AI - Status Update

 This man is an AI Start-Up founder and CEO and he has a hard time. When he was close to tears I remembered an IT specific. Doc Google helped me find it:

The theory that adding more developers to a software project can slow it down instead of speeding it up is known as Brooks's Law. 
This principle is established in the book "The Mythical Man-Month: Essays on Software Engineering" by Fred Brooks, first published in 1975. 
 

Key Details from the Book
    The Core Statement: "Adding manpower to a late software project makes it later".
    Why It Happens: Brooks explains that software development is not a perfectly divisible 

      task. When new people are added, they require training from existing staff (ramping up), 

      which slows down current productivity. Additionally, the communication complexity 

      increases exponentially, requiring more time for coordination than for coding.
    The "Man-Month" Myth: Brooks argues that it is a mistake to think that man-months (a 

      unit representing one person's work for one month) are interchangeable with product 

      volume. For example, a task requiring one person five months cannot be completed by 

      five people in one month.
    Context: The book is based on Brooks's experience managing the development of the IBM 

     OS/360 operating system in the 1960s. 

The Mythical Man-Month remains a classic text in software engineering, with an anniversary edition published in 1995 that reaffirms these principles in the context of modern development.

 

1975. A book. I was not even born than. That's what no drugs whats so ever and a library pass did to me. I was more than a Skateboard in Baggy Pants with a Butterfly knife.

 

Fred Brooks explains that software development is very different to manufacturing. In manufacturing adding more machines and men into an assembly factory linear increases production.

Software development is different. It is not the factory part, but the designer part. Most industrial mass produced core products were designed by one man. The Nuclear Bomb, the most dangerous mass produced item, needed a team to be developed.

Many investors and even managers mix up that creating software is a constant design like process and not at all the manufacturing known from mass production.

 

Software development means to organize a workforce into teams. Schedule their work and coordinate their output. In large teams the communication skills, the usually secondary soft skills called character set of a human, become eventually primary skills for success. 

Software can be million lines of code. That software exists of Classes, Objects and Functions. 

Class (Top Level - Blueprint/Abstract): Represents the high-level design, template, or blueprint of a system component, defining data and behavior.
Object (Instance Level - Concrete): A specific, unique instantiation of that class with defined data.
Function/Method (Behavior Level - Action): The executable code or behavior that acts upon the object's data.

In modern coding, running a several dozen big team, the idea of reusing any of these three parts over and over again instead of creating it multiple times without knowing from each other, is a structure with its own name: Object Oriented Programming.

So, everyone does that. They all have learned that, but tools and organizational structures to manage that in large teams are rare. You need a Coder Manager being a human.

 

 

Managing large teams to prevent redundant code requires a combination of visibility, discovery, and standardization. Beyond basic version control, specific tools focus on helping developers find existing solutions before they write new ones. 

1. Code Discovery & Intelligence
These tools allow developers to search across the entire organization's repositories to find existing functions or libraries, preventing them from "reinventing the wheel." 

    Sourcegraph Cody: This tool excels at multi-repository semantic search. It allows  

       developers to ask questions like "Where is the payment token validated?" in natural 

       language, surfacing existing code across different repositories to avoid duplicate 

       implementations.
    GitHub and GitLab: Both platforms offer advanced search capabilities. GitHub’s 

       InnerSource features specifically encourage teams to open up proprietary code 

       internally so it can be reused across different departments. 

2. Redundancy & Quality Detection
Static analysis tools can automatically scan codebases to flag "near-matches" or structural clones that indicate redundant logic.

    SonarQube: An enterprise-grade platform that identifies code duplication (copy-paste 

       blocks) and "technical debt" across multiple projects. It provides historical trends to 

       help large teams monitor and reduce redundancy over time.
    CodeAnt AI: Unlike simple scanners, it can spot duplication across different files and even 

       separate repositories.
    Codacy: Provides immediate feedback on code complexity and duplication during the pull 

       request process, ensuring redundant code isn't merged in the first place. 

3. Knowledge Management & Documentation
   Centralizing "how-to" guides and component libraries ensures that team members know          what has already been built 

    Confluence: Often used alongside Jira, it serves as a central hub for documenting shared 

      libraries, architectural decisions, and APIs, making them accessible to the whole  

      organization.
    Notion: Teams use Notion to build shared code snippet libraries, categorizing reusable 

      functions so they are easily searchable for new projects.
    Mintlify: Automatically converts code comments into live documentation, helping 

      maintain a clear "index" of existing tools for the team 

 

4. Component & Infrastructure Orchestration
For teams using modern architectures, these tools manage shared assets and environments.

    Bitbucket: Integrates deeply with Jira to link code changes to specific tasks, ensuring that 

       if a feature is already being worked on, it is visible to others.
    Nx: A build system for monorepos that helps manage dependencies between different  

       projects, ensuring they share the same core libraries efficiently. 

Are you more interested in tools that automatically detect existing duplicates in your current codebase, or are you looking for a portal/library where developers can manually browse for reusable components? 

 

That means, that collecting money, hiring staff and setting up an Office by having desks, laptops, screens and a meeting room with a white board, is falling very short of what is needed in reality to make success likely, but exponentially. 

You have to actually design an Office, an IT Coding Workspace dedicated to a given set up developers. 

 

While the Industry is having its next hype, even IT guys are like your average Secretary. One step ahead of TipEx on the Screen.

 

 #cyberpunkcoltoure

#deggers