Newsletter #30: Games, Gaps, and Agents

What's been brewing this week?

Author

Raphael Kim, PhD
August 21, 2025

Minimalist logo design featuring the stylized letters “B” and “A” in bold black geometric lines, paired with the words “Biodesign Academy” in uppercase sans-serif font, set against a solid light aqua background to represent innovation and modernity in design education.

Dear reader,

This week, we step into the space where biology and design meet: a playful vocabulary game, new research on accessibility in AI tools, and a framework mapping how intelligent agents shape biomaterials. They invite us to think more fluidly, more deeply, across boundaries. Here’s what happens when fields begin to speak to each other.

TL;DR (30-Second Summary)

  • Design–Biology Vocabulary Match Game
    A browser-based tool matching 72 biology and design terms to build cross-disciplinary fluency, with immediate feedback and optional source code access for Foundational Tier members.

  • New Research on Accessible AI Tools in Biodesign
    Our peer-reviewed Mindtrek 2025 paper highlights accessibility gaps in tools like AlphaFold and PyMOL, offering practical strategies to support diverse learners and reframe accessibility as a core design principle.

  • Six Roles for AI Agents in Biomaterials
    Framework identifying six categories of AI agents—from generative design to lab automation—helping founders, students, and educators navigate emerging biodesign workflows and technical stacks.

Photograph of a minimalist study desk with neatly stacked biology textbooks, a black hardcover labeled “Biology,” and modern digital devices including a laptop, tablet, and smartphone displaying an interactive vocabulary learning app, illustrating the integration of traditional biology education with digital tools, presented by Biodesign Academy.

New Learning Tool: The Design–Biology Vocabulary Match Game

Want to get better at connecting biology and design concepts? We've created a vocabulary matching game that pairs 72 scientific terms with design analogies. It's browser-based, gives you immediate feedback, and might actually help you think more fluidly across both fields.

We first built this game while teaching a cross-disciplinary course at Biodesign Academy, after noticing how students struggled to translate biological ideas into design vocabulary.

Screenshot of an interactive vocabulary matching game titled “Design–Biology Vocabulary Match,” showing biology terms such as Signal Peptide, Chemotaxis, Metabolite, Gradient Descent, Toxicity, and Affinity on the left, paired with design analogies like Chemical-directed movement, Failure side-effect, and Binding strength on the right, with tracking of attempts, accuracy, and progress, created by Biodesign Academy.

The idea is straightforward: matching biological concepts with design thinking helps you build the kind of cross-disciplinary vocabulary that's useful when you're working at the intersection of science and creativity. Plus, you'll know right away whether you're getting the connections right.

What you get:

  • Practice with core vocabulary from both domains

  • Better instincts for cross-disciplinary collaboration

  • Immediate feedback as you work through the matches

We’ve tested this tool with our selection of subscribers and it had revealed surprising knowledge gap at the intersection of biology and design.

You can play it directly in your browser—nothing to download. If you're a Foundational Tier member, you can also grab the source code to use or develop further for your own teaching.

It’s not a replacement for full coursework of course, but it’s a playful entry point for building fluency.

A modern classroom scene shows students seated at desks with laptops while instructors in lab coats present complex 3D molecular visualizations on large digital screens, illustrating protein structures and biochemical interactions in a collaborative teaching environment, created for advanced biotechnology and design education at Biodesign Academy.

New Paper: Making AI Tools in Biodesign More Accessible to Learn, Teach, and Trust

As molecular and protein design become central to biodesign practice, many students and educators are turning to tools like AlphaFold, ColabFold, ESMFold, and PyMOL. But these tools often assume technical fluency, sighted users, and a background in computational biology.

What if you're teaching a class where some students rely on screen readers?
What if your curriculum includes non-coders, neurodivergent learners, or designers just beginning to explore molecular sketching?

Our newly accepted Academic Mindtrek 2025 paper—Inclusive Molecular Sketching: Accessibility Barriers in AI-Driven Protein Biodesign Workflows—offers insight into why these tools exclude, how they can be reimagined, and what inclusive learning environments need.

This research, led by Raphael Kim and Jiwei Zhou, was peer-reviewed and accepted for the upcoming Mindtrek conference in October 2025.

Key takeaways of the paper for biodesign students and educators:

  • Understand the hidden assumptions baked into today's leading molecular design tools

  • Identify where POUR (Perceivable, Operable, Understandable, Robust) guidelines fall short for AI-native systems

  • Access practical design and teaching strategies for supporting diverse learners in biodesign

  • Reframe accessibility as a core design principle, not an afterthought

If you’re building biodesign syllabi, mentoring interdisciplinary students, or learning these tools yourself—this paper offers a roadmap for making these platforms more learnable, navigable, and trustworthy.

Foundational Tier Members can read the full pre-print now, ahead of the October conference. Not a member yet? Join here to access this and other field-defining research, early.

A robotic arm equipped with precision tools manipulates glass beakers and flasks filled with colorful chemical solutions in a brightly lit laboratory, showcasing automation and AI-driven experimentation in biotechnology research, developed as part of advanced scientific training at Biodesign Academy

Six Roles for AI Agents in Biomaterials and Biofabrication

We've been tracking how AI agents are actually being used in protein design and biomaterials work.

By AI agents, we mean software systems that can perform specific tasks with some degree of autonomy—whether that's generating designs, analyzing data, or controlling lab equipment. What we're seeing is that most of these agents tend to fall into six fairly clear categories:

  1. Generative Design Agents – create new protein sequences or structures

  2. Predictive Modeling Agents – forecast folding, stability, or material performance

  3. Data Parsing Agents – mine literature, patents, or datasets

  4. Simulation & Digital Twin Agents – model folding, assembly, or mechanical behaviour

  5. Lab Automation Agents – run experiments, control robots, analyze results

  6. Workflow Orchestration Systems – combine several agents into end-to-end loops

We put together an overview of these categories that might be useful:

An artistic illustration shows a humanoid robot with a sleek white design extending its hand toward a golden molecular chain, surrounded by flowing abstract lines symbolizing data and connectivity, representing the intersection of artificial intelligence and biotechnology research, created as part of the Biodesign Academy visual playbook.

This framework is part of our Protein Futures project, which is mainly focused on founders and CTOs building protein engineering and biomaterials companies. The idea is to cut through some of the hype and make it easier to compare tools and think about technical stacks.

The categories aren't just for startups though. Students might find them helpful as a way to think about where AI fits into their own biodesign practice, from literature reviews through to actual design work.

And educators could use them to structure workshops or projects around AI-driven workflows—giving students vocabulary that connects to what's happening in industry.

If this kind of content is useful to you: weekly strategy notes, case studies, and developments in AI for protein engineering—we have a separate newsletter for that:

Each tool and insight we share is a quiet call to reimagine biodesign as both precise and poetic. Thanks for exploring with us, where science meets imagination, and learning becomes a creative act.

Until next week,
Raphael, Biodesign Academy

Oil painting–style scene of a dimly lit laboratory desk, featuring a vintage microscope, glass bottles, a large closed book, and a wooden rack of test tubes, all bathed in warm golden light streaming through a curtained window with the sun low on the horizon, evoking scientific curiosity and discovery, connected with Biodesign Academy.

Start learning AI in 2025

Keeping up with AI is hard – we get it!

That’s why over 1M professionals read Superhuman AI to stay ahead.

  • Get daily AI news, tools, and tutorials

  • Learn new AI skills you can use at work in 3 mins a day

  • Become 10X more productive

Frequently Asked Questions

Q: What is the Design–Biology Vocabulary Match Game?
A browser-based learning tool that pairs 72 biology and design terms, helping learners build cross-disciplinary fluency with immediate feedback. Foundational Tier members can also access the source code for teaching and customization.

Q: Why is accessibility in molecular design tools important?
Many tools like AlphaFold and PyMOL assume coding fluency and sighted use, excluding non-coders and learners with diverse needs. Our peer-reviewed Mindtrek 2025 paper provides strategies to make biodesign tools more inclusive, following POUR guidelines and emphasizing accessibility as a design principle.

Q: What are the six roles of AI agents in biomaterials and biodesign?
AI agents generally fall into six categories:

  1. Generative Design Agents

  2. Predictive Modeling Agents

  3. Data Parsing Agents

  4. Simulation & Digital Twin Agents

  5. Lab Automation Agents

  6. Workflow Orchestration Systems

Q: Who benefits from these resources and frameworks?

  • Students: Gain fluency in biodesign vocabulary and AI workflows.

  • Educators: Use structured frameworks to design accessible teaching materials.

  • Founders & CTOs: Compare technical stacks and apply AI agents to biomaterials innovation.