Trying something new can make people nervous, especially if they are unsure of what the results will be. This can be even more nerve-wracking when it comes to healthcare solutions and therapy. Matthew Michelson, a machine learning expert and accomplished technologist, works closely with his colleague, Steven Minton, to help ease many concerns that customers have. Their California-based company, Evid Science, uses “breakthrough technology” to to provide innovative solution in Healthcare and for Pharmaceutical Companies. Learn more about how Matt is modernizing medicine.

Tamara: Can you share a story that inspired you to get involved in AI?

Matt: I’ve always loved languages and video games, and in college I realized I could combine the two via computer science and eventually AI. In This same deep interest in somehow mimicking what we do as people inspired me all the way through my PhD. And to be honestly geeky, I’ve always loved Data from Star Trek and hoped that in some tiny way I could contribute to the path that eventually leads to him!

Tamara: Describe your company and the AI/predictive analytics/data analytics products/services you offer.

Matt: Evid Science provides a platform for pharmaceutical, life sciences and healthcare companies to automatically compare therapy effectiveness. This means someone can ask our system how effective one therapy is compared to another, or how safe it is, or even how much it costs relative to other therapy choices. To do this at massive scale, for any drug, disease and outcome studied, we use an AI system that is capable of “reading and understanding” medical papers. That way the computers can pull out the results in the published papers and make sense of them, rather than relying on a person to do so, which would be too time consuming and expensive.

Tamara: How do you see the AI/data analytics/predictive analysis industry evolving in the future?

Matt: I think the evolution is going to be more about acceptance and pervasiveness, rather than technical achievement. Sure the algorithms will get better (they always do), but the bigger change is that people will become more and more comfortable with AI working on their behalf in all sorts of capacities, whether that is recommending what to eat or watch; driving you to work; or even planning a trip for you (let alone suggesting what to invest in and what disease areas show the most promise for your company’s research). Further, as people accept AI into more and more of their lives, there will simply be more of it around, since the barrier to deploying it will be lower. I am not the first one to say this, but I agree with the sentiment that AI will be the new electricity. Everything will have AI in it as a feature (not the main one, by the way!), and this goes hand in hand with acceptance. As it becomes more acceptable and useful, it will be included in more stuff – and the cycle will continue until AI just falls into the background.

Tamara: What is the biggest challenge facing the industry today in your opinion?

Matt: The biggest challenge is hype versus reality. People often expect an AI to have super-human abilities, when in reality the goal is often to meet human ability, and the “super human” aspect is that the machine won’t sleep or eat. So it can do the same task, only faster and at massive scale. Sometimes people expect their AI-enabled software should be writing Shakespeare when all it really is required to do is spit out the right text message at the right time. And those of us in AI are guilty – the promise of the technology is so amazing that we sometimes forget to ground our technology in reality when we explain it to people.

Tamara: How do you see your products/services evolving going forward?

Matt: Our AI will continue to improve its reading ability over time. Right now, our AI can read, more or less, at the level of a sophisticated college student or early medical student. That means it can understand quite a lot of medical text, but not all areas yet. Our goal is to get our AI to be as sophisticated a reader as a seasoned physician or scientist. We will certainly get there; it’s just a matter of time.

Tamara: What is your favorite AI movie and why?

Matt: Definitely can’t pick just one! So here are a few!

A.I. – It showed that people can and should have empathy any with a conscience, even if people constructed it

Star Wars – C3PO!

Star Trek TNG – (not a movie, ok) Data is the epitome of AI

Terminator 2 – We need to be conscientious about the dangers of AI. Also, Arnold…

Tamara: What type of advice would you give my readers about AI?

Matt: Learn what you can to make an informed decision – there is a lot of hype, uncertainty and fear and much of it is unfounded. Think of it this way, a self driving car is simply an amazing feat. And a computer program that can teach itself to win at Go or play video games perfectly is amazing too. But the self-driving car can’t play Go and recommend songs and it may never be able to. I think we are more likely to live in a world of many, many super-specialized AIs than one of a general AI that can do anything – and it is that myth, of the self-sentient, all powerful AI that seems to frighten people.

Tamara: How does AI, particularly your product/service, bring goodness to the world? Can you explain how you help people?

Matt: There is a lot of important healthcare evidence buried within the published medical literature, and humans simply don’t have the capacity to read and interpret everything that is written. A paper from a while back estimated it would take 86% of your waking hours in a month just to stay current on what’s published within general medicine. That’s clearly not a feasible task. And yet, there are so many important results that are published that we don’t want to let slip by. Our goal is to surface that evidence and make it instantly accessible so that healthcare decisions can be made using the latest, most relevant, evidence. We will start in the pharmaceutical and life sciences industries, making their research and other processes more efficient, and move into healthcare long term.

Tamara: What would be the funniest or most interesting story that occurred to you during your company’s evolution?

Matt: At a major pharmaceutical conference, I somehow got invited to a cocktail party full of large company CEOs and government representatives. I had no idea how this happened (but I went of course) and it was funny to engage in some of these conversations – I mostly just nodded and said things about AI, but steered clear of the fact that our company was close to 1000 times smaller than theirs!

Tamara: What are the 3-5 things that most excite you about AI? Why? (industry specific)

Matt: 1. Someone will make a breakthrough drug discovery or health decision based on AI, and it will be correct. This could be the genesis of the next wave of science: AI-enabled pharmaceutical and life sciences research where the machines work with people to hypothesize and test ideas. Because this would be so much more efficient, we could see more breakthroughs than ever before.

2. Bedside AI – At some point, I hope that physicians will find a way to integrate AI into their workflow so they can get back to the human aspects of medicine. Doctors are quite time constrained, so if an AI can help mitigate some of the mundane and repetitive tasks, doctors can spend more time simply talking to their patients

3. The ubiquity of AI – I think AI will become more accepted and pervasive, and it will flow into all sorts of different areas in the pharmaceutical and life sciences industries. I suppose what I am excited for is that there will be all sorts of new AI-enabled processes and products in the pharmaceutical and life sciences industry that no one has even thought of yet, and I can’t wait to see what they (and we!) come up with.

Tamara: What are the 3-5 things worry you about AI? Why? (industry specific)

Matt: 1. I worry that a decision that was outsourced to an AI could be wrong, with serious consequences. Even if a human would have made the same decision, because it was an AI it should be taken more seriously – anytime you have new technology and something related to health, this is a huge worry.

2. That there will be a backlash from all of the hyper surrounding AI, much of which is unwarranted. If people keep promising that AI will cure disease or find the next blockbuster drug, and then it doesn’t, it makes it harder for the rest of us to continue to be taken seriously.

3. That I won’t get to see all the amazing things in the future that a true age of AI could enable. If we can truly get to a point where AI’s can make intelligent, evidence-based, data-driven health and life sciences decisions, we could make serious headway against some of the major challenges in life science and medicine.

Tamara: Over the next three years, name at least one thing that we can expect in the future related to AI?

Matt: More evidence-based and data-driven decision making in the pharmaceutical, life sciences and healthcare industry. There are many companies taking an AI approach to handle all of the disparate, related data such as ours in the literature, others in the –omics or biome, some with real-time (mobile) data, etc. It might not be us, but some company is going to figure out how to make sense of all of this data for better decision-making, and the only way to do that is with AI.