My Journey with AI: From Excited Explorer to Concerned Observer
A Gradual Embrace: My First Encounters with AI
I've been fascinated by the rise of AI for quite some time. As a teenager, it felt like this technology was rapidly becoming a normal part of everyday life, starting with the buzz around platforms like ChatGPT. My first real hands-on experience came with Copilot in the Edge browser – and honestly, it felt like a monumental leap forward for the browser, especially compared to its older version.
For someone with ADHD like me, navigating websites could often be a frustrating task. Copilot was surprisingly helpful in finding things I would otherwise miss, offering a much more intuitive way to explore online content. However, back then, its use felt somewhat limited.
Diving Deeper: Privacy and Local LLMs
Over the past year or so, my engagement with AI has shifted. I've moved beyond casual exploration and started actively trying to understand how these models work and how I can use them effectively – all while prioritizing my privacy. Before exploring local Large Language Models (LLMs), I’ve been using AI for practical tasks like proofreading my writing, outlining project ideas, and gaining clarity on concepts where my knowledge is lacking – things that used to require human interaction.
As I've become more aware of the implications of adulthood, protecting my personal information has become a major concern. The risk of sensitive data being exploited for social engineering or identity theft has led me down the path of local LLMs. This allows me to leverage AI power without exposing my private details. My go-to models are Google Gemma 3 and a DeepSeek model specifically configured for my hardware, with careful attention paid to token usage to ensure I can effectively utilize each one. This entire journey has been driven by a genuine curiosity about AI and a personal interest in exploring its potential.
A Growing Unease: My Concerns About the AI Overwhelm
Initially, the promise of AI was exciting – the idea of a more human-like way to access information was appealing. However, my enthusiasm has slowly been replaced by a growing unease. I'm increasingly concerned about the pervasive and often forced integration of AI into almost every facet of technology.
It feels like major tech companies – Google, Apple, Microsoft, Facebook, and pretty much everyone except Apple – are relentlessly pushing AI into their products and services. It’s estimated that a significant portion (20-30%) of their new code is now AI-generated. While the code generation itself isn't necessarily a bad thing, I've witnessed firsthand some of the issues that can arise. For example, the recent widespread outage on Google's AI services seemed to stem from an "AI-like bug."
My own experience with AI-assisted coding has often involved extensive debugging and patching up errors – a constant need for human oversight to ensure reliability and proper testing. These experiences have reinforced my belief that while AI has potential, it shouldn’t be the sole driver of development.
The Future I Plan: To have to only use local llm’s and a Worried Outlook
Ultimately, my goal with AI is to leverage its power locally. I don't see any problem with using AI as a tool. My concern lies in the feeling that AI is being aggressively pushed onto us without fully considering the implications. I also have a deeper worry about the future of the field I aspire to work in – a fear that the increasing reliance on AI could diminish the value and opportunity for human expertise.