Contemplative software engineer argues AI discussions lack context, creating false hype

A software engineer argues that the divide between AI enthusiasts and skeptics stems from a fundamental lack of context in how people describe their experiences with large language models. Writing on his blog “A Place Where Even Mammoths Fly,” Dmitrii “Mamut” Dimandt contends that the AI industry has become dominated by “magical, wishful thinking” similar to the cryptocurrency hype cycle, where questioning AI capabilities leads to accusations of being a “clueless moron.”

The core problem: Discussions about AI effectiveness lack crucial contextual information that would allow meaningful comparisons between different users’ experiences.

• People sharing their AI experiences rarely specify which projects they work on, what type of codebase they’re using (greenfield, mature, proprietary), or their level of expertise in the relevant domain.
• The amount of additional work required for reviewing, fixing, and deploying AI-generated code is typically unknown.
• AI systems are completely non-deterministic, meaning what works now may not work “even 1 minute from now for the exact same problem.”

Why this matters: The lack of context makes it impossible to compare a senior engineer’s experience with a greenfield React project to a non-coding designer working with proprietary OCaml code, yet both experiences are often treated as equivalent evidence.

What industry leaders are saying: The author criticizes vague testimonials that receive massive engagement despite providing no useful details.

• One widely shared post with 1.8k likes described Claude Code as “absolutely ruthless in chewing through legacy bugs” and “like a wood chipper fueled by dollars.”
• The post provided no information about codebase size, specific bugs addressed, programming languages used, or additional oversight required.

The author’s perspective: Despite being a critic, Dimandt actively uses AI tools across multiple projects, including building apps with Vercel’s v0, creating a SwiftUI monitoring app with Claude Code, and generating event posters with Midjourney.

• “Like most skeptics and critics, I use these tools daily.”
• His assessment: “50% of the time they work 50% of the time. It’s a non-deterministic statistical machine.”
• He argues the discourse incorrectly assumes AI tools are either “magic” or “engineering” when they’re actually statistical machines that can feel magical when they work.

The bigger picture: The author draws parallels between current AI hype and the cryptocurrency boom, suggesting that critical thinking about AI capabilities is being suppressed by industry enthusiasm and social pressure to embrace the technology unconditionally.