Image credits: https://youtu.be/d1g1tltlVr0?si=8CqX1hcB9JXcR_ap

Can we stop talking about AI for one damn minute?

Updated

I am technically an unpaid intern but I still consider myself unemployed, because I’m not being paid and the work is nonexistent to say the least. Gives me something to put on a resume and stretch the truth of what my responsibilities actually are (for legal reasons this is a joke) I suppose until I find a reliable source of income while being able to extract some amount of job satisfaction, so I’ll be withholding my application to my local McDonald’s for a while unless I am suddenly discover my passion for flipping burgers or working with people in retail…

Anyone who looks at any tech news at this point knows about the AI LLM explosion in popularity. As outsiders looking into the AI world, at least for content generation, it feels like the hype just grew exponentially, and is still growing, ever since OpenAI released ChatGPT. What’s crazy is that they had existing APIs for their less powerful GPT models before and people could’ve launched their own form of “Chat with GPT” applications. I remember creating smut using unfiltered text davinci in their web beta playground and trying to create a discord chatbot acting as a catgirl childhood friend. The latter was definitely more of a waste of time because you can get better versions of that by talking to real people, but let’s move on from darker times…

Anyways before ChatGPT maybe one can point to the release of a neural network architecture called Transformers that started things off. Andrej Karpathy was already amazed with what he saw RNNs could do, and we can see the magic of all the smart AI researchers and engineers work in the many LLMs today:

This is not even mentioning any of the AI Inference Compute providers like Replicate and Groq. The best performance comes from the big models, while smaller models will probably be used in applications where it makes the most sense for meeting some business’s need and their budget.

The talk of the town in some parts of tech twitter is “how does this impact my job security as a swe or aspiring swe?”

When it comes to developers solving problems, creating customer or shareholder value, writing code, and actually creating usable software there are two main different experiences of this process that will divide a lot of people on how they feel about this AI progress:

  1. People who enjoy the process of development.
  2. People who enjoy the results that software creates.

Of course you can feel both and the workers’ subjective experiences is not something corporations care about anyway, but we are talking about people’s experiences and the unknown ripple effects of potentially many developers offloading their thinking to AI tools and creating more technical debt (more on that later).

The software development experience depends on what you are building. Websites/web applications, SAAS, B2B soft, video games, low level hardware drivers, operating systems, kernels, complex and high distributed and efficient backend systems, databases, new snapchat filters, etc.

For tasks that involve implementing some feature whose result is fundamentally just good working code, the experience of doing the thing is like: problem/puzzle solving, feedback loop with making changes, running your code, viewing the output of your program, making more changes, etc, or just stepping through and viewing the state changes of your program in a debugger, typing and editing code — especially if you’re using vim motions.

Then there’s the experience of delivering/deploying/finishing the thing and reaping what you sowed. There’s some people who just don’t like the physical act of typing for instance or don’t care for the feedback loop of working through a problem and hate debugging. Some people also just don’t care about the dev experience, whether they like parts of it or hate parts of it — it doesn’t matter. They are completely focused on results, a die hard pragmatist if you will.

In our world dominated by companies and mega corps, it does feel like if you want to survive results are everything. Generally in the real world, nobody cares about how hard you tried, the results is all anybody cares about. If you want and are seeking affirmation for your hard work that may or may not pan out, seek God or truly believe in your mission. My mind jumps immediately to the story of the developer behind the indie video game Stardew Valley. “Just grind it out bro, day in day out, 10 hour days, 7 days a week, for 4.5 years and it’ll work out for your work of art like it did for Eric Barone.” That’s a hard sell, but maybe that’s what it takes to make it. Good luck.

One should never bet against the rise of new technology that gives more power than what was previously available — just look at history. It’s always been adapt or die. This doesn’t mean that for your personal enjoyment you should stop riding a horse at all if cars are invented to get around to places. Maybe you’ll learn to enjoy driving, or maybe you’ll never enjoy driving and only drive if you don’t have enough time and need the speed and practicality of the car over your life long horse. 

So I think newbies who might feel they’ve been jebaited into majoring CS for college, if you do enjoy the dev experience learn to use the AI tools for productivity. Coding assistant AIs can generate a lot of code and it just works. You prompt it either with English or with some code comments or a start of your code snippet and it will autocomplete that shit like it’s nothing, given you are not working on something super advanced, complicated, or novel that the model has never seen before ofc…

If you’ve never used an AI coding tool, some options I know include:

If you never enjoyed the dev experience and always cared more about the end goal of your creation, then you’re probably already slurping up and trying out all of the different AI tools. And I’m sure there are many other aspects of the dev experience that you actually care for that this whole thing going on is a wake up call to you that you don’t care about software, and that you actually cared about X instead. 


What does all this news mean for say an unemployed, partially Gen Z brainrotted, recent Comp Sci graduate who’s neither a normie or a cracked engineer, perhaps possibly much more closer to the stereotypical midwit than they would like to admit, that is aspiring to have a career in software development?

I’m not the only one asking this question of course.

Maybe we have nobody to blame because the root cause of our collective problem is that we are just not skilled enough and need to “lock in” and grind more. At the end of the day many of us losers need to actually develop our social skills, because the wall that needs to be overcome is convincing a real person who has access to lots of money that you are worth something to them to pay you money. “My dad is the CEO” and other nepo babies I’m not talking to you right now.

Many people lament “wasting” time in college and make videos asking about whether college or CS at college is worth it:

Most people just go to college because it is the tried and true path to getting a job and starting your adult life where you actually have agency over your life. You should still probably go to college / pursue higher education if not for learning and growing as a human, for meeting people of the same age.

AI can automate so much and has taken people’s jobs already like for graphic design and copywriters. This can be scary for many, and I certainly feel anxiety in general in this world we live in, but what can we do but hedge our bets on actions and plans that we hope will give us the life we want in the future. “It is what it is, and we just have to say fuck it we ball.” As individuals we have to be careful to never let it automate our thinking. “I think, therefore I am.” Let it be an augmentation to your thinking process, another input among your scattered stimuli to churn and churn into some thoughts that might lead to action in the real world. Next time choose an employable skill that can’t be automated (jk, that’s unfair, but fr tho…)

This kind of brings another question like should any governing entities do anything about some probably soon to be significant fraction of their citizens not having a job/stable source of income.  If we look at history as a reference point, many jobs have disappeared over time and I’m pretty sure those people were just told “eat shit and die” or maybe “figure it out” for their questions of “what do I do now?” Your investment dried up, you lost your main source of income, and maybe you lost a big part of your fulfillment in life. What are you going to do now?

Whatever you do, I wish you great fortune.