Skip to content

Instantly share code, notes, and snippets.

@buwilliams
Created March 17, 2025 16:31
Show Gist options
  • Save buwilliams/4799bda61b50f10abbac0ca1d5614c15 to your computer and use it in GitHub Desktop.
Save buwilliams/4799bda61b50f10abbac0ca1d5614c15 to your computer and use it in GitHub Desktop.

Any advice for an AI enthusiast who wants to start creating?

Let's start with situational awareness in three categories: Purpose of Technology, AI Maturity, and the Economy. Finally, I'll add my suggestions.

I. The Purpose of Technology

TLDR - It’s all about time.

  1. In the most fundamental sense, technology is the lever for time. If you haven't given much thought to my last statement, it's worth reflecting on. It provides the means for extending our lives and allowing us to choose what we want to do. There are two dimensions of time: quantity and quality. Maybe, eventually, we’ll add another dimension: position.
  2. If you look at it sideways, AI is sort of the last invention of humans. It’s mastery over intelligence—perhaps not the form of human intelligence, but certainly its function.

II. AI Maturity

TLDR - It’s moving super duper fast.

  1. AI Maturity: Theory > Components > Applications
    1. Theory: represents the latest research in the field.
    2. Components: the building blocks of AI systems (audio, video, image, text, robotics.)
    3. Applications: tangible workflows and systems.
  2. AI is improving exponentially. Depending on the type of product/service offered, building long-term solutions with current capabilities could be an expensive mistake if enough people are involved. What is rapidly evolving? All of it—the theory, the components, and the applications. It’s a moving target.
  3. Centralization of software functions has historically been essential, but with the capabilities we see in agentic workflows, this historical architecture may not hold. In theory, one centralized application will likely split into many applications. Think of a workforce of agents working on niche functions and communicating with other niche agents.
  4. Frontier model companies may add functions in the form of agents. This is analogous to Apple adding a feature to iOS and removing the need for a previous application. Due to agent evolution, many startups will lose their bets.

III. Economy

TLDR - Nobody knows.

  1. The applications space will offer many opportunities because many inefficient businesses exist. There are currently several market challenges:
    1. Sellers must pivot their businesses, adding new employment delays and hire freezes.
    2. Sellers are trying to determine what to offer buyers.
    3. Buyers do not know what to buy because it’s new and confusing. This is where the primary opportunities exist.
  2. Since we live in a capitalistic economy, buyers will purchase based on the bottom-line rationale.
    1. In all cases, production capacity will exponentially increase. The measurements of my productivity have shown two orders of magnitude. I expect this to grow.
    2. Labor demand will stall and then decrease, eventually falling to near zero. The timeline is unclear since it depends on exponential progress, geopolitics, and distribution. These are significant influences.
    3. This is a departure from traditional business economics between labor and capital. Expect political ramifications. Themes will emerge, but the AI market maturity is in the early days.
  3. The primary buyers are everyone. Opportunities are abundant.

IV. My advice for an AI enthusiast who wants to start creating?

TLDR - Learn. Make stuff.

  1. Work at whatever job pays you money. Look for opportunities to use AI.
  2. Learn how AI works.
  3. Build AI-powered solutions for your life and turn them into a portfolio. The goal is AI fluency.
  4. How would you talk to an AI system if you knew it was more intelligent than you? Talk to these AI chatbots like that. Ask the big questions. Be truthful. Be vulnerable with your hesitations and hang-ups. Ask AI what questions you should ask.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment