People have been trying to make machine generated music for a long time. Some of the earliest examples were musicians punching holes in piano roles to create complex melodies unplayable by humans (see Conlon Nancarrow, 1947).
More recently, it’s looked like electronic music in the form of MIDI files, where, by specifying various attributes —the instrument, pitch, duration, and timing—songs can be symbolically represented. But what does it look like for AI to run the whole generation process?
After starting chess, my rating immediately settled in the 800s. This is considered low novice territory; 2889 is the highest all-time rating, held by Magnus Carlsen. I wasn’t getting better regardless of the time spent.
So, you’re a PM and there’s been a lot of talk about how AI is revolutionizing the industry, how your company is data-driven, or how your product has a deep analytics focus. But what exactly does that mean? Without understanding how data-powered applications work, your PRD’s will be long on buzzwords and short on substance. We can talk ourselves in circles about the value of analytics, but without understanding how to productize machine learning, we’ll never actually ship.
This article will walk you through the five steps of building and deploying machine learning models, using time-series anomaly detection to demonstrate…
Designing cities for the future of mobility
The adoption of cars during the 20th Century dramatically improved standard of living while driving massive social changes and more subtle requirements in urban planning. Over time city planners developed precise zoning requirements from the Institute of Transportation Engineers’ (ITE) Parking Generation Manual, which have been shown to be arbitrary and not data-driven. Even in low-income housing projects where many residents don’t own cars, zoning codes require vast lots to house these non-existing vehicles. These regulations were well-intentioned but often resulted in bizarre outcomes.