A Developers Guide to Judgment, Trust, and Agency in the Age of AI: 10 evidence-based practices and methodology (Video Course)

Description

The first video course built on Asymmetric Collaboration, the only AI development methodology extracted from a fully documented, real-world project, teaching professionals how to work with AI without ceding judgment, agency, or craft.

Most AI development content teaches tools. This course teaches a methodology. Asymmetric Collaboration is a conceptual framework that reframes the humanAI relationship accurately: Humans and AI do not contribute the same things at different levels. They contribute fundamentally different things, and the collaboration only works when both sides are understood clearly.

This course is built on about 100 days of documented, real-world AI-assisted software development, 500+ commits, development journals, dual-perspective session logs, and quantified productivity metrics. The result: 10 transferable practices organized into four progressive layers, each grounded in specific project evidence, including the failure cases that generated the practices in the first place.

Learners leave with a structured framework they can apply immediately: a way to communicate intent precisely, verify AI outputs systematically, maintain human judgment where it matters, calibrate trust based on evidence, and protect their agency as AI handles more execution. The methodology is tool-agnostic. Whether learners use GitHub Copilot, Claude, ChatGPT, or any successor, the 10 practices apply.  .

Skill Level:

• Intermediate

Learn How To:

• Apply the Asymmetric Collaboration framework to any AI development workflow
• Communicate intent precisely so AI output reflects what you actually need
• Run incremental verification cycles that prevent compounding AI errors
• Maintain human judgment on architectural and consequential decisions while delegating execution to AI
• Build and maintain a trust calibration record for AI outputs by domain and error type
• Surface AI misalignment early before it becomes expensive to correct
• Identify and protect explicit decision points that AI velocity tends to bypass
• Produce context-enriched deliverables that transfer knowledge AI cannot infer
• Sustain professional agency and craft across an extended AI-assisted collaboration

Course requirement:

Individuals should have knowledge and experience with:

• Professional experience in software development or a knowledge-work domain (technical writing, analysis, localization)
• Familiarity with at least one AI coding assistant or generative AI tool (any: GitHub Copilot, Claude, ChatGPT, Cursor, etc.)
• No specific programming language is required; examples are language-agnostic
• No academic background in AI required; the methodology is practitioner-facing

Who Should Take This Course:

Primary Audience

• Software developers and engineers who are already using AI coding assistants and finding that productivity gains come with new forms of friction, quality risk, and loss of clarity about who is really making decisions. They know how to code. What they need is a structured framework for collaboration, not more prompting tips.

Secondary Audience

• Technical leads, architects, and senior contributors who are evaluating AI integration for their teams and need a methodology grounded in documented evidence rather than vendor claims or speculation. Also valuable for knowledge workers in adjacent fields, technical writers, localization specialists, analysts who are adopting AI tools and want a principled approach to maintaining quality and judgment.

About Pearson Video Training:

Pearson publishes expert-led video tutorials covering a wide selection of technology topics designed to teach you the skills you need to succeed. These professional and personal technology videos feature world-leading author instructors published by your trusted technology brands: Addison-Wesley, Cisco Press, Pearson IT Certification, Prentice Hall, Sams, and Que. Topics include IT Certification, Network Security, Cisco Technology, Programming, Web Development, Mobile Development, and more.  Learn more about Pearson Video training at  http://www.informit.com/video.

Video Lessons are available for download for offline viewing within the streaming format. Look for the green arrow in each lesson.

Sample Content

Table of Contents

Introduction

Lesson 1: The Asymmetric Collaboration Framework

Learning objectives

1.1 Understand why asymmetric collaboration differs from standard productivity frameworks

1.2 Interpret the evidence base: project data, commit history, and what it proves

1.3 Apply Practice 1: communicate intent precisely through iterative prompting

1.4 Apply Practice 2: verify AI outputs incrementally to prevent error compounding

1.5 Observe P1 and P2 applied together in a live real-task demonstration

Lesson 2: Daily Work: Communication, Velocity, and Judgment   

Learning objectives

2.1 Apply Practice 3: regulate AI collaboration velocity to preserve comprehension and control

2.2 Apply Practice 4: surface AI misalignment early before it becomes expensive to correct           

2.3 Analyze the Docker Cache Incident as a case study in late misalignment detection          

2.4 Apply Practice 5: retain human judgment on architectural and high-abstraction decisions

2.5 Recognize the early signs of architectural drift in an AI-assisted project

Lesson 3: Trust, Error, and Control

Learning objectives

3.1 Apply Practice 6: calibrate AI trust through structured documentation of outputs over time

3.2 Build a domain-specific trust map that reflects observed AI reliability by task category

3.3 Apply Practice 7: attribute AI errors systematically to identify patterns and recovery strategies

3.4 Distinguish the four AI error categories and select the appropriate recovery strategy for each

3.5 Apply Practice 8: define and protect explicit decision points that AI velocity tends to bypass

3.6 Explain the control paradox and why unchecked velocity can feel like productive momentum

Lesson 4: Transfer and Sustained Agency

Learning objectives

4.1 Apply Practice 9: produce context-enriched deliverables that transfer knowledge AI cannot infer

4.2 Analyze the i18n case study: eliminating context errors across three languages using P9

4.3 Extend Practice 9 to non-software knowledge work contexts

4.4 Apply Practice 10: sustain professional agency and craft across an extended AI collaboration

4.5 Assess your own irreplaceable contribution using the human-value self-assessment framework

Lesson 5:  Integration: A Full Collaboration Session

Learning objectives

5.1 Identify consequential decisions in advance of a collaboration session using a pre-session protocol

5.2 Observe all 10 practices operating in a real AI-assisted work session: annotated walkthrough, Part 1

5.3 Observe all 10 practices operating in a real AI-assisted work session: annotated walkthrough, Part 2

5.4 Debrief the session and identify what each practice protected against

5.5 Evaluate what changes and what does not when AI handles more execution in professional work

Summary

Updates

Submit Errata