Evidence

Format Web Page

"AI coding tools can generate convincing prototypes quickly while leaving inexperienced makers struggling to debug and finish them."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title The 70% problem - Hard truths about AI-assisted coding

Source authors Addy Osmani

Source year 2024

Confidence high

Importance mid

Format Web Page

"Rapid AI-assisted prototyping can encourage feature creep and additional work rather than simply reducing the developer’s workload."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title 10 things I learned from burning myself out with AI coding agents

Source authors Benj Edwards

Source year 2026

Confidence high

Importance mid

Format Online paper / PDF

"Language models can have more difficulty retrieving relevant information from the middle of a long input than from its beginning or end."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title Lost in the Middle - How Language Models Use Long Contexts

Source authors Nelson F. Liu; Kevin Lin; John Hewitt; Ashwin Paranjape; Michele Bevilacqua; Fabio Petroni; Percy Liang

Source year 2023

Confidence high

Importance mid

Format Online paper / PDF

"LLM performance can become less reliable as input contexts grow, even when models nominally support long context windows."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title Context Rot - How Increasing Input Tokens Impacts LLM Performance

Source authors Kelly Hong; Anton Troynikov; Jeff Huber

Source year 2025

Confidence high

Importance mid

Format Online paper / PDF

"Csikszentmihalyi’s concept of flow centers on an absorbing state in which task challenges are matched to a person’s abilities."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title A Scoping Review of Flow Research

Source authors Corinna Peifer et al.

Source year 2022

Confidence high

Importance high

Format Online paper / PDF

"Multiline slot-machine play can produce dark flow, a highly absorbing state associated with repetitive gambling activity."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title Dark Flow, Depression and Multiline Slot Machine Play

Source authors Mike J. Dixon; Madison Stange; Chanel J. Larche; Candice Graydon; Jonathan A. Fugelsang; Kevin A. Harrigan

Source year 2017

Confidence high

Importance high

Format Online paper / PDF

"In multiline slot machines, a payout smaller than the original wager can be accompanied by celebratory feedback resembling a win."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title Losses disguised as wins in modern multi-line video slot machines

Source authors Mike J. Dixon; Kevin A. Harrigan; Rajwant Sandhu; Karen Collins; Jonathan A. Fugelsang

Source year 2010

Confidence high

Importance high

Format Web Page

"Nearly 30% of senior developers in Fastly’s 2025 survey reported that editing and auditing AI output offset most of their initial time savings."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title Vibe Shift? Senior Developers Ship nearly 2.5x more AI Code than Junior Counterparts

Source authors Alina Lehtinen-Vela

Source year 2025

Confidence high

Importance high

Format Web Page

"CodeRabbit reported 1.4 times as many critical issues in AI-generated code within its comparative sample."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title A semantic history - How the term 'vibe coding' went from a tweet to prod

Source authors David Kravets

Source year 2026

Confidence mid

Importance mid

Format Web Page

"CodeRabbit reported 1.7 times as many bugs and quality issues in AI-generated code within its comparative sample."

View Source

Location PART 3 - THE THIRD AXIS CHAPTER 15: DEPLOY NOW, PRAY LATER Gambling Builds No Skill

Source title A semantic history - How the term 'vibe coding' went from a tweet to prod

Source authors David Kravets

Source year 2026

Confidence mid

Importance mid