Understanding and Leveraging Abstractions in Technology
Written on
Chapter 1: Introduction to Abstractions
Every technology we use today is fueled by the concept of abstractions.
To truly benefit from abstractions, it’s essential to grasp their nature first. We often create and utilize abstractions instinctively without recognizing it. While many abstractions occur unconsciously, this discussion will focus on explicit abstractions that allow us to tackle everyday challenges with greater efficiency and control. Computer science serves as a rich source of ideas for crafting these abstractions.
This piece is part of a series that explores practical applications of computer science. We will start by defining what an abstraction is, followed by an illustrative example to build our understanding. Finally, we will examine advanced topics such as abstraction layers and their associated costs.
By the end of this discussion, you should feel confident thinking about abstractions intuitively and gain familiarity with the basics of designing effective abstractions for real-world problem-solving. Let's begin with a clear definition.
What is an Abstraction?
An abstraction is a technique used to simplify the complexity of a system. Imagine you are faced with a complex system; using an abstraction allows you to conceal certain intricate components, making the system simpler and more user-friendly.
One of the primary purposes of abstractions is to render complicated systems more accessible and comprehensible. A practical example of an effective abstraction is a coffee machine.
A coffee machine consists of numerous intricate components and subsystems that most users do not understand. Designers create abstractions that allow users to interact with the machine as if it were a straightforward black box.
When we use a coffee machine, our focus is solely on the inputs (milk, sugar, coffee beans, etc.) and the output (coffee). The internal workings are insignificant to us. This is the essence of a successful abstraction—it transforms a complex system into something manageable.
However, abstractions are not limited to household appliances; they apply to any intricate system. To deepen our understanding, let’s consider the complexity of human systems.
How Human Systems Utilize Abstractions
You may not have considered it this way, but organizations can be seen as complex human systems. For example, suppose you are the CEO of a tech firm. Your organization consists of distinct departments: 'tech', 'business', and 'customer support'.
In this structure, department heads from 'tech', 'business', and 'customer support' report directly to you while managing their respective teams. This hierarchical structure serves as an abstraction that simplifies the complexities of the organization. A less effective structure would involve you micromanaging all departments, which could quickly lead to chaos and inefficiency.
An effective abstraction allows us to break down a complex system into understandable components with clearly defined roles. In your organization, separating tech development, business operations, and customer support fosters smoother workflows, as everyone knows their responsibilities.
Yet, we must consider the broader context. We are only examining one layer of abstraction. When we introduce multiple layers, we may encounter new challenges.
Layers of Abstraction in Human Organizations
Imagine one day your chief of tech approaches you with a concern. He feels overwhelmed due to the overlapping responsibilities between departments at a lower layer of abstraction.
This overlap creates confusion about task ownership, as seen in the following illustration:
In this scenario, 'bug fixers' are shared between 'tech' and 'customer support', while 'customer channels' are split between 'business' and 'customer support'. You need to address this issue promptly to prevent it from impacting productivity.
Challenges with Layers of Abstraction in Organizations
One potential solution is to create two distinct teams for 'bug fixers': one for 'tech' and another for 'customer support'. Alternatively, you could consolidate roles, assigning 'bug fixers' solely to 'customer support' and 'customer channels' exclusively to 'business'.
However, splitting or merging departments often has drawbacks. Not every division or combination yields positive results, but some configurations will be more effective than others. We will explore how to identify better solutions soon.
For now, remember that although layers of abstraction can present challenges, they often simplify our lives.
How Computers Leverage Abstractions
To resolve your organizational issues, we can draw parallels from computer systems. When companies like AMD or Intel launch a new processor, they also introduce an abstraction layer, known as "assembly instructions" or "machine instructions," which the processor utilizes.
Human interaction with machine language is challenging, so we have developed programming languages (like C or Python) that are closer to human languages. When you write a program, an "interpreter" translates it into machine language for the processor.
This translation process incurs a time cost, referred to as the cost of abstraction. Although it is possible to work directly with machine code, doing so is comparable to a CEO micromanaging every aspect of their organization.
Now that we've examined how computers benefit from abstractions, let's discuss how to evaluate different abstractions based on their costs.
Calculating the Cost of Abstraction
Consider a Python program you’ve written to address a specific issue in your organization. It runs for 90 seconds and costs $0.02 per hour of runtime. If executed 500 times, the total cost would be:
Cost = (Runtime in hours) x (Cost per hour) x (Number of runs)
Cost = (90 seconds / 3600 seconds) x $0.02 x 500 = $0.25
Now suppose one of your top tech employees devises a more efficient program in assembly language that runs in 10 seconds. However, implementing this program requires 1 hour of her time, costing $100.
The question then arises: is it worth implementing this new program? To answer, calculate how many executions of the new program would be necessary to offset the $100 implementation cost:
Time saved = 80 seconds
Cost of new program = $100
Total executions needed = $100 / (Runtime saved per execution cost)
If the program is expected to run frequently—say, over 225,000 times—it could justify the implementation cost.
How Your Organization Can Benefit from Abstractions
The analysis of the computer program illustrates how to evaluate resource trade-offs involving different abstractions. Similarly, in your organization, it is wise to simulate various departmental configurations based on their cost and efficiency.
Based on these simulations, you can experiment with several options before settling on the most practical solution that enhances productivity while remaining approachable for all involved.
Final Thoughts
We have seen that designing and implementing abstractions involves costs. As with the programming language and assembly language example, introducing various components often slows performance due to increased communication.
However, in most situations, the benefits of abstraction outweigh the costs, as they not only simplify the handling of complex systems but also facilitate maintenance and error resolution.
For further reading, consider exploring: "Why Are Analogue Computers Really On The Rise Again?" and "How To Benefit From Computer Science In Real Life (I)."
If you appreciate this content, consider supporting me on Patreon.
The first video, "Abstraction in Software Design," delves into how software design principles utilize abstractions to simplify complexity and enhance functionality.
The second video, "Abstraction in Reinforcement Learning," explores the role of abstractions in reinforcement learning, focusing on how they streamline complex decision-making processes.