The Future of Machine Learning

Nathanael O'Donnell

 This week, I would like to open up the field and contemplate the future of machine learning. In a relatively short period of time—barely one decade—machine learning has gone from being a relatively niche academic domain to becoming an integral part of our global information infrastructure. And the impacts of machine learning are being felt not just in the arena that we traditionally consider "software"; rather, machine learning is enabling intelligent behavior in all sorts of physical devices, leading to the "internet of things" that many have be prophesying.

Developments that once sounded like science fiction—such as a computer program beating a world champion in Go—are now science fact. The question is, what comes next? Rather than speculate using only my own imagination, I will survey the plans and predictions of industry leaders and pioneering academics.

AI Will Enhance Creativity

A great first stop for those interested in how machine learning might impact creative professions in the future is the Google Creative Lab. This "lab" is really an assortment of projects, funded by Google, that explore ways in which AI can enhance the creative process in a variety of disciplines. For example, the project "AI + Writing" explores ways that AI can help writers come up with ideas, flesh out unfinished text, or even chat with their characters in real time to generate dialogue! The "AI + Music" project allows musicians—amateur or professional—to experiment with novel sound-generation techniques, such as conducting an orchestra by moving their arms in front of the camera and playing a piano that accompanies itself. "AI + Drawing" lets people draw together with a neural network or generate handwriting that fits their style. 

A writer chats with Google's "Banter Bot" AI (source: Experiments with Google)

Factories and Warehouses will be Human-Free

Although many factories have been highly automated for quite some time now, there are still a large number of important tasks that robots have yet to mater. For example, Amazon still employs human workers to place oddly-shaped products in boxes for shipping. As the capabilities of AI robotics continue to steadily advance, however, the number of tasks that resist automation will dwindle, and we will begin to see workplaces that are nearly empty of human oversight, let alone manual labor. 

Just last year, Amazon began installing machines in its warehouses that package products in custom-sized boxes, thus eliminating the need for a human worker to perform this step. And while there are still other tasks that require humans to perform at Amazon, the list grows shorter every year. 
A robot sorts materials in an Amazon warehouse (source: Reuters)

Home Construction will be Automated

Even as the assembly of automobiles, electronic devices, and myriad household appliances has been automated, home construction is still a painfully slow and expensive process, and is dependent on skilled human labor to perform. However, as a result of advances in AI and robotics, this is already changing, and will eventually radically transform the construction industry as we know it.

There are currently more than 50 companies independently working on robotic construction methods, and many of them use sophisticated machine learning techniques to accomplish their tasks. Notably, computer vision algorithms are critical in performing some of the tasks that require extra precision, such as laying a brick exactly on top of another one (see below).


A robotic arm that lays bricks (Source: Fastbrick, via fbr.com.au)

The Long Term

With all of the above advancements—and more—coming to fruition in the next decade or two, what are the long-term implications of AI and robotics? This question elicits a variety of conflicting responses from experts on the subject, but there is consensus on the prediction that AI will obviate a large number of jobs (perhaps 50% of current jobs), both blue- and white-collar. Will these jobs be replaced by new ones? Will our work week be reduced to 15 hours, and John Maynard Keynes famously predicted? How will humans find meaning once the basic tasks required to sustain life have been automated? 

These questions do not have discrete boundaries, and in some ways they are already being answered. For example, as a result of the COVID-19 outbreak in the U.S., a large percentage of workers have been revealed to be "non-essential." This raises a perhaps surprising question:if so many jobs are "non-essential," then why do we place such pressure on all working-age adults to find full-time employment? Perhaps our utopian, automated future has already arrived, at least in part. As with all else, time will tell!  

Comments

Post a Comment

Popular posts from this blog

Applications of Neural Networks

Image
Nathanael O'Donnell Neural Networks - An Introduction In comparison with most other machine learning techniques—including linear regression and logistic regression—neural networks are significantly more complex. The simplest way of explaining neural networks is as multi-dimensional logistic regression, in which the "layers" of the neural network representing additional transformations. Somewhat intuitively (because they are based on logistic regression), neural networks, in their more basic forms, are classification models. However, more complex neural networks (those with more neurons) can serve as regression models as well.  So if neural networks are so much more complicated than other machine learning methods, then why do practitioners use them? The simple answer is that they work better than other approaches in many situations. The complexity of neural networks is what makes them such a flexible and powerful approach for more complex problems. Neurons can be...
Read more

Applications of Logistic Regression

Image
Nathanael O'Donnell What is Logistic Regression?  Along with linear regression, logistic regression is a widely used machine learning algorithm. In spite of its name, however—and in contrast to linear regression—logistic regression is not actually a regression algorithm. Rather, it is a classification algorithm , meaning that it is useful when categorizing (aka "classifying") observations based on any number of feature variables. Mathematically, logistic regression works by mapping observations to points along a logistic function , whose standard equation and graph are below: Source: Andrew Ng Note that g(z) , called the logistic function or sigmoid function, tends towards 1 as z approaches infinity, and tends towards 0 as z approaches negative infinity. This property of the logistic function is what makes it useful for classification: the output of the function will also be a number between 0 and 1, and more extreme values of z —in either direction...
Read more