What do you do if artificial intelligence is reshaping healthcare practices?

Embracing AI in healthcare offers immense potential for improving patient care and advancing public health. However, it is vital to approach this transformation with caution, ensuring that the benefits of AI do not lead to the centralization and privatization of medical knowledge. By advocating open source, encouraging collaboration, and prioritizing equitable healthcare systems, we can leverage the power of AI to address global health challenges. The goal should be to leverage AI as a tool for empowerment, ensuring that all individuals, regardless of their socio-economic status, can access the benefits. In doing so, we can work towards a future where healthcare innovation serves the common good, rather than the interests of a select few.

For healthcare professionals, this evolution means staying informed about the latest AI developments and actively seeking out training opportunities to enhance their skills. Continuous learning and adaptability are key to leveraging AI tools effectively and improving patient care.

Keep up to date with the latest developments in AI and how they can directly affect and impact your workspace and specifically your niche. Aim to be the go-to person about these advancements and prepare the information in a format that can be shared if needed.

In my experience, staying up-to-date with fast-moving technologies like AI requires active engagement, particularly through conferences. These events provide direct exposure to the latest developments, insights, and ideas. One of the most valuable practices I’ve adopted is hosting “failure sessions,” where we discuss what didn’t work in our projects. This open exchange helps us learn from one another’s mistakes and fosters continuous improvement. While academic journals and articles offer foundational knowledge, they often fall behind the rapid pace of technological advancements. By the time research is published, it can already be outdated. That’s why hands-on learning and connecting with others in the AI space is so important.

When health data is derivatized for AI model training, it often becomes part of proprietary systems rather than open-access resources. This raises ethical questions about the ownership of data and the benefits derived from it. While the potential for improved health outcomes through AI is significant, the transformation of personal health data into Intellectual Property (IP) can lead to inequalities in access to the benefits of such research. If the data is used to create commercially valuable AI models, the original contributors may not receive any of the benefits, which contradicts the ethical principle of justice in research.

I always find it uncomfortable to talk about this topic, because from a corporate perspective, I believe we often approach ethics more rigorously than in the medical field. Let’s be honest—doctors frequently share patient information, such as CT scans and details, over platforms like WhatsApp to seek second opinions from colleagues. If this is happening, I’m quite certain that many doctors, who aren’t typically trained in data management, may not fully understand which patient data can be shared when it comes to AI applications, and which information is more sensitive. AI literacy is incredibly important in this context.

The integration of AI into healthcare presents both opportunities and challenges for physicians. By prioritizing AI education in medical curricula and fostering the development of key competencies, we can prepare future physicians to harness the power of AI effectively. This proactive approach will not only enhance patient care but also empower physicians to navigate the evolving landscape of medicine with confidence. As we move forward, it is essential to recognize that AI is a tool that can augment human capabilities, ultimately leading to better health outcomes for patients and a more efficient healthcare system.

In addition to the points mentioned, I believe it’s crucial to regularly check in with patients when developing AI tools intended for their use. It's important to ensure that these tools are something they feel comfortable with, and to gather feedback on what else we can improve to maximize their benefit. I’m passionate about patient empowerment and the idea that technology should enable patients to take control of their health. We need to create tools that are not only user-friendly but also enhance patients' understanding of their own health throughout the process.

From my perspective, it’s also important to emphasise the role of mentorship in skill development. Experienced professionals who are well-versed in AI can mentor others, ensuring that the transition to an AI-integrated healthcare environment is smoother and more collaborative. Additionally, developing soft skills such as critical thinking and communication can further enhance the effective use of AI tools.

To mitigate biases, adopting a global collaborative approach is essential. Over the past two years, we've observed a growing trend in open-source-based global collaborations. These open and shared R&D strategies bring several key benefits, fostering innovation, inclusivity, and project efficiency. Recently, this trend has begun to influence the medical AI R&D community. The shift from closed to more open collaborations is a promising development, as it not only democratizes AI but also helps reduce information asymmetries that contribute to global inequalities. This transformation has the potential to revolutionize the way we address some of the world’s most pressing health challenges, creating a more equitable & accessible future for all.

In my experience, the only way to truly move forward with AI in healthcare is through open communication and building trust. There's a great deal of fear from both doctors and patients when it comes to adopting new technology. They worry—can they trust it? Where is their data going? And if something goes wrong, will there be a real human available to listen to their concerns? The key to addressing these fears is to be present and truly listen. We must consistently engage with patients, understand their real needs, and maintain ongoing communication. Only by doing this can we help foster trust and ensure that AI is embraced in a way that benefits everyone involved.

I think that the phrasing “embrace AI” suggests that it is being pushed from outside and you are sort of standing there. That viewpoint may not be sufficient to keep you in a state of being the one in control of the proceedings. I personally first encountered AI in 1996, as a teenager obsessed with technology, and went deep into programming. My suggestion is that you take your pro-activity to a new level. Merely embracing it means that you are only capable of squeezing it in your arms, trying to contain it. Unfortunately, that is not a sustainable mindset - because if you try to squeeze a technology as revolutionary as AI it could break your arms. Rather, get into the driver’s seat, be in control. Take responsibility for training AI models

One thing I’ve found helpful is accepting that we may never know it all. I’m fully aware that I have gaps in my knowledge, especially when it comes to AI in healthcare, but I’m always learning from other experts and exploring how to blend different technologies. We can only do so much, but if we approach our work with integrity and poise, we show others that we take what we do seriously. And if we don’t have the answers, we commit to finding them. With this mindset, I believe we can move forward confidently toward the future of care.