Life Sciences and Artificial Intelligence? Only ten per cent of AI is used in a truly innovative way, the rest are common applications
05 January 2024
Pavlína Koutecká, Scientific Analyst at i&i Biotech Fund.
Pavlína Koutecká is a computer scientist specializing in artificial intelligence and bioinformatics. Since last year, she has been working as a Scientific Analyst at i&i Biotech Fund. Her career path reflects a blend of deep technical knowledge, cutting-edge research and a startup environment. Pavlína’s working experience has been marked by a strong focus on projects at the intersection of AI, Life Sciences, and healthcare. During her time at the Czech Institute of Informatics, Robotics, and Cybernetics, she contributed to research in the field of mathematical optimization and machine learning. "I see AI as a tool that now and in the future will help in various aspects of our lives," she says in the following interview.
How did you get involved with artificial intelligence and its application in the Life Sciences?
It all started during my studies at the Czech Technical University in Prague, where I focused on using artificial intelligence for detecting cancer cells in histological images, using AI for the analysis of cardiac or motion signals, and analyzing the human genome using AI.
You placed second in the IT SPY 2023 with your thesis, a competition for the best IT theses in the Czech Republic and Slovakia. What was your thesis about?
My thesis was dedicated to the use of AI in smart scheduling of surgeries in hospitals. Specifically, it involved devising a system that allows the most efficient allocation of patients to operating rooms. I considered aspects like time, costs and other factors.
Have you worked with AI in fields unrelated to human health?
Yes, during my studies I also worked as a machine learning developer in a startup that dealt with integrating AI into various companies. For example, a group of farmers from Germany approached us, wanting to use AI for weed detection in their business.
How can artificial intelligence help in controlling weeds?
Here, the idea was that farm employees used drones to fly over fields and detect the presence of weeds. The collected information was then transmitted to tractors, which subsequently applied sprays smartly, only where needed. We helped them implement artificial intelligence in this process to automate, expedite it, and especially to reduce the use of pesticides.
AI, with its ability to analyze vast amounts of data, can identify patterns that are otherwise difficult for humans to recognize. This way, it can reduce the role of chance in drug development and also speed up and refine this process.
Scientific Analyst at i&i Biotech Fund
Let's talk about your current work for the investment fund i&i Bio. What do you focus on in the fund?
In the fund, I am responsible for assessing the credibility, meaningfulness, and uniqueness of the integration of AI in the technologies of companies seeking investment from us. Following that, I analyze the competition, the company's results, potential buyers of the technology, or look at the expertise of the team behind the project. Besides this passive analysis, I also actively search for projects that we could support. Based on the prediction of future trends, I try to find those that could bring something unique to the market.
What percentage of the projects seeking investment in the fund involve AI?
Roughly estimated, it could currently be 15-20%, and this proportion will increase in the future. This will happen not only because there will generally be more of these projects, but also because startups will realize that we are focusing on this area and start approaching us more. However, it's questionable whether an increase in the number of AI projects will equate to an increase in the innovativeness of AI applications. I'm more sceptical about that.
In what percentage of Life Sciences projects can we talk about truly innovative use of AI?
Based on my experience working for i&i Bio, it's about 10%; the rest are standard applications.
Is there a fundamental difference in the use of AI in Life Sciences compared to other fields?
Fundamentally, the use of AI is the same, whether it's in developing a new drug or generating text. The difference, however, is that in addition to knowing AI, you must understand biology, which allows you to apply generally known algorithms in innovative ways.
AI is a very useful assistant, which science will very likely not do without in the future. But it will never work without human curiosity, research, and the desire for knowledge.
Scientific Analyst at i&i Biotech Fund
How specifically can AI help, for example, in discovering new drugs?
Today, chance plays a significant role in the discovery of new drugs. Many scientific breakthroughs in pharmacy have come from unexpected observations in experiments that were originally aimed elsewhere. AI, with its ability to analyze vast amounts of data, can identify patterns that are otherwise difficult for humans to recognize. This way, it can reduce the role of chance in drug development and also speed up and refine this process.
One thing is the use of AI in startups, another is the perspective of big market players. For example, how do large pharmaceutical companies view the use of artificial intelligence?
Large pharmaceutical companies are increasingly investing in AI projects. It's an obvious trend, explained by the fact that these companies simply don't want to miss out on something that could be groundbreaking. Successful implementation of AI can save time and resources for pharmaceutical companies compared to current methods. So why not use it?
To face these and other risks, critical thinking, educating ourselves and those around us, and continuous discussion are essential. A legal framework for the use of AI must also be defined.
Scientific Analyst at i&i Biotech Fund
Do you think there will ever be a situation where AI completely replaces the work of a scientist?
I don't think so. AI is a very useful assistant, which science will very likely not do without in the future. But it will never work without human curiosity, research, and the desire for knowledge.
Finally, I would like to ask for your comment on the risks associated with the use of AI. What do you think are the biggest risks, and how should we face them? And what risks do you see in the use of AI specifically in the natural sciences?
There are several risks associated with the use of AI. One of the fundamental problems, for example, is the data we use to train models. They often reflect existing biases in society, leading to skewed decisions by the model. An increasing problem is also distinguishing between real and AI-generated content. In today's world, it is becoming increasingly difficult to determine whether images, videos, or texts are created by a human or a machine. This can have serious consequences, such as in the area of media manipulation or the spread of disinformation. Concerns about privacy violations and misuse of sensitive data are also significant. To face these and other risks, critical thinking, educating ourselves and those around us, and continuous discussion are essential. A legal framework for the use of AI must also be defined.
In the Life Sciences, to the aforementioned general risks, issues such as ownership rights are added. If AI designs a molecule de novo, the question arises as to who is its discoverer and who owns the rights to it. This issue has broad implications for patenting.
So should we be afraid that AI will one day wipe out humanity?
Haha, probably not. I see AI as a tool that now and in the future will help in various aspects of our lives. The idea of its autonomy and desire to exterminate humanity is quite a bit of sci-fi for me.
Thank you for your time!
Author: Martin Kovalčík