Wisdom From The Women Leading The AI Industry, with Emily Lewis

An Interview With David Leichner

Not only for information collation, but for the ability for AI to “think” and problem solve on our behalf. With enough training on large, unbiased datasets we can work towards AI that utilizes logic and rationality rather than just recall.

As part of our series about the future of Artificial Intelligence, I had the pleasure of interviewing Emily Lewis.

Emily is a thought leader within digital medicine, building AI driven care solutions within her role in Digital Business Transformation. She is a digital humanist, firmly believing that technology like artificial intelligence is meant to augment and supplement human capabilities, rather than replace them. She is also a strong proponent of algorithmic justice, utilizing her expertise in AI to democratize access to care, tackle unmet patient needs, and ameliorate disease burden.

Thank you so much for joining us in this interview series! Can you share with us the ‘backstory” of how you decided to pursue this career path in AI?

Let’s face it. I’m a millennial through and through. While I remember the analog age of cord phones and dial up internet, I am mostly a digital native. Most of my adult life has been spent on a computer or phone, utilizing broadband internet. I like to think I was born at a “sweet spot” where I remember the “before times” and appreciate the values of hard work and perseverance and yet I have this insatiable appetite to learn and adopt all that there is to the disciplines of technology and innovation.

I can vividly remember my grandpa saying to me, “What’s all this dot com stuff?!” Me and my family still joke about that comment to this day. At some point technology can pass you by and it requires diligent work to stay abreast of what the latest and greatest capabilities are. I would say that my initial interests in AI came when the transition from MapQuest to GPS devices like Garmins allowed us to better navigate unknown territory. I’ll also never forget when I test drove my first Tesla. I was utterly amazed at what the car could see (traffic cones included!) and do (automatic breaking, course correction, and self-driving capabilities) just based upon narrow AI. I would say that was the first time I “experienced” AI firsthand. However, my journey into the realm of healthcare AI started in earnest when I began working on software products and solutions. I’ve worked with some of the brightest companies in the bay area to create products that delight users because AI allows them, at a fundamental level, to better navigate their life (pun intended!).

What lessons can others learn from your story?

Just like AI can drive your GPS to “re-route” you on a different path based upon the training of its algorithms, it can also allow you to explore new areas of a city, save time, be alerted to that crash ahead. Life will constantly throw unknowns at you, that is for certain. But if you “program” yourself to say I’ve been trained to tackle this challenge by being agile (e.g., “re-routing”) and being open to a growth mindset, you can absolutely upskill yourself to stay abreast of this burgeoning field, meeting the right people, and be in the right place to allow your career to soar. There are so many opportunities in this space, and those opportunities will only multiply over time.

Can you tell our readers about the most interesting projects you are working on now?

I am currently working to build a rare disease platform which is poised to be utilized as Software as a Medical Device (SaMD). The app is initially being built to serve patients with Myasthenia Gravis; a neuromuscular condition involving muscle weakness, double vision, and difficulties with speech and swallowing. Utilizing computer vision, we are building features which will allow patients to independently track their disease symptoms and symptom severity in remote settings using their smartphone. Patients can perform eyelid fatiguability exercises and then take a selfie of their face to objectively measure eyelid droop (ptosis) with the assistance of an algorithm. Additionally, they can follow prompts for vocal exercises and subsequently record audio samples of their voice to measure symptoms of dysarthria and dysphonia. While we are still in the process of deploying the Digital Medicine Society’s V3 framework (verification, analytical validation, and clinical validation), we are gathering the clinical evidence needed to support the deployment such features in direct-to-patient decentralized clinical trials (DCTs). We are also planning the human factors studies needed to support the development of this technology. Ultimately, we believe that multi-sided platforms like this one will not only help patients be more informed about their disease progression, but allow them to better self-manage it where they are able to. What’s more, they will be able to better communicate with their healthcare providers by sharing data rather than relying on recall which can be subject to recall bias.

None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?

Thanks to the internet, I can follow many leaders in this industry without needing to attend every conference! One woman who inspires me within this field is Najat Khan, Chief Data Science Officer at Janssen. She so eloquently speaks of the most pertinent AI use cases we’re seeing in the life sciecnes industry and has a tremendous backstory. You can learn about her backstory here in her interview with PharmaVoice for Woman of the Week. While I am not a data scientist by training, I love to listen to folks in bioinformatics to better understand how they approach problems and use data to arrive at solutions.

What are the 5 things that most excite you about the AI industry? Why?

1. The potential for the reduction in mundane, administrative tasks and ability to free up time. Everyone needs a helping hand, right?
2. Not only for information collation, but for the ability for AI to “think” and problem solve on our behalf. With enough training on large, unbiased datasets we can work towards AI that utilizes logic and rationality rather than just recall.
3. We have moved way beyond image recognition and classification (“Is this a hot dog or not a hot dog?”) and are now utilizing generative AI to create images. Large language models (LLMs) like like DALL-E can create realistic, original images and art from natural language, combining concepts, attributes, and styles.
4. While generalized LLMs like ChatGPT have been created to serve the general public, LLMs like BioGPT and MedPalm are now being trained on medical information (e.g. using PubMed) for use in the life sciences.
5. AI has the potential to help discover, and select, new compounds for pre-clinical testing for higher throughout screening. Companies like Generate Biomedicines are on the bleeding edge of combining machine learning, biomedical engineering, and medicine to utilize generative biology to create breakthrough medicines.

What are the 5 things that concern you about the AI industry? Why?

1. Unfortunately, much of the data that exists has some inherent bias in it. We know for a fact that for many years research was largely only open to white males which is problematic in many dimensions. Did you know that women were not even allowed in research in this country until 1991 when the U.S. Department of Health and Human Services established the Office on Women’s Health? Issues of diversity, equity, and inclusion are always top of mind for me and I worry that AI can perpetuate errors and mistakes of the past.
2. Access to the right datasets to build AI is oftentimes behind paywalls or only able to be accessed by academia or nonprofits. Industry has a difficult time obtaining data, believe it or not! This is why the pooling of data through consortia is becoming more and more important for us.
3. AI transparency is a huge bugaboo for me. I am concerned about how the healthcare industry can be transparent about the algorithms they build for patients and healthcare providers within the context of clinical decision support. While the end user should absolutely know what data was used to train the algorithm and where it came from, in generative AI contexts this can sometimes be impossible to do because they are so many data sources. You also have to recall that, for industry, sometimes the algorithms are the “secret sauce” that allow them to make money (and put more money into AI research) so having everything open source just is not practical as a business model. It’s a delicate balance, but I believe we can work to strive to achieve transparency (and accountability) while preserving intellectual property.
4. The AI industry has catalyzed corruption. We’ve seen a huge rise in data brokers, and particularly those who sell personal health data. Data is the new currency and it easily gets into the wrong hands. I believe that individuals should own their own data and have greater agency over its control and use.
5. Algorithmic change and the ability for companies who build AI backed products to not be overly burdened by the regulatory and bureaucratic red tape that stifles innovation. Products with an AI backbone, by their very nature, need to be iterated upon. In theory they are constantly improving as they are fed more data, so the product itself is changing. While this can be worrisome from a quality assurance perspective, regulators are coming up with creative solutions (e.g., the FDA Pre-Certification program) to certify whole companies based upon their practices and policies rather than requiring the repetitive review of individual software products).

As you know, there is an ongoing debate between prominent scientists, (personified as a debate between Elon Musk and Mark Zuckerberg,) about whether advanced AI has the future potential to pose a danger to humanity. What is your position about this?

In the most extreme metaphor, I liken AI to fire. It’s a tool and can either be used to keep you warm, ignite an engine to propel you somewhere, or it can totally incinerate your entire home. There should be parameters for it’s use to keep people safe and its power is not to be taken lightly. I think AI done well does not pose a danger and can tremendously help solve some of humanity’s crises (across energy, healthcare, etc.). Unfortunately, as we saw in the case of Tesla’s car crashes, AI done poorly can result in injury and even death. The problem is, sometimes trial and error is needed in order to make “good” AI.

What can be done to prevent such concerns from materializing? And what can be done to assure the public that there is nothing to be concerned about?

My data colleagues would say we need to ensure the test and training datasets are separate. From a macro lens, however, I would say we need to more algorithmic transparency allowing users to better understand how the AI came up with the answer it did. In some cases, we need to train the users to question the AI’s response. For clinical decision support, for example, we need to ensure that doctors see the AI’s suggestion and say, “Hm. That’s interesting. That does/does not make sense and align with my clinical training” and be empowered to take the AI’s suggestion (or not).

How have you used your success to bring goodness to the world? Can you share a story?

I enjoy giving back in many arenas. I have lots of people direct message me on LinkedIn asking to chat but I also formally give back by volunteering and mentoring with companies like HITLAB, the Healthcare Businesswomen’s Association, Newchip, Prime Health, Astia, and Multiple. I really enjoy the healthcare startup and accelerator ecosystem and love learning about how people pitch and invest in new ideas. I love hackathons like MIT Hacking Medicine and others which allow me to sit with entrepreneurs and walk through the development of their product/solution with them. While these events are compressed into 48 hour periods, it brings me so much joy to think about where a group started on day 0 to where they end up on day 2. And many even win some money (and networking opportunities)!

As you know, there are not that many women in your industry. Can you advise what is needed to engage more women into the AI industry?

My mother always recounts that “back in her day” the only two viable jobs outside of the home were that of a teacher or a nurse. When computers became more mainstream and came into the home in the 80s, the fear was still there that women couldn’t use these machines and they were for men. It’s ridiculous to think about in retrospect, but women are just as technically savvy as men now and this “fear” that women couldn’t do this work was all a complete myth. Now, more than ever, women need these four things: 1) opportunity, 2) access, 3) support and 4) community. There are many great initiatives emerging that I am a part of, which particularly serve women, that I would like to highlight. Women of Wearables is a fabulous community of like-minded entrepreneurs and businesswomen in the life sciences space focused on the development of technology by women for women. It focuses on all things FemTech and their founder writes some really great pieces to highlight these women as a contributor to Forbes. There are also groups like HITLAB which put on events like the Women’s Health Tech Challenge and PharmaVoice as mentioned earlier which has a “Woman of the Week” segment. Women in Product is another great group as well as Women in Innovation (WIN). The options to get connected are honestly endless these days. We’re also seeing professional societies like The Alliance for Artificial Intelligence in Healthcare (AAIH), which can bring companies from the industry itself together.

What is your favorite “Life Lesson Quote”? Can you share a story of how that had relevance to your own life?

Lately, the quote that has lingered in my head was from my colleague Jennifer Goldsack of the DIME society. When asked about the consumerization of healthcare she replied something to the effect of, “I don’t care if triple bypasses are two for a dollar, no one should want to CONSUME healthcare.” I thought that was a really fresh perspective and it’s something that I’ve kept in the back of my mind. While I want to build products that are easy for patients to use and that delight patients when using them, I don’t actually want patients to have to use them. So much of the healthcare industry was built around generating a need for healthcare that may not have existed before. Think about the last time you heard a commercial for a drug on television — you didn’t think you needed that product until you heard that it existed. This is also much of why I work in “beyond the pill” digital medicine. While traditional drugs have their place in modern medicine, they are not a panacea and do not allow for a holistic approach to medicine treating the whole patient. This is where digital can play a huge role in healthcare to plug those gaps by providing care navigation, symptom tracking, medication reminders, digital therapeutics, and the like.

You are a person of great influence. If you could start a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. :-)

I really enjoy being a part of the Algorithmic Justice League. It’s outlets like this that allow me to both remain educated about issues in AI, but also be a force of change. I see myself, hopefully, using my knowledge of this field to affect change or help shape new policy in government oversight and regulatory bodies like the Food and Drug Administration (FDA) and Centers for Medicare and Medicaid Services (CMS).

How can our readers further follow your work online?

They can follow or add me on LinkedIn at https://www.linkedin.com/in/emily-kunka-lewis-ms-ccrp-a243842/. I’m always looking to network with other folks interested in AI and technology.

Thank you so much for joining us. This was very inspirational, and we wish you continued success in your important work.

About The Interviewer: David Leichner is a veteran of the Israeli high-tech industry with significant experience in the areas of cyber and security, enterprise software and communications. At Cybellum, a leading provider of Product Security Lifecycle Management, David is responsible for creating and executing the marketing strategy and managing the global marketing team that forms the foundation for Cybellum’s product and market penetration. Prior to Cybellum, David was CMO at SQream and VP Sales and Marketing at endpoint protection vendor, Cynet. David is a member of the Board of Trustees of the Jerusalem Technology College. He holds a BA in Information Systems Management and an MBA in International Business from the City University of New York.