Coding with AI Credit: reddit.com
Introduction
Generative AI (Gen AI) is a subset of artificial intelligence in which machines create new content in the form of text, code, voice, images, videos, processes and even the 3D structure of proteins. Some forms of Gen AI have been well established in this decade, but it was a Large Language Model (LLM) powering an easily accessible chat interface that enabled Gen AI to have its breakthrough moment and surprised many experts working in software development. How does Gen AI differ from traditional AI? The primary difference between traditional AI and Gen AI is that the latter can create novel output that appears to be generated by humans. The coherent writing and hyper-realistic images that have captured public and business interest are examples of Gen AI models outputting data in ways once only possible with human thought, creativity, and effort.
The Evolution of Programming
Looking at the evolution of programming one can distinguish the following three steps:
- Moving from punch cards to digital systems and machine language
- Moving from machine language to compiled languages
- Moving from compiled languages to interpreted languages
Associated with each move is a quantum leap in technology whereby traditionalists or purists believe that the new approach is causing a loss of knowledge. They think it makes people less good at their job because they need to understand less about what coding or programming is all about. When developers went from machine language to compiled languages, the accessibility of coding went up. One did not have to understand what registers are or what a jump-not-equal event is. One just needed to understand what an if-then-else is or what a variable assignment is. Consequently this was all one needed to know and with this knowledge one could build code required for the problem to be solved. The third jump from compiled languages to interpreted languages made things like the application of Java much more available to more individuals who wanted to work as developers. One could just put a file somewhere, and if it is loaded the correct way, it will be executed exactly as one expects. This accessibility led to more interesting and better tools being built and with these tools solutions could be architected that made prior systems obsolete.
The Evolution of Coding with AI
Just a few years ago, in the best case, developers were scanning through documentation and writing mostly functional code. At worst, one was chasing typos and strange bugs for hours, to no avail. But things have changed. We now live in a world where software developers who do not use AI tooling to write code are becoming extinct. The technology available, either extremely cheap or free, is so ubiquitously useful and impactful that coding as it was five or 10 years ago will be completely replaced. This shift comes with several advantages in terms of productivity and speed, but it also heightened a lot of fears and exposed potential problems worth exploring.
The clear advantage of using AI is simple. AI generates code faster than a human could ever write. ChatGPT, Claude, and others are capable of writing code, producing software that actually works. For an experienced developer, these tools are incredibly useful because they can create code that one can quickly scan and approve or reject. However, it is also obvious that you have to know good code to approve good code. This means two things:
- For experienced developers, these tools are massively impactful on the quality of work.
- For people learning to code, there is still a necessity to understand how coding works to judge the AI’s results effectively.
This is the core of a new problem. If one is learning to code with AI, one might never truly understand the solutions you need to judge the code against. If you do not know how to code, you cannot imagine what the overall system will look like and to judge what the auto-generated code is and does.
What Developers must learn
Just as some inexperienced developers get themselves into trouble by using AI code blindly, fully trusting its responses and not correcting its mistakes, there might be developers that cut and paste code, thinking that they had a working solution but as a result, create all kinds of problems for themselves. Development is less about where the lines of code come from and so much more about the thoughtful architecture and assembly of code into an efficient, working solution. There is much industry pressure for every organization to be seen as in front of the AI revolution. Everyone wants to be perceived as competently wielding an amazing new technology, handling rapid change where others find it hard to follow. Whether accurate or meaningful, the percentage of AI written code is going up. This will seduce developers making the wrong moves to rapidly get in front of AI, or the wrong calls as a backlash. Consequently organizations should not set a specific target for AI generated code. The goal is efficiency and AI is just one tool that may support it getting there. An increase in reliance on AI generated code without proper scrutiny is likely to decrease the efficiency with bloated code bases that are harder to maintain. Regardless of the cost and time spent to incorporate this technology, AI-enabled coding will continue to grow, benefiting its users. It lowers barriers to entry for coding and allows new ways to learn and experiment.
The Advantage of Coding with AI
According to Andrew Ng, well known expert on coding strategies with AI, some developers are discouraging others from learning programming on the grounds that AI will automate the necessary coding. In his view this advice can be seen as some of the worst career advice ever given. In the 1960s, when programming moved from punch cards to keyboards with terminals, programming became easier. Over the past few decades, as programming has moved from assembly language to higher-level languages like C, from desktop to cloud, from raw text editors to AI-assisted coding, frequently developers no longer examine the generated code, and this process is getting easier with every iteration for solving a problem. Tech-savvy people that use AI generating tools have 10 times the impact of the average individual in their field. According to Andrew Ng the best way for many developers to accomplish this level of proficiency is not as a consumer of AI applications, but to learn enough coding to use AI-assisted coding tools effectively. One question asked most often is what someone should do who is worried about job displacement by AI. Fact is, that there is a great demand for experienced developers who understand how to generate software because of their knowledge of coding and Large Language Models (LLMs). They know precisely how to apply these tools in order to get the best results. Consequently this is the best time yet to learn to code, to learn the language of software and learn to make computers do exactly what one wants them to do.
The Road Ahead
Many developers are defensive about change, feeling insulted that tasks they spent years mastering can now be done so easily. A lot of developers are hyper-defensive about this because coding as it is applied by them has taken a long time to learn and understand and now in a very short time, generating code has become very easy. But the benefits are undeniable as developers benefit massively from the increased speed of generating code. It is always easier to judge code, to look at it and read it, than to come up with it. That is the main reason why coding is now fundamentally different because developers get to not just write code, but to parse code, understand code and validate it. In this new era of AI-assisted coding, the role of developers is changing as they become architects, instructors and editors of code rather than just writers. And while the transition may be challenging for some, it opens up exciting new possibilities for innovation and efficiency in software development. As organizations move forward, they need to embrace this change, ensuring that developers have the skills needed to effectively guide and instruct AI in creating the software solutions of tomorrow. The future of coding is here, and it is a thrilling blend of human creativity and machine efficiency.