This essay was originally published as a weekly newsletter via Substack. If you enjoyed this essay, please consider sharing it with a friend or subscribing to the newsletter.
One of the mistakes that people tend to make when choosing colleagues is selecting for credentials, rather than curiosity. This can work for status games but breaks down if the job requires working together for longer than a few months. For long-term endeavors, I believe that curiosity is a much better indicator of success.
This mistake seems to be so common because credentials are an objective measure of ability, whereas curiosity is dependent on the topic in question. Choosing who to hire is expensive, and credentials provide a paper trail that limits the downside of making a decision. If a well-credentialed candidate turns out to be a dud, hiring managers can throw up their hands and say, “how was I supposed to know, they looked so good on paper.” Curious candidates may be more qualified, but pose a greater risk to the individual responsible for hiring if the curiosity is feigned or misaligned with the work. …
Views and opinions expressed in this article are my own, and do not reflect those of my employer. Any information shared is drawn from publicly available sources and do not contain PHI/PII.
I work in healthcare and am surrounded by opportunities to improve the status quo with ML techniques. It recently came to my attention that the Centers for Medicare and Medicaid Services (CMS) announced the fee for service billing error rate had dropped from 9.5% to 8.1%. Unfortunately, an 8.1% error rate still means a loss of $31.6 billion dollars to the Medicare program.
This article will be the first of a series where I capture my attempts to build automated solutions for labor-intensive tasks that drive inefficiencies within our healthcare system. …
Always gon’ be a whip that’s better than the the one you got
Always gon’ be some clothes that’s fresher than the one’s you rock
J. Cole — Love Yourz
I’m going to describe a person that you know: They always seem to find themselves in the right place at the right time. Even though they aren’t particularly smart or hard-working, they’ve become successful. A series of lucky breaks propelled them into their current position.
You can’t help but check on their progress. Looking at their social media is a guilty pleasure that you would deny if anyone caught you in the act. …
Standing before a group of students, Nobel Laureate and behavioral economist Richard Thaler asked this question in what would become a landmark experiment. Thaler’s research found that subjects tended to systematically overvalue near-term versus long-term rewards. Put simply: the majority of subjects were willing to extend payment, even if it meant paying an extra fee.
This is one of many psychological biases that shape our daily choices. As humans, we tend to maintain the status quo, avoid difficult decisions, and undervalue the future. These pitfalls can manifest in various ways: neglecting exercise, procrastinating, or under-allocating money for retirement. For individuals these tendencies are merely problematic. …
Take a moment to check the applications you have running. How many Microsoft Office products are on that list?
Word, Excel, and PowerPoint are required skills for most entry-level positions. College graduates joining the workforce can expect to find these tools required for 4 out of 5 open positions.
This is unsurprising. Productivity tools serve as general-purpose applications that come pre-installed with each desktop. They are good enough for most office tasks and mysterious terms like “VBA” and “macros” float around college business classes. …
Elon Musk’s secretive brain-computer interface (BCI) company Neuralink presented the results of their work over the past two years.
The presentation revealed a device that sits behind your ear and will allow you to control your phone with your thoughts. Use of the device requires undergoing a medical procedure where an automated robot implants thousands of electrodes into your brain. The first product is called the N-1 Sensor and will be controlled via an iPhone app. In-human clinical studies are forecasted to start in 2020.
A collection of interesting headlines from (mostly) reputable sources.
5. The Chinese territory of Xinjiang is quickly becoming a testing ground for the securitization of facial recognition technology under the “One Belt, One Road” initiative. …
This post was inspired by a recent conversation where I was asked to describe my dream job. I’ve been fortunate to work in some incredible environments and I hope this post will serve as a benchmark.
A good job is challenging. A good job requires a multi-disciplinary skill set and encourages collaborative behavior. A bad job is repetitive, narrowly focused, and discourages sharing information.
A good job allows for open communication and promotes concise and clear writing at all levels. A bad job involves posturing, pointless meetings, and bureaucracy.
A good job rewards those who develop solutions to problems that are not found in the back of the textbook. A bad job rewards those who are most effective at publicizing their results, regardless of actual impact. …
Today’s youth will become tomorrow’s patients; how will they seek treatment?
The widespread adoption of artificial intelligence is posed to reshape the way we make strategic decisions about our health. The healthcare industry is particularly ripe for innovation as we attempt to include the individual actions of billions into our projections for the future. This point of view will explore a possible timeline for mankind with two guiding assumptions to shape the vision:
· The amount of patient specific data available will continue to grow exponentially.
· The future of healthcare will be focused on personalized, value based care.
Using these guidelines as a framework for discussion, current trends and market forces can be extrapolated to forecast a sense of how decisions today will impact our eventual global well-being. Specifically, this text will focus on the advancement of predictive models and artificial intelligence as they relate to healthcare. These technologies continue to benefit greatly from the explosion of individual specific data being generated through the proliferation of mobile devices and technological integration with existing systems. …
This is the final part of a series documenting the end to end process to develop a generalized linear model designed to output Airbnb rental price based on a number of features. As a whole, the series will include a description of dataset analysis, advice for additional data collection via web scraping methods, feature engineering (specifically for unstructured images), model selection and results. Readers should find this document helpful when developing their own predictive models, or when looking for a framework to organize their thoughts. Most importantly, I hope to demystify some of the process behind “data science” by breaking down a typical workflow into distinct and modular activities that can be reproduced for many types of problems. …