About AIxponential

Building a more equitable future through AI education, trustworthiness, and global collaboration.

Disclaimer: AI was used in the creation of this article with a human in the loop

Our Story

AIxponential was founded on the belief that artificial intelligence should be a force for equity, knowledge, and truth. Our mission is to ensure that AI technology benefits everyone, especially in education.

At our core, we believe the true promise of AI is the democratization of expertise. For the first time in history, Subject Matter Experts in any field have the power to leverage the full breadth of software and code. We are moving toward a world where the ability to build and create is limited only by our ability to imagine and describe.

We work to bridge the gap between AI innovation and practical, ethical applications in classrooms and communities worldwide. Through research, tools, and education, we're creating pathways for AI to enhance human potential rather than replace it.

We navigate this rapidly changing landscape with radical humility, recognizing that progress requires us to admit when we are wrong and to relentlessly pursue "better", "wiser", and "fairer" solutions.

Our vision is a world where AI literacy is universal, where educators are empowered with AI tools that align with their values, and where students develop critical thinking skills alongside AI assistance.

The Mission

The Democratization of Expertise

The current explosion of Artificial Intelligence is often viewed through a narrow lens of market capital and automated efficiency. At AIxponential, we believe the true promise of AI is far more profound: it is the democratization of expertise.

For the first time in history, Subject Matter Experts (SMEs) in any field - from history to horticulture - have the power to leverage the full breadth of software and code. We are moving toward a world where the ability to build and create is limited only by our ability to imagine and describe.

Navigating the "Peak of Mount Stupid"

In my opinion, we are in the middle of a rapidly changing technology environment with respect to AI. Advances are occurring so rapidly that knowledge that was state of the art several months ago becomes obsolete. Experiences with tools from several months ago are no longer relevant as rapid iteration of capabilities have moved the dial and become faster, better, and easier.

What this means is that as a society, we are currently navigating the Dunning-Kruger effect at scale. All of us, myself included are frequently finding ourselves at the "Peak of Mount Stupid" - that precarious point where we know just enough to be dangerous, but not enough to be truly wise.

We invest time into obtaining greater knowledge and find that after a month, those learnings and experiences have become obsolete and we start yet again on a revised and updated journey. It's scary and exciting in the same breath, but looking beyond software development, the landscape is littered with gems, detritus, and slop.

Peak of Mount Stupid - Dunning-Kruger Effect

Navigating this landscape is daunting, so we see all the typical behaviors observed in periods of rapid change... denial, mania, anxiety, anger... If we want to find some historical parallels, my favorite is described by Dan Carlin in his podcast "Prophets of Doom". The period described is the circumstances after Martin Luther unleashes Pandora's Box which coincided with the Printing Press.

For some people, AI has thrown a cat into the pigeons, for others it is the best thing since sliced bread, and if you believe the hype it will lead to the obsolescence of the knowledge worker. These all maybe right, but from my perspective the most important goal needs to be solving problems like human suffering, human happiness, creating an equitable society, and preserving stability.

We all may differ on the relative priority of each of these, but considering the world around us, these seem to absorb most of the time, money, and discourse of our society. That's a long-winded way of saying what should be the goal of using AI, and that is what I formed AIxponential to explore.

At AIxponential, we approach this with radical humility, recognizing that progress requires us to admit when we are wrong and to relentlessly pursue "better", "wiser", and "fairer".

Empowering the Architects of Learning

Part of the circumstances we are in now is due to the massive economic investment in AI. One side effect: it prioritizes "black box," packaged delivery vehicles designed for monetization. This creates a sales urgency looming over the potential of the technology.

A key area of investment is the people who are effectively providing the picks and shovels of the workforce that will be needed in the Post-Knowledge Era. I am referring to the Instructional Designers, Teachers, Educational institutions, Trainers, and Designers who will be helping sort through the signal to noise and inspiring the vision of the people who will be using the technology.

Architects of Learning - Educators collaborating

Our educators will need to help us navigate through the complex path of what the state of knowledge at a given time asserts to be the best, but it's important to remember that in hindsight we are going to reflect on some of these recommendations in the same way that we look back at Radium tonic, Mercury for Teething powders, Lobotomies, Bloodletting, and Tobacco Smoke Enemas. There will be mistakes but doing nothing can also be a mistake.

Our Approach

AIxponential operates on a decentralized, collaborative model inspired by the principles of Maker Spaces and the open-source ethos. We believe in the power of community-driven innovation and voluntary contributions.

Unlike traditional organizations, we don't have rigid hierarchies or fixed roles. Instead, we foster an environment where educators, technologists, researchers, and advocates can contribute their unique skills and perspectives to shared goals. This organic and collaborative approach encourages experimentation and empowers individuals. More than any other goal, we believe fostering a joy of learning is key to innovative solutions.

This approach allows us to be nimble and experimental. We believe in the power of diverse viewpoints which means:

Open collaboration across disciplines and backgrounds
Practical solutions grounded in real-world needs
Transparent processes and ethical considerations
Equitable access to AI education and tools
Radical humility in admitting mistakes and pursuing 'better, wiser, fairer' solutions

Automated Abstention

We develop open standards for AI systems that use confidence scoring to simply say "I don't know" when they can't meet a threshold of accuracy.

This honest acknowledgment of limitations is crucial for educational applications where misinformation can have lasting impacts. Agency over the criteria of "imagination" is also essential. Knowing which is being used and its approach is back to transparency and trust.

Neuro-Symbolic Integration

Using techniques like Model Context Protocol (MCP), we combine the creative power of large language models with the rigid, logical truth of traditional code. APIs are another example, except the interface details can be made common for typical situations.

This hybrid approach delivers both flexibility and reliability - essential for educational tools.

Open Infrastructure

We champion open-access architectures that allow educators to understand and control the AI tools they use, rather than being locked into proprietary black boxes.

Transparency empowers educators to make informed decisions about when and how to use AI in their classrooms.

Knowledge Graphs

We partner with initiatives like the Learning Commons Knowledge Graph, Quill.org, and AIEdu.org to enable AI that references structured, verifiable knowledge.

This approach provides an alternative to relying solely on the hidden weights of commercial models.

AI in Education

The AI in Education initiative aims to transform education by thoughtfully integrating AI technologies in ways that enhance human teaching and learning rather than replacing them. We work at the intersection of education, technology, and ethics to create tools, frameworks, and resources that empower educators and learners to navigate the AI revolution.

Two-Pronged Approach

Our innovative two-pronged approach to AI integration in education:

1. Tool-First Perspective

Treats AI as a powerful tool for problem-solving, emphasizing how AI can augment reasoning and expedite work while still requiring human oversight and interpretation of results.

2. Learning-First Perspective

Focuses on how understanding can be enhanced by exploring AI's limitations and capabilities, using AI errors and reasoning patterns as opportunities for deeper engagement.

Prompting Course

Our comprehensive course on effective prompting techniques helps educators and students master the art of communicating with AI systems to achieve better results. Learn how to craft effective prompts for different educational contexts and integrate AI tools meaningfully into your teaching practice.

Access the full course

AI Trustworthiness

The AI Trustworthiness initiative specializes in unbiased AI evaluation, with a particular emphasis on Large Language Models (LLMs). As these powerful systems become increasingly integrated into critical aspects of society, ensuring their trustworthiness through objective, rigorous evaluation becomes essential. We are committed to developing methodologies that go beyond surface-level metrics to assess the true capabilities, limitations, and potential impacts of AI systems across diverse contexts and user populations.

Key Evaluation Dimensions

Our unbiased evaluation frameworks examine LLMs across multiple critical dimensions:

Factual Accuracy: Assessing whether model outputs contain verifiable truths
Fairness & Bias: Measuring representation across demographics and topics
Robustness: Testing performance across different phrasings and contexts
Safety: Evaluating resistance to generating harmful content
Alignment: Determining adherence to human values and ethical principles

Why Unbiased Evaluation Matters

While many LLM evaluations exist, most suffer from significant limitations:

Over-focus on English and Western perspectives, neglecting multilingual and multicultural dimensions
Narrow benchmark datasets that don't reflect real-world complexity
Models optimized for specific benchmarks rather than real-world performance
Evaluation methodologies that aren't fully transparent or reproducible

Ready to join the mission?

Help us build transparent, reliable AI tools that empower educators and learners.

Get Involved Contact Us