A career in the software industry is going through a major transformation with artificial intelligence. Quite frankly, it’s a lot of people who call themselves software engineers who don’t realize that job titles automatically render them redundant and moreso, worthless in the IT landscape today. 

Strong advice - avoid the term “software engineer” as it is as irrelevant as the term “web master” from the 1990s. You have to retool and recalibrate your technology skills to stay relevant and  land jobs in this current environment. 

One term I keep seeing on social media is the concept of “vibe coding” - a term that no real professional would use to their clients or employer. Vibe coding is someone creating prompts of the expected output and allowing AI to create the code of the software. While this may sound hipster, this method of coding is inefficient and not scalable and a seriously flawed concept. 

The problem is the vibe coder is mostly freestyling prompts and allowing AI to deliver the output and going with whatever the AI puts out. The other problem is the AI can miss the big picture of the application being developed and create code that is fragmented all over the place. 

This is no different than the Blavity Afrotech goofy that bragging they built a mobile app - did you ever see them actually release version upgrades every quarter? No, that app was a one-trick pony and they had no skills to scale up and scale out. Look at how lame Blavity is and irrelevant at this juncture in AI. Vibe coding creates the same one-trick pony where if customers demand a new feature, it will be difficult to upgrade due to refactoring complexities of AI generated code. 

Information Architecture and AI Approach

Today, in order to stay relevant and competitive as a software engineer is to no longer be a software engineer - you become an information architect. Then you transition to AI by becoming an intelligence engineer. 

Becoming an information architect is critical but the good thing is AI can guide you along the way. Anybody who wants to learn can become an information architect. The key is the following: 

Know what you want to build. You need to define the platform you want to manifest. You have to say you want an e-course portal, a turnkey hotel management system, a media streaming platform. You have to focus on 3 things - your customer, your service, and facilitating payment. Write all this down and analyze with AI to refine your vision statement. 

Look at your competitor's features. Look at existing patterns and practices and take note of the features. Are they using Google/Apple for the login? Do they have a dashboard to show token use or billing? A newsletter for marketing? Support section? Chat agents? List your competitors and look at everything from UX design, workflow value stream, to how they collect payment. 

Understand the core components. This is where you get to the nuts and bolts. You learn what the core functions are. Some will be reusable and you create a toolkit - this is what our Toshikiso API does. Simple functions such as send email, load file, save file, upload to cloud, download from cloud, login, logout. Every feature you create will likely use a simple function - make everything break down to a simple function such as Calculate Risk, Poll Incoming Folder, On Mouse Click Event. Break it all the way down! 

Stub the sum of all parts. This is the beautiful part about AI - you don’t need to code. You just structure an outline of your application with a comment like “send email function” and “save to database” and structure your code with the functions. On the UX side, you focus on prompts such as “show spinning circle to process form submit” or “populate the location after the user enters zip code / postal code” and just create the structure. 

At that point, you established a shell that should make sense to you as the infrastructure and architecture of your application. If you worked with Claude, it will help you create an infrastructure and create the outline and high level for you. You can also use SeedDream or GPTImage to create architectural diagrams based on your design, Claude can read all your requirements and propose the overall design for you. 

Intelligence Engineering through Management

The days of a software engineer or developer being a loner that keeps to themselves are over - that’s why they are laying them off by the thousands at companies worldwide. Today, you are going to have management and manager skills of people onsite, people offsite or offshore, and most important - AI agents. 

When you have the skills to define the features, design the prototype, stub out the functions, then assign the tasks to people and AI to do the work, now you are relevant in the technology industry today. The industry want a person like you to wipe out that entire line of $200K/year  goofy management, senior software engineer, junior software engineer, and that project manager - they just need you and when you built yourself up like that, they begging you to take the job while the rest of these clowns are fired. 

Look at these characters on YouTube with the sob story - I can look at them 5 seconds in and know these are irrelevant jokers that just rode a wave and did not make waves in their career, seriously look at them. That Black chick probably attended a Blavity Afrotech conference. Don’t be like them, be the apex predator - be the wolf. They are straight blabbing hard in these sob story YouTube videos, aren’t they? 

You got to be relevant and you got to be efficient and return value for the money they are paying you. So if you can show you got AI agents and how you created them, customize them, and the work they perform on your behalf, you just made yourself the most valuable person and sought after in the industry right now. 

With intelligence engineering, the goal is to have specialized AI agents focus on a specific task with specialization and divide up the work between all of them. You do not want one jack-of-all trades AI agent, you want to model them as if you are managing a team of developers. This way, you have them all work in parallel on the same code base doing different functions. 

You want AI agents that work on the business logic and features such as computing business rules, managing the inventory database. Then you want AI agents that can work on the design UX and know how to write React and HTML5 and are compatible with both Apple and Windows platforms or create desktop applications. You want AI agents that can test code and AI agents that can deploy code and inform you that a new prototype is ready for your review. 

Then you focus on fine-tuning your AI agents to do the work that privileged but mediocre White dudes in IT running a good ol boy network in the tech industry should have been doing but now they replacing mediocre White tech bros with H1-B workers and you knocking both of them out as a Black man or Black woman they tried to underpay and take credit for your work. You are the superstar now, mastering AI agents and running the show as a lone wolf, apex predator. 

Let’s summary the steps and please remember - you can get started in IT right now from nothing by just jumping in with the AI technology and outdo that clown in the YouTube video above that been working 25 years as a software engineer and now on YouTube blabbing woe-is-me crybaby stuff instead of retooling and repurposing. 

Focus on the big picture.  Design your application by discussing with AI and analyzing your competitors. Even Gemini and Claude can assist you in creating diagrams and other materials to print out and stick on your wall. 

Build out the model and prototype. Have your workflow, business logic, and data structure defined and work with AI to refine them. You have to know how to set up a code environment with the IDE that uses AI agents like CoPilot or Claude Code or Cusor and how they read your instructions like a claude.md file at the root directory. Learn how AI interacts in these dev environments like VS Code. 

Build out stubs of components.  Your code should be structured where you have a core component section where you create a communicationManager library that sends e-mail, SMS, or mobile notification, or send to social media or messaging apps. Then you have a specific library of functions related to your application purpose like board customer, upsell customer. 

Let AI agents do the tedious tasks and process. Learn to divide and conquer with multiple AI agents with specialty to build and test and deploy the code - this is the tedious work that had previous software engineers working tier coding for hours - you are just waiting for the AI to finish and review the final project and not get lost in the tedious stuff. 

You deliver the final project. Your job is to make sure the application and code works as expected and can attract customers, have customers do what they came for, and money is made through the value stream.

Pivot to an Information Architect and Intelligence Engineer

Trust me, this new software skill path with AI and being at the bigger picture management and executive level of designing, building and delivering applications is not hard to master and the place where you will be relevant in a period of transformation. 

The nice benefit is you will not be doing all the tedious coding anymore and fixing bugs all night long. There will be moments your AI agent will goof up, but you should know from interacting with AI on coding and design how to fix those issues manually by intervening or asking AI to help you fix your AI agent mess. You are always going to win. 

If you are a brotha, you don’t have to attend a goofy Blavity Afrotech conference anymore. Today, you can just go to the coffeeshop with your Asian chick while your AI agents are performing their tasks. And instead of working up all night and spending your dollars on ratchet sisters at the strip club, you can make passionate love with your Asian chick throughout the night as your AI agents work 24/7 to get the job done. 

If you are one of my Akibas out there - you need to have the same skillset to take you anywhere in the Afroasiatic ecosystem. Our breed and our caliber will need AI agents doing the tedious work for us while we pursue the bigger plans and bigger picture. It is also how we are going to generate revenue streams to support our lifestyle overseas and have a base back home. 

Akibas cannot just be a cosplayer in the Afroasiatic movement - we are making moves and manifesting in this paradigm shift. You should also learn tools like AI agents like OpenClaw and AI tools like GitHub Copilot  and AI workflows like N8N to build out business models like Japanese fashion, export kawaii cultures, build chillhop cafes and produce anime music and videos for Akibas. I hope you understand the opportunities ahead of you are bigger than the problems you are currently worried about, my sista. 

This is the future and direction of how we are moving in tech and carry on the mission of developing your self-ecosystem with AI and start going for the world as ours to thrive in.