---
title: "Dense Slides: No Vibes Allowed: Solving Hard Problems in Complex Codebases – Dex Horthy, HumanLayer"
category: "slides"
video_id: "rmvDxxNubIg"
sourceLabels: ["Captured video frames", "Local OpenCV slide-region detection"]
---

# Dense Slides: No Vibes Allowed: Solving Hard Problems in Complex Codebases – Dex Horthy, HumanLayer

## Source Video
[No Vibes Allowed: Solving Hard Problems in Complex Codebases – Dex Horthy, HumanLayer](https://www.youtube.com/watch?v=rmvDxxNubIg)

## Method
This deck is slide-only. The existing captured video frame set supplies candidate frames, then local OpenCV rejects sponsor/title/speaker-only frames, crops visible slide surfaces, deduplicates, and saves the cropped slide images.

## Cropped Visible Slides
![[assets/dense-slides/rmvDxxNubIg/slide-001.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/dense/rmvDxxNubIg/slide-001.html)
- AI slide classifier: `content_slide` confidence `0.99`
- Text source: advanced OCR `rapidocr-live/bright-screen/opencv-adaptive`.
- OCR decision: ready — Dense quadrant diagram with multiple labels and small text; OCR is the cheaper/more reliable path.

Slide text:

> Task Complexity High Complex tasks require deeper human insight +10-15% Constrained by outdated code & intricate dependencies +0-10%
> Low These tasks are often repetitive and well-defined +35-40% Legacy projects still benefit on simpler tasks +15-20%
> :Greenfield Project Maturity.. Brownfield:
> AIE/CODE CODEBASES NO VIBES ALLOWED: SOLVING
> Google DeepMind Presenied dy DEX HORTHY / Founder & CEO. HumanLayer.

![[assets/dense-slides/rmvDxxNubIg/slide-002.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/dense/rmvDxxNubIg/slide-002.html)
- AI slide classifier: `content_slide` confidence `0.98`
- Text source: advanced OCR `rapidocr-live/bright-screen/contrast`.
- OCR decision: ready — Multi-column diagram with small labels and a quote; OCR should recover details better than manual transcription in this pass.

Slide text:

> Syiten Iuatructlons CASE.Od Systee Instruetltes CALOE.SE
> Cloode BalItin Tools Intentional Compaction Clonte lulltlo Tools NOTeol
> User mssoe "Write everything we did Reod proyress.nd User sessoge
> Asslstam Messoje Great,row Aac Uuer nesroge Reod) Sesrch) Eritet) we're working on" we're taking, the steps we've done so far, so far to progress.md, and the current failure ensure to note the approach pregreas.nd sritel Briteo Reodt)
> ritet)
> Assistontlesscge
> morize progreas to progress.nd User nessoge progress,nd
> no vibes allowed adexhorthy Code Summit
> AIE /CODE graph Mahesh Murag ANTHROPIC Natalie Serrino gimlet Eno Reyes FACTORY Lee Robinson CURSOR Nik Pash cline
> Google DeepMind PRESENTEDBY Cognition 采FACTORY Graphite:e04j autho Worko

![[assets/dense-slides/rmvDxxNubIg/slide-003.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/dense/rmvDxxNubIg/slide-003.html)
- AI slide classifier: `content_slide` confidence `0.95`
- Text source: agent_vision.

Slide text:

> Hyper Engineering
> Context engineering for coding agents

![[assets/dense-slides/rmvDxxNubIg/slide-004.jpg]]

- Recreated text/layout view: [open HTML recreation](/assets/slide-recreations/dense/rmvDxxNubIg/slide-004.html)
- AI slide classifier: `content_slide` confidence `0.97`
- Text source: agent_vision.
- OCR decision: ready — Code diff / test result screenshot with small text and UI labels; OCR is appropriate.

Slide text:

> test #1252


Classification audit: `raw/sources/slide-ai-classification/dense/rmvDxxNubIg/audit.json`