A Trial Writer

A company reached out to me on Friday about working not as a Frontend Engineer but as a Technical Writer. I love the company, and I see promise in what they do so it would be a no brainer to decline. Based on their Core-Virtues there were some things I felt that really hit it off with me from the start.

Company Mission: “Our mission is to equip businesses with accessible AI technology so they can deliver transformational customer experiences.”

I can see this being used in my previous organization, filled with students wanted to learn AI for education or leverage tools to model without knowing the nitty gritty of Linear Regression or Stochastic Gradient Descent. My previous work could benefit off this once completed too, since they perform analysis on anomalies, and we used Linear Regression + spreadheets for callibration. Applying AI for detection and classification was always discussed but we lacked an expert.

  1. Understanding of the Target Audience (Customer Focused)
    • Students and Developers like me, who want to break into the AI industry but lack the experience. Likewise I understand the difficulties in trying to break into it.
    • “A.I tools are designed for experts”
  2. Continuous Improvement
    • Ex. I started off knowing nothing as a Full-Stack Engineer, non of it was taught at University. I learned it through developing applications, reading articles, and watching videos. Similarly, I want to apply this in my work at Mage. Understand how to use ML Models and then share it on Collab/Github. It’s a plus to be able to create a website where users can interact with the code itself and have a playground!
  3. Open to collaboration and cooperation (Victorious to team)
    • I have prior experience using Google Collab, sharing, importing data/libraries, and training models. When I took Applied ML in University (via Zoom). I collaborated a lot with TAs and students to ask around about what engine is the best for this dataset, what parameters to use to get a good accuracy. For my Intership at CROSS I worked with on Collab and Github to perform partioning on Pandas.
  4. Make AI accessible for all, whether large or small.
    • Its cool to see AI in action, but when everyday people try to pick it up or look stuff up about it its turn them off. They need to understand math formulas and symbols, which drives most people off. I want to create “blogs” as technical articles filled with personality. This makes it engaging to the reader and they don’t need to read or understand all the math. It’s 100% more friendly than reading the documentation of Unix Systems or a Butter-Bandpass filter.

Resonate: Coming from a Research and Labratory background, I’ve met professors who have spent their life studying Systems struggling to transition to AI when it first came out. This resulted in one department growing incredibly quick, whule everything slowed down for others departments, as resources (funding, supply of workers) become scarce if you weren’t in AI. Then, there was a problem where a research group wanted to apply AI there, but lacked a willing expert to build/train the model. On the other hand, there were plenty of experts that wanted to build/train, but lacked the data. My reearch department was a group that sampled data and had milestones to ship it to an AI/ML Data Scientist in the future.

What I really like about this position

  • I get to learn about AI. I get to code. I get to write with feeling and personality. Its a good balance of professional and informal. It’s not a Thesis Paper, but also not a text you would casually send a friend.

  • Potential to “level up” and evolve after building a strong foundation and understanding of all 40 models.

  • In university we often didn’t have enough time to do everything, usually we would gloss over things the teacher didn’t place a high importance on. Here, I’ll be able to touch practically everything currently released by a year or so. “Our roles don’t define our responsibilities” is a great team motto.

  • Turn documentation into an “Intelligence Network”. It’s currently being done at OpenAI, through the (beta) Open AI community forum. Tons of apps, help, and self-start guides everywhere with an active community of moderators and staff.
  • True quality comes from new members asking questions like this, and bright community responses.

What I want to see

  • Google Search Results for AI Terms (SGD, Pytorch Transformations) should not be scholarly articles or boring docuementation but links to user friendly blogs or technical writing.

Questions

  1. Are AI solutions currently overpriced? From what I’ve seen so far hiring an AI engineer, requires a PhD and in most cases asking for 2-3 publications?! There’s a low supply due to its time consuming and expensive barrier to entry.
    • b. How do you plan to make AI affordable? Is it a priced per usage kind of deal?

  2. AI is currently it right now, and theres a soft transition for VR and AR applications. Are there plans to explore those fields deeper in the future?

  3. During my time in university, one of my Capstone teachers tried to pitch and launch a new course called AI and Ethics. Can you tell me more about the company stance on “ethical” AI applications and how it’s enforced.

  4. Are you thinking of partnerships with other companies? OpenAI while its in beta has amazing applications too and their newest product Codex generates code from words. Combined with the articles (trained documentation) I think there’s a possibility of Codex being able to code an idea from scratch.

What’s the Trial

I’m going to be writing about AI and ML techniques, but in layman’s terms so anyone can understand. My goal is to have it be short, entertaining, and helpful regardless of prior knowledge. My first article will be on Regression Metrics!