach, quickly releasing a functional version and then iteratively improving it. This dedication is driven by a genuine interest in science and a belief in the project's long-term impact, rather than immediate profitability.

ENCOURAGING CONTRIBUTION AND COMMUNITY GROWTH

The founders encourage anyone interested to contribute by annotating papers on Fermat's Library, sharing the platform with their peers, or using it within their university journal clubs or research groups. The platform is free to use for educational and research purposes. They also accept cryptocurrency donations. The core idea is to build a community around shared learning and annotation, fostering a collaborative environment where knowledge is readily exchanged and enhanced, ultimately contributing to the broader goals of open science.

Mentioned in This Episode

●Products

●Software & Apps

●Organizations

●Books

●Concepts

●People Referenced

Making Science Accessible: Key Takeaways

Practical takeaways from this episode

Do This

Annotate papers to clarify complex concepts and add context.

Embrace open science principles: share data, code, and publish in open-access journals.

Prioritize clear and accessible writing to maximize impact.

Leverage online platforms for collaboration and knowledge sharing.

Consider annotating books to enhance learning and understanding.

Be mindful of the potential for data misinterpretation (e.g., Simpson's Paradox).

Encourage the publication of negative results and address issues like p-value hacking.

Explore new metrics for evaluating scientific impact beyond citations.

Avoid This

Don't let valuable insights from reading or research remain unshared.

Don't rely solely on traditional, paywalled publishing models.

Don't assume authors are always the best annotators of their own work.

Don't neglect the potential for online collaboration to solve complex problems.

Don't dismiss the value of accessible, bite-sized scientific content (e.g., on social media).

Common Questions

Fermat's Library is a platform designed for annotating academic papers. It allows users to add comments and supplementary content directly alongside PDFs in a browser, aiming to make complex research more understandable and foster collaboration.

Topics

Mindset & Self-Improvement Technology & Innovation Science & Mathematics Data Interpretation Research Collaboration Scientific Communication Academic Research Scientific Publishing Knowledge Sharing Annotation Tools

Mentioned in this video

People

Terence Tao

A Fields Medalist mathematician who collaborated online to solve the Erdos discrepancy problem.

Freeman Dyson

Author of seminal papers on Dyson Spheres and Feynman's derivation of the Schrodinger equation, admired for his clear and concise writing.

Dan Carlin

Host of the Hardcore History podcast, cited as an example of successful long-form educational content with a large audience, similar to the appeal of learning through well-explained scientific concepts.

Charlie Munger

An investor and vice chairman of Berkshire Hathaway, whose non-technical paper was annotated on Fermat's Library, attracting a large audience.

Lawrence Lessig

Commented on the Bitcoin paper on Fermat's Library.

Tim Gowers

A Fields Medalist mathematician who started the Polymath Project to explore online collaboration in solving math problems and founded the open-access journal Discrete Analysis.

Paul Erdős

A famous mathematician who posed the Erdos discrepancy problem.

Software & Apps

arXiv

A preprint server where scientific papers are published before peer review, crucial for fast-moving fields like machine learning.

Quantum Journal

An overlay journal on arXiv for quantum physics, collaborating with Fermat's Library to integrate its annotation interface.

Fermat's Library

A platform for annotating academic papers, allowing users to add comments, explanations, and supporting content to PDFs within a browser interface.

Discrete Analysis

An open-access journal founded by Tim Gowers, showcasing a new model for scientific publishing that collaborates with arXiv.

Concepts

Erdos discrepancy

A mathematical problem posed by Paul Erdos, famously solved through online collaboration, highlighting the power of shared knowledge.

hot hand phenomenon

The belief in basketball that a player who has just made a shot is more likely to make the next one; debunked by analysis suggesting players are simply more consistent or that aggregate data can be misleading.

Dyson Sphere

A theoretical megastructure proposed by Freeman Dyson to capture a star's energy output.

Simpsons paradox

A statistical paradox where aggregate data can lead to a different conclusion than disaggregated data, used to explain phenomena like the hot hand fallacy.

Quantum Electrodynamics

A theory in physics, unified and made accessible by Freeman Dyson's paper, building upon independent work by Feynman, Schwinger, and Tomonaga.

open science

A movement aiming to make scientific research more accessible and collaborative through open data, open code, and open publishing (non-paywalled journals).

Media

Joe Rogan Experience

A podcast mentioned in relation to discussions about scientific publishing and data scrubbing.

Companies

Stack Overflow

A Q&A platform for programmers, serving as another example of successful online collaboration and knowledge sharing.

GitHub

A platform that has significantly enabled open-source collaboration, analogous to what Fermat's Library aims to achieve for scientific research.

Products

Fitbit

A wearable device used to track sleep, which led one of the speakers to research papers on sleep and mortality.

Organizations

Nature

A prestigious scientific journal, used as a benchmark for the reputation that open-access journals aim to achieve.

CERN

The European Organization for Nuclear Research, used as a VPN point for researchers to access paywalled scientific papers.

Wikipedia

A platform for general knowledge creation through collaborative, open editing, serving as a model for scientific collaboration platforms.

Science

A highly regarded scientific journal, mentioned as a benchmark for the reputation that open-access journals aspire to.

MIT

Massachusetts Institute of Technology, where access to research journals is often readily available within the campus network, highlighting disparities in access globally.

Books

Bitcoin paper

The foundational paper on Bitcoin, mentioned as the first paper annotated on Fermat's Library and remaining the most commented.

Ethereum White Paper

A foundational document for Ethereum, mentioned as an example where authors may not be the best annotators as they've internalized much of the information.

Feynman's derivation of Schrodinger equation

A paper by Freeman Dyson explaining Richard Feynman's approach to deriving the Schrodinger equation, noted for its clarity.