Beyond DeSci Part 3: From Concept to Code

After two installments where I explored the theoretical foundations of an aspirational modern scientific trust architecture, I’m pleased to share tangible progress: SciValidate has evolved from concept to functional working code, with a sample interface now available at scivalidate.org/example. I need to give credit where credit is due—my coding skills are dated, and development needs working knowledge of at least four different computer languages. Without Anthropic’s Claude AI, this would be much further behind.

From Theory to Implementation

The sample demonstrates the core concept: replacing binary social media metrics (“likes”) with meaningful indicators of scientific validity. I’ve populated a test database with public data from a single academic department, using ORCID for identity tracking and OpenAlex for keywords and coauthor networks, automatically inferring fields of expertise and reputation from language analysis.

This serves as a seed database — a starting point that infers reputation from existing metrics to enable the transition to something more profound. The current implementation offers a view “under the hood” where faculty members and their publication history have been analyzed using natural language processing algorithms to determine reputation scores (0-10) in esoteric scientific fields and begin to describe a network of scientific expertise.

I would like to point out that this academic-based approach is only scaffolding. SciValidate must transcend traditional academic metrics that can be (and routinely are) gamed. Publication counts or citation indices don’t determine actual scientific reputation. Instead, we must measure how effectively knowledge is communicated, verified, and applied. In the final system, reputation will be judged primarily by readers and knowledge-seekers, not just academic peers who benefit from the same incentive structures.

For transparency and collaborative development, all code is now available on GitHub:

github.com/jburbs/scivalidate

Technical Challenges Worth Solving

Several non-trivial challenges have emerged during implementation:

1. Scaling the seed network

Starting with just Rensselaer Polytechnic Institute’s Chemistry Department (14 faculty with publications and unambiguous ORCIDs), the database quickly expanded to over 5,000 potential participants across multiple institutions. This exponential growth highlights the challenge and opportunity of building a comprehensive scientific trust network. [If you, or someone you know, is enamored with “big data” or “natural language processing”, this is a treasure trove.]

2. Platform integration

For SciValidate to achieve its purpose, verification badges must integrate with existing platforms. Initial exploration revealed significant technical hurdles: API limitations restrict what third-party applications can do, each platform has different technical requirements, and social media companies may resist validation systems that don’t align with engagement-based business models or require more transparency than is comfortable. At the same time, these platforms are all under fire for promoting disinformation, so they have an incentive to outsource curation (consider Wikipedia if you don’t think social curation is practical!).

3. The reputation paradox

The most challenging question is: How can reputation be measured objectively?

Academic scientists already monetize their publication records in countless ways, from grant applications to promotion packages. SciValidate cannot simply become another h-index equivalent to keep score in a game. Our algorithm currently infers reputation from public information as a necessary starting point, but this cannot be the endgame.

The system must evolve to recognize those who excel at science’s essential skills: verification, communication, and constructive debate. Unlike traditional academic metrics, which primarily reflect peer judgment, genuine scientific reputation is ultimately determined by knowledge consumers who benefit from clear, honest, and accessible scientific communication.

Why This Matters Now

Recent data shows that over 50% of paper retractions in 2022 were attributed to deliberate fraud—triple the average between 1988 and 2018. This industrialization of fraudulent research threatens to create a catastrophic feedback loop as AI systems train on scientific texts presumed to be human-authored.

From someone who has spent decades in and around scientific research, this is more than just a troubling development. The integrity of scientific discourse—the foundation of humanity’s collective problem-solving capability—is at stake.

Moving Development to GitHub: A Call for Collective Action

Neither LinkedIn nor Substack is the right platform for what is essentially a software development project: GitHub will serve as our primary collaboration space for technical development. The repository at github.com/jburbs/scivalidate includes:

• Data collection tools for gathering researcher information
• Database components for organizing profiles and metrics
• Analysis modules for calculating reputation scores
• A functional web interface with verification badges.

While GitHub excels as a technical collaboration platform, its interface can be intimidating for those unfamiliar with software development workflows. Here’s how you can contribute:

For developers and technically-inclined contributors:

• Fork the repository directly on GitHub
• Submit pull requests for code improvements or bug fixes
• Use GitHub Issues to highlight specific technical challenges.

For scientists and non-technical contributors:

• GitHub’s “Issues” feature is accessible even without coding knowledge
• Create a free GitHub account and navigate to our Issues page
• Click “New Issue” and select from our templates for different feedback types
• Tag your issue with relevant categories (e.g., #reputation-metrics, #data-integration)

For all interested in following the progress:

• Look at the mockup at scivalidate.org/mock, the interface at scivalidate.org/example, and star the GitHub repository to receive update notifications. Contribute your insights.
• Share this initiative with colleagues who might contribute valuable expertise.

My goal is to build a community of problem-solvers across disciplines. The GitHub repository represents a transition from conceptual discussion to practical implementation, where your contributions directly shape the system’s development.

The problems of scientific trust online are too essential to abandon because the path forward is difficult. By sharing both the vision and implementation as open source, I aim to attract collaborators who can help turn this concept into reality.

The problems of scientific trust online deserve an engineering solution, not just a theoretical discussion. The scientific method teaches us that empiricism trumps pure theorizing—hypotheses must be tested against reality.

I’ve built the prototype of a working seed rather than writing another essay on potential approaches. I’ve moved from commentary to code, from conceptualization to concrete implementation. Join me in testing, refining, and building this vision into reality.

The interface is clunky, the design needs refinement, and challenges remain significant—but progress begins with functional implementations, not perfect theories.

“It doesn’t matter how beautiful your theory is, it doesn’t matter how smart you are. If it doesn’t agree with experiment, it’s wrong.” – Richard Feynman.

Which challenge would you tackle first? Your technical expertise, subject matter knowledge, or even thoughtful criticism could help reshape how scientific truth propagates through our information ecosystem.

#ScienceIntegrity #OpenSource #ScientificCommunication #Research