To die, or not to die, that is the question. Using AI to predict when may be the answer.

Okay so my title isn’t exactly William Shakespeare.  I’ve written a lot about Artificial Intelligence (AI) over the years, but can it really be used to predict when I might die? If so, do I want to know? Talk about AI is everywhere lately. Last week, 60 Minutes discussed the impact on the workforce of the future. I am also watching season 3 of the Netflix Series Travelers about time travelers from the future following the directives of a highly advanced quantum AI program known as The Director. I’ll attempt to tie all three of these things together in the next few paragraphs.

Stanford Medicine is exploring the possibility of using AI to guide doctors, patients, and their families through end-of-life-care. Dr. Stephanie Harman, the clinical chief of palliative care at Stanford Health Care is leading the pilot program. The tool isn’t designed to predict when a healthy person may die or telling patients nearing end-of-life exactly how many months or years they have left to live. Instead, the predictive analytics model focuses on identifying patients who have a high probability of dying in the next 3 to 12 months. Someone in palliative care is not expected to live longer than 12 months.

Inaccurate predictions or physicians providing overly optimistic timeframes delay families from having important discussions. It’s an understandable human tendency to want to provide patients and their families with the best possible outcome. However, Stanford is removing the human element, often influenced by emotion, from the equation by using a fact-based approach trained on several million patient records. Stanford’s AI model identifies patterns that a single doctor is simply not able to catch. What they’ve found is their algorithm can forecast a palliative care patient’s death with an astonishing 90 percent accuracy.

By providing a more accurate prediction of time of death, it allows the patient and family to plan accordingly. The following stat is from the Stanford website, “While 80% of Americans prefer to spend their final days in their home, only 20% actually do. More than 60% of deaths in the US happen in an acute care hospital, most of the patients receiving aggressive care in their final days.” To read the paper in its entirety here is a link: https://stanfordmlgroup.github.io/projects/improving-palliative-care/ .

Now to my latest Netflix binge – Travelers. As indicated in my intro, it is about time travelers from the future traveling back to the present time. Directed by a sophisticated AI program the travelers take the identity of people that otherwise would have died on that day in history. Like any AI model the travelers have been trained on the person’s identity in advance and attempt to assimilate into their new life undetected while carrying out their missions for The Director whose predictive analysis has forecast how their actions will change the course of history.  I won’t ruin this for people that may be interested in watching. Here is a link to the official site and reviews from Rotten Tomatoes.

Lastly, the 60 Minutes interview with AI expert Kai Fu Lee. This was probably too big of a topic to lump into this blog, but I’ll share a couple relatable nuggets along with a link to the complete interview. Lee believes AI will have a bigger impact on the world than electricity did when history looks back. When asked specifically if he thinks a machine will ever be able to think like a human, Lee answered this way, “So, the bar keeps moving higher. I think that's, I guess, more motivation for us to work harder. But if you're talking about AGI, artificial general intelligence, I would say not within the next 30 years, and possibly never. I believe there's a lot of love and compassion that is not explainable in terms of neural networks and computation algorithms. And I currently see no way of solving them.”

It’s unlikely that the AI presented in science fiction will ever become our reality, but despite the long road that the technology has in front of it, advancements are making substantive impacts on the lives we’re living today.  Our piece of the puzzle at Extract is to continue to invest in AI to constantly improve accuracy for your data capture and redaction needs.


About the Author: Troy Burke

With 30 years of experience providing clients with stellar service and strategic solutions for growth and development, Troy is committed to ensuring his customers receive the highest quality solution, training and support with every implementation. He frequently speaks on the topic of redaction and is actively involved with National Association of Court Management, Property Records Industry Association and several other government organizations.