Something Big Is Happening. A personal note for non-tech friends and family on what AI is starting to change.

Reid Burke

AI capabilities have reached an inflection point where models can now autonomously complete complex, multi-hour tasks and even contribute to building newer versions of themselves. Recent releases like GPT-5.3 Codex and Claude Opus 4.6 demonstrate judgment-like decision-making that eliminates the need for human oversight in many knowledge work tasks. Industry leaders predict 50% of entry-level white-collar jobs could be eliminated within 1-5 years as AI becomes a general substitute for cognitive work across law, finance, medicine, software engineering, and other fields. The author urges immediate adoption of paid AI tools, daily experimentation, and financial preparation, emphasizing that the window to gain competitive advantage by being early is closing rapidly.

Something Big Is Happening

I know this is real because it happened to me first

Tens or hundreds of thousand lines of code isn’t a flex like you think it is. Code rabbit’s report is that AI generated code has 1.7 as many problems as human written code. Each line of code is a liability that must be maintained as well as a potential attack surface
AI has a ton of problems ahead of it too that remain completely unanswered:
<ul>
<li>Polluted training data. The early days of scrapping Reddit are over because most of the internet has AI generated content EVERYWHERE</li>
<li>Models aren’t having exponential progress anymore no matter how you slice it. With the current LLM approach, it looks like we are looking at diminishing returns</li>
<li>NOBODY has made a profit running an AI model. Not OpenAI, not Anthropic and not even Google. When that 20 subscription becomes a potential 200 dollar subscription for less usage, your ‘AI Knowledge’ becomes worthless as soon as you get rate limited</li>
</ul>

An AI-bro-Owner-of-an-AI-company, talks how AI is so great that it will replace everyone. The funnier part was comparing it somehow to covid19 pandemic.

“Obviously you’re not a golfer.” is probably the first thought that comes to mind after reading that
“Judging AI based on free-tier ChatGPT is like evaluating the state of smartphones by using a flip phone.” Now, let’s be reasonable. It would be more like by using a blackberry. Come on, false equivalence. But I digress.
Lot of mentioning “AI Models”, no specifying which kind… are we to presume LLM is what’s being considered the prima donna. No mention of neuro-symbolic or anything else.

What are you using that writes for four hours unsupervised?