top of page

Fanfare vs. Reality: Tech Trends Meet Healthcare

  • Writer: Shubhdeep Singh
    Shubhdeep Singh
  • May 16, 2024
  • 3 min read

A Single Technology is Rarely the Answer.


 

Those of us involved with the world of healthcare have seen firsthand how the introduction of modern and cutting-edge technologies has transformed the face of the industry. Robotic-aided surgery, remote patient monitoring through wearables and sensors, use of AR/VR for training as well as telemedicine are just a few examples of real-world applications that would have been considered works of science fiction just a couple decades ago. However, the one field that has been the talk of the town is AI, specifically Generative AI.

 

Even as I’m writing this blog, just a few days ago OpenAI announced the launch of its new GPT-4o model. While it remains to be seen exactly how much the new update improves the model’s performance, Altman’s industry disruptor is hardly the only player in this realm. Just last week, Google shared a white paper on their work on Med-Gemini, the latest healthcare-optimized AI model that seemingly surpasses all prior models. This, along with their recent work on Vertex AI, MedLM, and the Healthcare Data Engine, provides for a strong suite of tools for healthcare professionals to utilize.

 

Despite all the promise in the models and engines from Google, OpenAI, and others, the real-world applications of these technologies have not had the same impact. Although there have been some encouraging outcomes, the challenges and limitations to using Gen AI in healthcare are abundant. Issues such as regulatory compliance, patient & data security, and inaccurate results have resulted in a lot of people questioning the authenticity of the AI-powered future being sold today.

 

One such take comes in an opinion piece by Julia Angwin in The New York Times, where she compares the hype around Gen AI to that of blockchain and cryptocurrency a few years ago. She bases her argument on various lackluster results of AI testing, mostly showing how AI can work hard but not smart. In talking about its usage, she mentions “…A.I. models can often prepare a decent first draft”, which is an accurate representation of AI models being helping hands/assistants and not replacements for human experts.

 

While the article does a great job of questioning AI’s true capabilities, it is the comparison with blockchain that caught my attention. The biggest ethical considerations around Gen AI use in healthcare are rooted in bias, transparency, and accountability. Interestingly, these issues that plague AI are the exact ones that blockchain seeks to address and eliminate. In fact, there is a definite synergy between the two technologies where not only can AI benefit from the trustless and decentralized nature of blockchain but it can also help aid blockchain with its own limitations like latency, scalability, and interoperability.

 

AI can be used within a blockchain network to enhance smart contracts to be responsive and relevant, enforce adherence to consensus rules, review transaction data for anomalies, and so much more resulting in a crucial balance of privacy and transparency. Many experts believe, this symbiotic relationship between the two innovative technologies is the key to unlocking their full potentials and this has been reflected in many real-world applications across different industries that use AI and Blockchain in tandem. An example of the success that comes from such collaboration is the country of Estonia that has managed to digitalize 99% of its governance and public services on a backbone of cutting-edge technologies.

 

Within the realm of healthcare, both Gen AI and Blockchain are still emerging technologies that still require considerable real-world resting and refinement before they can see widespread adoption. I, like many others, do believe that blockchain and AI can work hand-in-hand to deliver the future. But how far away that future is from today is a different question. With the amount of energy required to harness the true potentials of these technologies, there are other problems that need to be addressed before we can truly adopt these technologies in practice to the extent that we hope to.

 

 

References and Helpful Links:

 
 
 

Comments


Let’s talk:
Schedule an introductory call

Thank you! We will get back to you shortly.

  • LinkedIn

©2023 by DS Digital Consulting

bottom of page