Smart machines explained in 2 x 2 Quadrant

I am sure over holidays you talked about Google home/Alexa, autonomous vehicles and smart toasters and their interaction with your existing smartphones and fitness trackers etc. These machines are becoming integral part of our daily life.

We love using these smart devices, they make my life a bit easier and fun, but do we need to understand these machines and its impact on our lives in coming decades? I think yes, we need basic understanding of this emerging new species. If machines are going diagnose my cancer report and suggest whether I take chemo therapy of go for tumor removal, I better know how it works. If machines are going to drive my kids to the school, I better make sure that they are safe.

What is smart machine? What varieties they come in?

Smart machine:

Figure 1 — Anatomy of smart machine

They are made up of ‘Body’ — Hardware and ‘Brain’ — Software (Artificial intelligence).

Body is a physical structure that gives us an interface to interact with. It also includes sensors that consumes input like voice, text, images. For example: Fitness tracker includes sensors for detecting heartbeat and tracking GPS location. More sensors mean more parameters a machine consumes. They are equivalent to our senses (sight, hearing, touch etc.) but much more accurate. These sensors also work together and feed into the central brain. In mumbo-jumbo terms it’s calls Internet of Things (IoT).

Brain is the software that does all the behind the scene work. They come in whole spectrum of varied complexity. They vary from a simple algorithm that analyse your heartbeats and locations and give you ‘Achievement’ badges on your fitness tracker to the self driving cars that takes decisions on behalf that can impact your life. These algorithm (or programs) are called AI algorithm, neural network, deep neural network and other terms that doesn’t mean much to the people who are not working in that specialized field. Main function of these brain is to find patterns and/or to do probabilistic determination.

The fast pace innovations of these sensors (body) and the software (brain) is making these machines more powerful everyday. Iteration cycles are short and hence evolution is pretty fast.

Using our age old classic quadrants (2×2), these smart machines can be divided in four groups based on these two parameters. Number of sensors and software complexity.

Figure 2 — Smart machine quadrant

1. Simple body and simple brain:

Example: Fitness tracker, Smart thermostats, Security systems

Role: Assist an individual in their task

Functions: They gather few parameters, churn it through a simple algorithm and give us suggestive results as ‘get-up and walk’ or take simple decisions like ‘Start heating when humans are in home’.

Impact on us: They are facilitator to our daily lives. They provide a bit more convenience and helps us in better decision making. These machines assist, but humans are in control. We make decisions. If you are not going to get-up and walk, machine is not going to withhold candy bar from you that you are thinking of devouring.

2. Simple body and complex brain:

Example: AlphaGo, Google home, Amazon Alexa, IBM Watson

Role: Take a decision for an individual

Functions: They have very few sensors, but their software is quite complex.

Impact on us: Their complex algorithm helps us delegate some of our tasks. They help us in reducing our mental fatigue. They can help doctors to diagnose disease or book my next appointment to the physiotherapist. This is the fastest growing segment of the smart machines as there is very little hardware cost and new features can be added by installing new version over the internet. Consumer won’t even notice and they will get new features and functionalities.

3. Complex body and simple brain:

Example: Smart parking lots, Smart City, Environment trackers

Role: Assist community in simple tasks

Functions: Many sensors of few variety are deployed in a geographical area. These sensors collect tons of data, and software aggregate this data to provide simple answers or to find patterns

Impact on us: Let’s say “City of Buzztown” installs hundreds of sensors around downtown that tracks parking spots, traffic patterns, air quality etc. All these sensors feed into an aggregator and help citizens to find empty parking spots or running routes. Another example: Let’s say one is studying caribou migration in the Canadian north. They can put sensors on these caribous and track their migration patterns and correlate with other local environment patterns. These machines help us collect data on much wider scale, aggregate them and find patterns. These types of connected smart machines will have a huge impact on how goods and people move around the world.

4. Complex body and complex brain:

Example: Autonomous vehicle

Role: Completes a complex physical task on behalf of an individual

Functions: They take decisions and perform actions on our behalf. We humans need to provide desirable outcome and machine will take care things.

Impact on us: We will delegate complex repetitive tasks to them. They will free us from doing the grunt work. Industrial robots did similar work for years in the confines of factories, but these new breeds of machines are very adaptable to changing circumstances and will be able to operate around humans in their day to day activities. These machines will change the dynamics of workforce and our economy.

These machines are here to stay and they are going to have impact on every aspect of our lives. Our challenged will be to find the balance so we can coexist and thrive with this new species.

If creatures from today’s world time travels to 2040 and look into the world with today’s eyes, they will see one harmonized world run by a man and a machine.

“The creatures outside looked from machine to man, and from man to machine, and from machine to man again; but already it will be impossible to say which was which.”