Post Scarcity Health care — Part 2 AI and Scarcity

Elliot Smith
5 min readMay 4, 2018

--

Last time we looked into what post scarcity is and what effects it has on society. To summarize briefly, post scarcity kicks in when we have more of a resource than we need, think oxygen. When we take away scarcity in health care we level the playing field. Health outcomes become about the best we have to offer rather than the best we can afford.With ‘the best we have to offer’ being an ever moving target what we really need is a way to take what we have today, make that as cheap and accessible as possible, grow what ‘best’ is then repeat this process. I believe that the way to do this is using Artificial Intelligence (AI).

AI has received a lot of media interest lately so I will assume the general concept of AI is not totally new to you. What I think is worth talking about is what AI can do today, what it should be able to do soon and what is still a long way off.

Today’s AI, with the right dataset, is very good at taking data and finding relationships to a desired output. This might be finding faces in images, classifying a cat and a dog or predicting change based on passed events. What AI is starting to do is explain why it made its choices. In the early days of machine learning this was an easier task, decision trees could explain themselves by showing their cutoffs. As models became more complex, explanation became more difficult. When models learn relationships between image data and text it is not always simple to explain why. There is however, growing progress in this area.

With this in mind we should expect, in many fields that there will be an increase in tasks done by AI. These will be tasks that humans could do but probably don’t need to. Things like sorting lists in order of importance or finding similar content, all possible using humans but much less work when done using AI. To give a rational view on what to expect, I think of it as raising the bar. Taking something like photo enhancement as an example, products like Google Photos already use AI to adjust your photos in a way that makes them more visually appealing. This raises the bar. People will find themselves spending far less time adjusting white balance and color profiles to get their photos looking right. There is a low chance however, that vanity fair will ditch its photographers and let AI set their cover photos. At least not this month.

Finally, what AI isn’t likely to do any time soon. Despite the hype and research in this area we are still a long way from replacing a human with AI. I’m certainly not saying it will never happen, in many ways I hope it does, but we have a long way to go before that day comes. The way AI is progressing it much like machines in the industrial revolution. Machines were made to remove the bottom 20% of tasks from a human, giving them 30% more in return. This process continues and in doing allows humans to produce more for less.

This doesn’t mean AI is set to take over all clinical jobs. I strongly feel that the day when AI (and probably robots) can be a drop in replacement for a clinician is still a long way in the future. As a result we feel the best thing we can do for post scarcity health care is to help clinicians deliver health care to more people. This means we can focus on the things AI can do well — pattern recognition, parsing large volumes of data — and let humans do what they do best, exploring, creating and empathizing.

It is worth looking at AI in the specific context of health care as there are already a number of exciting things happening in this space. While there are many more than I can mention here I have picked these examples as they have a clear path toward helping bring about post scarcity health care.

Babylon Health’s Chat App

Babylon Health are one example of a company publicly working towards these same goals. Their chat platform provides immediate and 24–7 contact to clinical expertise. First, users talk to their chat bot which helps to determine what the issue is and who best should handle it. This then facilitates connection to a real human doctor to continue the discussion. By digitizing this triage process Babylon can give the clinicians behind the scenes back some of their precious time to work on what they do best, treating these patients.

Butterfly Network’s forthcoming ultrasound device.

Butterfly Networks is tackling scarcity at the hardware level. Their soon to be released IQ device allows any clinician access to ultrasound imaging. Using AI to assist in image processing, their device becomes portable and more accessible to clinicians at all levels of training and at a small fraction of the cost of a traditional setup.

Lab work at Recursion Pharmaceuticals

Recursion Pharmaceuticals is using AI to run large volumes or parallel drug experiments looking for novel application of new and existing compounds. This effort helps bring more potential solutions to the market faster and has the potential to do so at lower cost than traditional methods.

All of these great companies and many more are bringing a world of post scarcity health care a little closer to reality. These technologies let us reach post scarcity because the cost to extend them to more people is minuscule. AI also has the added benefit of growing stronger with use. Every patient connected, every ultrasound and every drug tested adds to the pool of data behind these systems. This data refines and improves the system, allowing it to perform with more accuracy and begin to ask more and more complex questions. So few of our technologies have such virtuous usage patterns.

Our current tooling in development for prostate cancer reporting.

This virtuous cycle is a huge draw card to be working in AI. Knowing that we can create technologies that go on to help more people than you designed them for is a deeply rewarding feeling. At Maxwell MRI we’re working alongside these great companies to provide new diagnostic tools for clinicians. It’s our hope that by collecting and analyzing patient data over time we can spot changes before they become problems. Learn which things are dangerous and which are better left alone. Providing a specialist clinician in you, one that partners with clinicians to give everyone the best in health care and growing that ‘best’ along the way.

I’ll be giving a talk on Post Scarcity Health Care at the upcoming Myriad Festival. If you’re around it would be great to see you there.

--

--

Elliot Smith

CEO & Co-Founder of @maxwell_plus combining AI and medicine