In the pre-AI world most of the things devs used to do were local; hence affordable.
No doubt that some models like of Gemma or others can easily be run on edge devices but sticking being an AI native Engineer will (as of now atleast) require most of the stuff to be in the 3rd party hands and be paid 💰.
In the web-dev world, we are fighting for 500kb js bundle to run on edge device or server for 20 years, resulting in SSR, SSG, SPA, and other variants.
What is your take on the computational expenses, an AI native engineer has to deal with?
This article deeply resonated with me. As a former Agile Coach now on the journey to becoming an AI-Native Engineer, these points particularly struck a chord:
- The importance of a growth mindset, constantly learning and evolving.
- The significance of inspect and adapt, a core principle of Scrum.
- The perspective of Amplification, not just Augmentation, moving beyond simple capability enhancement.
- The proactivity to drive towards goals, rather than being led by tools.
The fact that becoming an AI-Native Engineer is a continuous journey, not a one-time transformation.
These points align perfectly with what I have always considered important.
Reading your article also brought more clarity to a thought I've been having recently. I believe that our relationship with AI can be viewed from the perspective of delegation, much like the 7 Levels of Delegation (https://management30.com/practice/delegation-poker/). Just as leaders must carefully consider which decisions to delegate to their teams, we need to thoughtfully determine which tasks we can entrust to AI and progressively build that trust.
Furthermore, I was greatly inspired by your previous post, "Vibe-Coding is not an excuse for low-quality work," which led me to create a GitHub repository on the topic of 'Agentic-Coding' (https://github.com/agentic-coding/agentic-coding.github.com) to further develop these ideas.
Your writing has truly sparked new ideas in my journey. Looking forward to learning more from your future insights!
Nice read.
But I am more worried about computational costs.
In the pre-AI world most of the things devs used to do were local; hence affordable.
No doubt that some models like of Gemma or others can easily be run on edge devices but sticking being an AI native Engineer will (as of now atleast) require most of the stuff to be in the 3rd party hands and be paid 💰.
In the web-dev world, we are fighting for 500kb js bundle to run on edge device or server for 20 years, resulting in SSR, SSG, SPA, and other variants.
What is your take on the computational expenses, an AI native engineer has to deal with?
Excellent analysis and subject matter, its a question of holding on for as long as you aren't in the way.
This article deeply resonated with me. As a former Agile Coach now on the journey to becoming an AI-Native Engineer, these points particularly struck a chord:
- The importance of a growth mindset, constantly learning and evolving.
- The significance of inspect and adapt, a core principle of Scrum.
- The perspective of Amplification, not just Augmentation, moving beyond simple capability enhancement.
- The proactivity to drive towards goals, rather than being led by tools.
The fact that becoming an AI-Native Engineer is a continuous journey, not a one-time transformation.
These points align perfectly with what I have always considered important.
Reading your article also brought more clarity to a thought I've been having recently. I believe that our relationship with AI can be viewed from the perspective of delegation, much like the 7 Levels of Delegation (https://management30.com/practice/delegation-poker/). Just as leaders must carefully consider which decisions to delegate to their teams, we need to thoughtfully determine which tasks we can entrust to AI and progressively build that trust.
Furthermore, I was greatly inspired by your previous post, "Vibe-Coding is not an excuse for low-quality work," which led me to create a GitHub repository on the topic of 'Agentic-Coding' (https://github.com/agentic-coding/agentic-coding.github.com) to further develop these ideas.
Your writing has truly sparked new ideas in my journey. Looking forward to learning more from your future insights!