Coding Away, or, Why AI Won't Kill Computer Science

"Coding is dead!"

"Don't go to school for computer science, you won't find a job!"

"The sky is falling!"

...okay, I made up that last one, but the point stands. It has become commonplace, and all to mainstream, to make statements like these, especially staring, ostensibly, the barrel of a combination of agentic coding workflow and/or AGI/ASI.

The panic is understandable, and there are likely quite a few that feel empathy toward it. The job market is already seeing massive and drastic cuts, including less hires and a shift toward augmented enterprises (a shift predicted over two years ago), especially as models like o3 become ever more capable.

What everyone is missing, however, is this: this change was inevitable. Whitehead assures us of as much we he cautioned that "The major advances of civilization are processes that all but wreck the societies in which they occur." Our society, in this case, is the computer science job market and the quite literal requirements of the profession, while the major advance, is, of course, the proliferation and oncoming ubiquity of GenAI.

This means that the change isn't unforeseen, and, more to the point, isn't catastrophic. It's expected that the job requirements will change. It's common and normal. That doesn't change, of course, the consternation behind it, so in this article I'd like to lay forth a challenge: this isn't an apocalypse but rather an opportunity.

If Whitehead was correct - and he's been proven time and again to be precisely so - then instead of chaining ourselves to the old job as it sinks into La Brea, let's have a hand in shaping the new future of what the career might look like. We'll even have AI help in which to do so. To wit:

# Define the output file name

output: my_output.txt

# Writing content

write: "Hello, this is the start of my document."

heading: "Main Title"

write: "Here’s some regular text that appears below the heading."

italic: "This line is italicized."

bold: "This line is bold."

newline

write: "And now we’re back to regular text."

If you're looking at the code above and wondering what language it's written in, the answer is TextLang, a very basic coding language that includes conditionals and flow control, as well as some simple output. "TextLang? Is that a derivative of LangChain?" you may be asking. The answer is no: TextLang won't appear on GitHub, it doesn't have support on any system, and no one can get a degree in it.

I created it.

In about an afternoon, using 4o MDAP agent design.

No, I'm not presenting this to show off the capabilities of GenAI - we're all intimately familiar with those - rather, this is a proof of concept to demonstrate what the future of computer science might be: a world in which each and every computer scientist defines his or her own programming language - much like an author has a writing style - and produces content in that language.

Is this a barrier toward team dynamics? Perhaps, if only until a bit of learning the syntax occurs. Instead what it offers is a chance for every single company with a bit of AI knowledge to ascribe not proprietary software but rather proprietary coding languages for their teams, languages built, maintained, and authored with the assistance of AI.

At the most basic level, this offers something as near perfect cybersecurity as one can get. It's difficult to hack a system in which one has no idea the language in which it was written, so long as it is bolstered by powerful portal control.

I invite you to think through additional capabilities - but most importantly, I invite you to think outside the box when it comes to utilizing GenAI in the field.

After all, this major advance in society has all but wrecked the field. We can curse the darkness, or we can light a candle.