Marc Vontobel is the CEO & Founder of Starmind, a platform that enables team members to connect to expert knowledge from anywhere in the organization they operate in, whenever they need it.
You’ve previously stated no one really understands what AI is, because no one really understands what intelligence is. Have your views changed, if not could you elaborate on this statement?
There is no official definition for intelligence, and depending on the context, you will get different answers. Is a mathematician intelligent? Yes. Is a cat intelligent? Probably. Is a tree intelligent? It’s debatable. However, a tree survives for hundreds of years. It takes nutrients from the soil, turns light into energy, and moves towards sunlight if it’s not getting enough. Therefore, it seems logical that a tree would be intelligent to some extent, right?
Because everyone views intelligence differently, people also have difficulties understanding AI and the limitations of machine learning. People believe a computer can easily become as intelligent as a human, or possibly, a tree. This is not the case.
A computer follows orders, which can trigger more complex algorithms than we can imagine. In the end, a computer never makes a decision based on the belief that it’s the right thing to do. Its decisions are derived from the inputs and algorithms.
Therefore, I believe no one really understands AI because no one fully understands or can completely define intelligence.
Could you share the genesis story behind Starmind?
Eleven years ago, we developed a platform that would enable scientists to share knowledge with one another. Instead of spending hours researching a problem, a specialist in the topic of research could access the Starmind solution to ask a question. Although the platform was well received, there was an imbalance in our target audience between users interested in providing knowledge and experience and users willing to pay a small fee to have their questions answered.
In 2013, the CIO of a large Swiss bank approached us and revealed that he had numerous questions daily, but no ability to get them answered. He asked if our technology could be applied to organizations, enabling people to get the right information from the right person while eliminating the time-consuming search effort. It inspired us to think about how large organizations handle their knowledge.
Eighty percent of knowledge resides in people’s heads. We realized that every large organization faces the exact same problem. Twenty percent of an organization’s knowledge is documented somewhere in different tools. Much of what is documented is outdated, hard to find, and once found, difficult to put into a new context. People are left with two options: Spend most of each day searching and consolidating information on their own to do their jobs or ask a colleague and hope they have the right answer.
It was then we realized there wasn’t a single technology that served as a primary source of information within an organization. We created a solution that enables every employee to tap on the virtual shoulder of the right colleague when needed, regardless of whether they know the person or not. We wanted employees to access it in a faster, more inclusive and collaborative way.
How does Starmind connect thousands of brains, to create a corporate brain?
Rather than connecting “brains”, we prefer to say that we break down knowledge barriers and connect “people” for greater knowledge collaboration across the entire organization. Our platform helps employees connect to expert knowledge from anywhere in the organization and when needed. Rather than relying on people within one’s own department or network, employees can anonymously pose a question through the knowledge collaboration platform. Our artificial intelligence can identify those within an organization who are best equipped to answer the question regardless of hierarchy or tenure. Within an hour of the query, the employee will receive a response. Answers are validated and ranked. The AI learns from these interactions by connecting terms or topics to identify how to map similar information in the future to identify who holds the knowledge as well as how data is “recycled” to make sense of it. In short, our platform breaks down knowledge barriers and information access within an organization. Our technology doesn’t replace human intelligence. It enhances it, making it more accessible and relevant than ever.
How does Starmind single out the most intelligent people in an organization that can best answer questions?
Over time, people share their knowledge and insight by addressing their colleagues’ questions. As employees rank responses, our AI becomes more competent at validating and recognizing who knows what in real-time. It learns from these interactions by connecting terms and topics to identify how to map similar information in the future. The platform not only identifies those who hold the knowledge but also recycles data in a way that makes sense.
Why is this solution superior to using deep learning or other types of AI to identify an answer?
AI and natural language processing certainly facilitate this process, but a more nuanced yet crucial component at play is human experience and contextualization. Eighty percent of knowledge resides within people’s heads, with only 20 percent documented. The documented information likely is outdated, inaccurate or redundant. Human experience can identify or contextualize data inconsistencies through knowledge sharing and collaboration across the organization. Consequently, it builds a rich, constantly evolving repository for employees to draw upon, consider and apply accordingly.
Could you give an example of a corporate use case for using Starmind?
A top global pharmaceutical company uses Starmind. Every month, they bring new drugs to the market, providing life-saving treatment to people worldwide. In addition, they must react to new viruses and strands like influenza and Covid. In these cases, time is of the essence, and they need to accomplish research and development with precision to bring treatments to the market as soon as possible. Developing these drugs requires knowledge and experience across multiple experts. Finding the right people can be challenging when your company is 125,000 strong across numerous countries and why they choose Starmind. It enables them to access expert knowledge quickly, connect people faster, and aid time-to-market speed.
What type of feedback have you received from the Starmind application?
We have been very well received throughout our industry and among our clients. Clients have noted that our platform helped them respond to customer inquiries quickly and effectively. It enables the onboarding of new employees faster, keeping them engaged. It ensures that technology is more accessible to all employees in the organization, fostering better collaboration innovation. Finally, it boosts productivity.
Are future developments in AI a threat to the Starmind business model?
I don’t think so. Even though AI and natural language processing can understand concepts and contextualize what people have written down, they don’t have access to the tacit knowledge that resides in people’s experiences and heads. Starmind encourages tacit knowledge to be shared. When written and captured, AI and natural language processing can contextualize the information that employees can access and use when needed. However, it requires a platform like Starmind’s to encourage and share that tacit knowledge.
AI potentially could automatically answer questions that someone has not previously answered. However, the AI response will always be based upon information provided by a person.
One major, unsolved problem is determining if particular data is still relevant, outdated, or correct. Just like humans, AI struggles to look at information and decipher if it is still valid or not.
Therefore, I believe human experience and contextualization of data will always be superior to AI.
Is there anything else that you would like to share about Starmind?
Knowledge is essential to organizations. When effectively harnessed, it can positively impact the way you work.
Imagine trying to travel from Paris, France to Stuttgart, Germany. You don’t know your way, and GPS does not exist yet. You have a general idea of the direction to go, but there is no clear path outlined for you. So you start out driving north to the Netherlands. Once there, you ask someone how to get to Germany. They tell you to travel to Berlin. Once in Berlin, you seek directions to Stuttgart. You’re advised to travel south through the Czech Republic, then finally to Stuttgart.
Finding the right information in an organization from the right person can be as indirect as your travel. Too many organizations continue to rely heavily on having a general idea or inaccurate directions of a well-meaning coworker, leaving employees taking an unnecessary European voyage. When, if they’d known the right person to ask, the best route would have been a seven-hour car ride from Paris straight to Stuttgart.
Today’s remote and hybrid workplace complicates the issue and creates additional digital silos. Remote workers often have even less access to the documentation to help them do their jobs. That’s why Starmind focuses more on the human experience and presents the best solutions, or directions, for employees.
Thank you for the great interview and for sharing your views on the nature of intelligence, readers who wish to learn more should visit Starmind.