Are we in the second wave of the healthcare AI gold rush?

Finding Gold

Over the past few months, I’ve seen a dramatic shift in how health systems are engaging with technology. COVID has exposed systemic weaknesses and driven a change in priorities, forcing health systems to seek out partners to help them navigate emerging technologies like AI, cloud, and cybersecurity. 

As a result, healthcare AI is getting more interesting and chaotic by the week. To help bring some sense to it all, I’m sending this note to share some novel insights we’re seeing in the healthcare AI landscape from the companies and health systems we work with. I hope these ideas will spark some thought.

First wave innovation vs. second wave innovation

In the early days of a new technology paradigm, it doesn’t make sense to build platforms. Think of Arpanet; when first developed, there wasn’t a need for a web browser because there were limited nodes and use cases. As the network grew, it quickly became apparent that a bottleneck to growth was the lack of a common way to access and use what we know today as the internet. The platform opportunity emerges once the ‘first wave’ of innovation occurs and the new technology paradigm matures. 

Timing matters. The explosion of AI applications signifies the gold rush has started, creating new and non-obvious opportunities that will become the picks and shovels of the AI gold rush. To date, the tools to enable AI adoption haven’t caught up to the healthcare industry’s needs. We know innovative early adopters will always do whatever it takes to try new technology. 

Eventually, however, we need common industry standards that lower the implementation cost for early majority users who have been sitting on the sidelines waiting for the dust to clear. During this ‘second wave’ immense value is created, typically after hard lessons have been learned. The intersection of needs and usability drives the creation of platform standards that enable high-velocity creativity and customer value. Coinbase is one of the latest examples and was founded a full four years after bitcoin’s debut.

As healthcare AI use cases rapidly grow, the complexity of management increases exponentially

It has never been easier to build healthcare AI algorithms. This is fantastic news for everybody, from patients to providers, and means there are significant opportunities for growth. A quick view of the industry tells us…

  • Currently, there are 1,200 AI companies in healthcare, up from 700 in 2020 – driven partly by a monstrous $80B in healthcare venture funding in 2020

  • The cost to develop clinical-grade AI has plummeted due to advances in tools and frameworks like Keras (2015), Tensorflow (2015), PyTorch (2016), AutoML, and pretrained models like BERT (2018)

  • Regulatory barriers to entry are lower, and the path is clearer thanks to the FDA 510(k) process (it’s literally called the ‘predicate device pathway’)

Based on our team’s experience with healthcare systems across three continents, we’re seeing healthcare executives come to us with a laundry list of healthcare AI applications (many that didn’t exist five years ago) that need to be installed and managed across their hospital system. Each of these AI applications requires a costly validation and integration process. 

The pain felt in healthcare is uniquely acute because of the complexity of inpatient and outpatient workflows. Healthcare leaders dealing with the complexity of a single AI application quickly find themselves needing to manage and integrate multiple AI applications. This means understanding how the AI applications interact together as well as with the vast web of healthcare data and workflows. As healthcare AI use cases grow, the customer pain increases exponentially, leading to the need for a secure and scalable platform to lower both the barriers to entry and the longer-term management costs. 

The chaos of COVID has accelerated the AI gold rush and kickstarted the second wave of healthcare AI, where value is primarily created through the lowering of transaction costs and barriers to entry. With this comes a sudden increase in AI use cases as algorithms are deployed across the health system, creating a demand for standardization and platform. 

Will the latest healthcare AI boom be the panacea that will fix our healthcare woes? What do you think? Drop us a note and let’s discuss! 

Picture of Pelu  Tran

Pelu Tran

Pelu is a serial healthtech entrepreneur; he studied both medicine and engineering at Stanford University and was four months away from receiving his MD when he dropped out to start his first company.

Contact Us

CASE STUDY

ARA Health Specialists

Use the button below to download your free case study and learn how our approach to validation has improved the number of clinically significant findings in AI software.

CASE STUDY

Sutter Health

Use the button below to download your free case study and learn how our approach to validation has improved the number of clinically significant findings in AI software.