ModuleQ’s Mission: AI that Advocates for Humanity
Human-centered AI augmentation for a new world of hybrid work
By: David Brunner, PhD and Anupriya Ankolekar, PhD
Artificial Intelligence has enormous potential to make our lives and societies better, freeing humans from drudgery and expanding our horizons. Yet too often, AI ends up working against us. AI can mislead and manipulate us, distract us away from our goals, waste our precious time and mental energy, and intensify our stress and anxiety. At work, AI is sometimes used to monitor and micromanage us, or automate us out of the picture entirely. As Shoshana Zuboff explains in The Age of Surveillance Capitalism, this unhappy “instrumentarian” state of affairs is a natural consequence of AI designed to advance others’ interests at your expense.
We founded ModuleQ to create AI that advocates for you. From the start, we rejected advertising-based business models, because they create a fundamental conflict between users’ interests and the financial success of the company. For ModuleQ’s People-Facing AI, you are not a means to an end. You are not an eyeball to be delivered to an advertiser. You are not an asset to be optimized, nor an employee to be “nudged.”.
You have goals that you are seeking to achieve, and ModuleQ’s AI is designed with the sole purpose of helping you achieve them. As founders, this is our commitment to you. Our commitment pervades the design of our technology, from structural safeguards to protect your privacy to how we measure our success. For example, we focus on real business outcomes that matter to you, such as customer meetings and revenue opportunities, and on your explicit feedback about the usefulness of the insights we deliver. And as we have built our business, we have embedded this commitment deeply into the fabric of our company by assembling a community of like-minded investors, board members, employees, and partners.
On our journey to create “AI that advocates for you,” we encountered a formidable challenge. To help you achieve your goals, People-Facing AI must first understand your goals. That’s difficult because your goals are complex and change over time. Moreover, while rudimentary understanding is sufficient for recommending advertisements or similar products, the level of understanding required for effective human-AI collaboration is much higher. Our solution is a unique invention that we call Personal Data Fusion. The product of decades of research, Personal Data Fusion uses your work data—emails, calendar meetings, contacts, etc.—to create a digital twin of you and your work.
Your digital twin is like a reflection of your professional self into the digital world. It is sufficiently accurate and detailed to be recognizable, but really just a very shallow representation compared to all the complexity of your mind in the real world. Although your digital twin is relatively simple and shallow, it enables ModuleQ’s People-Facing AI to understand your work context. Like an ever-diligent assistant, the AI knows about the business relationships and topics that you care about most. In fact, our highly refined algorithms are so attentive, they can predict the information that will be useful with up to 90% accuracy.
People-Facing AI is fundamentally different from apps that recommend books, movies, music, and dance videos. Those consumer apps use statistical data from millions or billions of users to make their services as addictive as possible. Statistical models work, but they don’t really understand anything about you. They are only as good as the data they learn from, making them brittle and prone to fail simple common-sense tests. That is why such AI can be tricked into thinking, for example, that a cat is guacamole.
In contrast to statistical engines that predict your behavior based on the behavior of other people, People-Facing AI actually reasons about you and your professional context from your work data. Of course, with access to this extremely sensitive data comes great responsibility. The best way to keep your data safe is to keep it inside the environment where it has already been secured. That’s why ModuleQ brings our AI to your data: People-Facing AI is distributed to our customers for behind-the-firewall operation.
Security alone is not enough. Since your digital twin mirrors your work mind, it must be protected from prying eyes and micromanaging bosses. Of course, as professionals, we understand that our work-related information is subject to examination under certain circumstances, but we don’t expect colleagues to barge into our offices and go through our half-finished drafts. This privacy must be extended to the digital world as well.
Privacy is good for business. Research by Harvard Business School Professor and ModuleQ Board Member Ethan Bernstein has shown that too much transparency is counterproductive, reducing innovation, impeding learning, and harming productivity. ModuleQ AI is designed to reveal only aggregate information about users, giving professionals the space they need to be creative and experiment with new ideas.
As founders, our commitment to helping people work smarter with AI has been a consistent theme throughout our professional careers.
As a child growing up in Silicon Valley, I was an avid consumer of science fiction and an eager student of new ideas and technologies, from lasers, holograms, and spectroscopy to fractals, chaos theory, and strange attractors, to computers and artificial intelligence. Above all, I was excited about the potential of technology to expand the horizons of humanity. I went to Stanford for my undergraduate education, where I majored in Computer Science.
At Stanford, I had the extraordinary good fortune to study under Professor Edward Feigenbaum, a pioneering figure in the field of Artificial Intelligence, Turing Award recipient, and former Chief Scientist for the Air Force. Professor Feigenbaum opened my eyes to the power of AI as a means of amplifying human intelligence.
Around the same time that I started studying with Professor Feigenbaum, I happened to be taking a class in ancient Etruscan art with Professor Jody Maxmin. Hearing about my research, she recommended a book by her sister-in-law, Harvard Business School Professor Emerita Shoshanna Zuboff: In the Age of the Smart Machine.
Professor Zuboff analyzed how computers were being used in different industries, and came to a startling conclusion. In most business settings, computers can be used to augment human creativity and enhance the role of humans in the production process, or, in the exact same business contexts, computers can be used for coercion and surveillance in ways that reduce human autonomy and diminish their value. From Professor Zuboff’s research, I understood the crucial importance of software companies committed to human augmentation and enhancement, and I resolved to create such a company.
My next step was to learn the basics of business by joining the Boston Consulting Group, first in Tokyo (I had learned Japanese at Stanford and as a visiting scholar at Shizuoka University) and then in San Francisco. Then, to better understand how AI and humans could work collaboratively together, I entered a joint business-computer science PhD program at Harvard University.
At Harvard, I was advised by Professor Marco Iansiti, a leading expert on digital transformation and Competing in the Age of AI. Marco now serves on the Board of ModuleQ. My PhD committee chair was Professor Mike Tushman, one of the foremost authorities in the fields of organization design and disruptive innovation.
In the course of my doctoral studies, it became clear to me that only by effectively dividing work between humans and computers will we be able to overcome information overload, the defining problem of the digital age. Humans excel at creative problem solving and strategic thinking, but cannot keep up with the deluge of digital data. AI excels at finding patterns in vast oceans of data, but cannot match the creative and strategic abilities of humans. Could AI tame the digital deluge, scouring the digital ocean on our behalf, and allow us humans to focus our mental energy on the most important and meaningful information?
This is the human-AI symbiosis that we are creating at ModuleQ.
From early on, I’ve been fascinated by human work and technology. Growing up with a technophile father who was also a business school professor, I was surrounded by dinnertime stories and theories of organizational change, human decision-making, and the role of technology in enabling nimble organizations. A compulsive reader in childhood, I was inspired by the humanism inherent in global literature and poetry, but also had an avid appetite for science fiction tales of technological promise and dysfunctional societies. These early influences convinced me of the positive, transformative role of technology in complementing humans and enabling human creativity, but also made me sensitive to the challenges of introducing technology into organizations and human society.
After an undergraduate and Master’s degree in Computer Science at RWTH Aachen, I joined a doctoral program in Human-Computer Interaction at Carnegie Mellon University (CMU) in 2000. There, I had the immense good fortune of attending a class by Herbert Simon, the visionary Nobel Prize, and Turing award winner, and was deeply influenced by his theory of the cognitive processes underlying human problem-solving. It began to dawn on me that despite the complementary abilities of AI and humans, to effectively aid humans, AI must support human cognitive processes. To explore what such an AI might look like, my doctoral work built a recommendation system to support software developers in open-source projects, by pulling together a curated set of bug reports, discussions, code commits, and peer developers for the problem they were working on. While at CMU, I also delved into intelligent agents, semantic technologies, and web services, but soon realized that the state of technology to extract meaning from human work traces on the web was not mature enough to enable the kind of human-centric AI I was envisioning.
In 2005, I joined the Karlsruhe Institute of Technology in Germany as an assistant professor to work on semantic AI and intelligent web services that could better represent and handle the semantics of information on the web. I worked on explicit knowledge modeling, using knowledge bases and data graphs to represent knowledge on the Web, and within organizations. Despite many significant advances in semantic technologies, natural language processing, and AI, I noticed that the benefits to organizations and humans were limited. There appeared to be a ‘last-mile’ problem: without being embedded in human workflows and contextualized to be relevant to users, information was not being effectively utilized. It wasn’t making a real difference to human work.
To better understand the information workflows of professionals, I moved to HP Labs in Palo Alto. During my nine years there, I observed knowledge work within a large enterprise; and designed, built and deployed technology prototypes and information markets to support knowledge professionals. I began to understand how the organization plays a key role in creating the right environment for creative human work and technology augmentation.
On meeting David, I realized we shared a vision for human-centered AI that augments human cognition and creativity. Within organizations, this kind of AI could be transformative because it could understand and support human work, deliver business-critical information and resources proactively and bring more relevant resources to bear on larger organizational issues.
Amplifying Serendipity, Catalyzing Creativity
Using AI to surface timely, relevant insights is not just about saving time and increasing efficiency. Receiving the right information at the right moment opens our eyes to new opportunities. AI amplifies serendipity, giving us more chances to make meaningful connections and unexpected discoveries.
To realize our vision for empathetic AI that puts humans first, we needed a patient investor committed to making the world better through technology. We found Taizo Son, the serial entrepreneur and social impact investor who helped launch Yahoo! Japan. Our vision for People-Facing AI reminded Taizo of Yahoo! founder Jerry Yang, who told Taizo that Yahoo! would be “an apple for a future Newton.” By connecting people with information they would never find otherwise, Yahoo! would surprise and inspire them. Taizo invested in ModuleQ through Mistletoe, his investment organization dedicated to creating a “sustainable human-centered future using technology.”
Yahoo! and other search engines put the world’s information at our fingertips. As the digital ocean has grown by orders of magnitude, search progressed from manual indexing to sophisticated algorithms. Yet digital information has grown so fast that humans simply cannot keep up. Somewhere out there in the digital ocean are crucial insights that can transform our work and our world, but we’re not even aware they exist.
ModuleQ turns the search paradigm on its head. Instead of placing the burden on humans to figure out when to search and what to search for, People-Facing AI searches tirelessly on your behalf. The AI is always looking for the game-changing insights that you’re most likely to find useful, eye-opening, and inspiring.
In our new world of remote and hybrid work, People-Facing AI is more important than ever. The pre-pandemic, industrial model of 9-to-5 office life has been eclipsed by more flexible and inclusive models of online collaboration. Although in-person work will surely endure to some degree, we can no longer rely on lunchtime discussions and hallway chats to keep us in the loop. We need AI to surface information when and where it is needed, connecting the dots across teams and organizational siloes.
Fortunately, our new online workspaces provide the ideal environment for human-AI collaboration. AI can interact with us in these workspaces much like a virtual colleague, listening to your conversations and chiming in with messages and updates. Of course, with AI working alongside us, empathy and ethics are paramount. No one wants disruptive, distracting, disloyal, or destructive coworkers, be they human or AI. We have every right to hold our AI collaborators to the same standards that we hold our human colleagues.
We are still in the early days of Human-AI collaboration. ModuleQ is committed to pioneering AI that advocates for you because we believe that a human-centered future must and will prevail.