liqiang
b119135836
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s
272 lines
8.7 KiB
Markdown
272 lines
8.7 KiB
Markdown
---
|
||
theme: seriph
|
||
title: "Lesson 42 — How Do Agent Teams Fail—and What Should We Build Next?"
|
||
info: "English video course for AI Agents in Depth"
|
||
author: Bojie Li
|
||
transition: slide-left
|
||
mdc: true
|
||
lineNumbers: false
|
||
monaco: false
|
||
aspectRatio: 16/9
|
||
canvasWidth: 980
|
||
layout: cover
|
||
class: cover
|
||
---
|
||
|
||
<div class="course-kicker">Expand · Chapter 10 · Multi-Agent Collaboration</div>
|
||
|
||
# How Do Agent Teams Fail—and What Should We Build Next?
|
||
|
||
<p class="course-subtitle">Conflicts, error cascades, Agent societies, and the course synthesis</p>
|
||
|
||
<div class="course-cover-meta">Lesson 42 of 42 · 18 minutes · Failure Modes; Agent Society; Economic Competition; Strategic Gameplay</div>
|
||
|
||
<!-- Presenter cue: Open with the concrete problem. Add personal context in your own words; do not read the slide. -->
|
||
|
||
---
|
||
layout: center
|
||
class: text-center
|
||
---
|
||
|
||
<div class="course-kicker">The central question</div>
|
||
|
||
<div class="course-big">How do we prevent local mistakes from becoming group failures while still allowing collective behavior to emerge?</div>
|
||
|
||
<!-- Presenter cue: Let the question sit for a moment, then state the failure mode the lesson will explain. -->
|
||
|
||
---
|
||
|
||
# Why this problem matters
|
||
|
||
<div class="grid grid-cols-3 gap-5 mt-6">
|
||
<div class="course-card blue">
|
||
<h3>Concurrency</h3>
|
||
<p>Shared files can lose updates or encode semantic conflicts.</p>
|
||
</div>
|
||
<div class="course-card green">
|
||
<h3>Cascades</h3>
|
||
<p>A wrong upstream claim is amplified by trusting downstream Agents.</p>
|
||
</div>
|
||
<div class="course-card orange">
|
||
<h3>Emergence</h3>
|
||
<p>Persistent Agents produce social, strategic, and economic behavior not explicitly scripted.</p>
|
||
</div>
|
||
</div>
|
||
|
||
<!-- Presenter cue: Connect each card to a product or experiment consequence. -->
|
||
|
||
---
|
||
|
||
# Three ideas to keep in view
|
||
|
||
<div class="grid grid-cols-3 gap-5 mt-6">
|
||
<div class="course-card purple">
|
||
<h3>Optimistic locking</h3>
|
||
<p>Detect version change before committing a shared write</p>
|
||
</div>
|
||
<div class="course-card blue">
|
||
<h3>Information control</h3>
|
||
<p>A code judge reveals only what each role may know</p>
|
||
</div>
|
||
<div class="course-card green">
|
||
<h3>External reward</h3>
|
||
<p>Society outcomes are scored by the environment—not self-report</p>
|
||
</div>
|
||
</div>
|
||
|
||
<!-- Presenter cue: Define unfamiliar terms in plain language; the audience is new to ML training and RL. -->
|
||
|
||
---
|
||
|
||
# The book's visual model
|
||
|
||
<img class="course-figure" src="/images/fig10-11.svg" alt="Voice Werewolf multi-Agent system">
|
||
|
||
<div class="course-caption">Voice Werewolf multi-Agent system</div>
|
||
|
||
<!-- Presenter cue: Trace the diagram in one direction and name the mechanism that matters for this lesson. -->
|
||
|
||
---
|
||
|
||
# Agent chat room vs. Governed society
|
||
|
||
<div class="grid grid-cols-2 gap-6 mt-5">
|
||
<div class="course-card orange"><h3>Agent chat room</h3><ul><li>Everyone sees everything</li><li>Loose role prompts</li><li>Claims spread unchecked</li></ul></div>
|
||
<div class="course-card green"><h3>Governed society</h3><ul><li>State authority in code</li><li>Role-scoped views</li><li>Auditable actions and rewards</li></ul></div>
|
||
</div>
|
||
|
||
<div class="course-caption course-caption-strong">Social complexity requires stronger state and information governance.</div>
|
||
|
||
<!-- Presenter cue: Explain the trade-off; avoid presenting the right column as universally superior. -->
|
||
|
||
---
|
||
|
||
# The judge owns truth and disclosure
|
||
|
||
~~~python
|
||
private_view = judge.view_for(player, global_state)
|
||
action = player.act(private_view)
|
||
judge.validate(action, role=player.role)
|
||
global_state = judge.apply(action)
|
||
audit.append(player.id, action, state_hash(global_state))
|
||
~~~
|
||
|
||
<!-- Presenter cue: Walk through the executable idea line by line; keep implementation details for the terminal. -->
|
||
|
||
---
|
||
|
||
# Test the claim
|
||
|
||
<div class="grid grid-cols-1 gap-4 mt-5">
|
||
<div class="course-card blue">
|
||
<div class="course-demo-head"><span>10-8 offline diagnostic</span><span>2 min</span></div>
|
||
<h3>Run a deterministic information-isolation game</h3>
|
||
<p><strong>Observe:</strong> Private role context, phase transitions, legal actions, votes, and winner gates</p>
|
||
</div>
|
||
</div>
|
||
|
||
<div class="course-caption course-caption-strong">Demo budget: 2 minutes · one contiguous terminal block</div>
|
||
|
||
<!-- Presenter cue: State the prediction before running anything. Name the observation that could disconfirm it. -->
|
||
|
||
---
|
||
class: course-terminal
|
||
---
|
||
|
||
<div class="course-kicker">Live demo</div>
|
||
|
||
# Switching to the terminal
|
||
|
||
~~~bash
|
||
$ cd chapter10/voice-werewolf && python demo.py --offline
|
||
~~~
|
||
|
||
<div class="course-terminal-watch">Run the command(s), narrate decisions, and point to the observation—not just the output.</div>
|
||
|
||
<!-- Presenter cue: Switch windows now. Return to the next slide after every listed experiment is complete. -->
|
||
|
||
---
|
||
|
||
# What the evidence supports
|
||
|
||
<div class="grid grid-cols-3 gap-5 mt-6">
|
||
<div class="course-card green">
|
||
<h3>Finding 1</h3>
|
||
<p>Coordination failures are distributed-systems failures plus probabilistic decision errors.</p>
|
||
</div>
|
||
<div class="course-card blue">
|
||
<h3>Finding 2</h3>
|
||
<p>A code-driven authority can preserve information asymmetry and rule integrity.</p>
|
||
</div>
|
||
<div class="course-card purple">
|
||
<h3>Finding 3</h3>
|
||
<p>Social and economic simulations can generate new experience—but also collusion and pathology.</p>
|
||
</div>
|
||
</div>
|
||
|
||
<!-- Presenter cue: Tie each finding to something viewers just observed; distinguish evidence from interpretation. -->
|
||
|
||
---
|
||
layout: center
|
||
---
|
||
|
||
<div class="course-kicker course-kicker-red">Where the claim stops</div>
|
||
|
||
# Boundary condition
|
||
|
||
<div class="course-boundary">An offline all-AI diagnostic does not satisfy the book's live human voice acceptance criteria.</div>
|
||
|
||
<!-- Presenter cue: Say explicitly what this experiment does not establish. -->
|
||
|
||
---
|
||
layout: center
|
||
---
|
||
|
||
<div class="course-kicker">Engineering takeaway</div>
|
||
|
||
# Design rule
|
||
|
||
<div class="course-rule">Keep shared truth, permissions, conflict detection, and final rewards outside the Agents that compete or collaborate.</div>
|
||
|
||
<!-- Presenter cue: Present this as a reusable decision rule, then give one counterexample or trade-off. -->
|
||
|
||
---
|
||
|
||
# Continue the experiment
|
||
|
||
<div class="grid grid-cols-2 gap-4 mt-6">
|
||
<a class="course-link" href="https://github.com/joonspk-research/generative_agents">
|
||
<span class="course-link-title">Experiment 10-7: Stanford Generative Agents</span>
|
||
<span class="course-link-path">https://github.com/joonspk-research/generative_agents</span>
|
||
</a>
|
||
<a class="course-link" href="../chapter10/voice-werewolf/README.md">
|
||
<span class="course-link-title">Experiment 10-8: consent-gated voice path</span>
|
||
<span class="course-link-path">chapter10/voice-werewolf/README.md</span>
|
||
</a>
|
||
<a class="course-link" href="../book-en/images/fig10-10.svg">
|
||
<span class="course-link-title">AI Town architecture</span>
|
||
<span class="course-link-path">book-en/images/fig10-10.svg</span>
|
||
</a>
|
||
<a class="course-link" href="../book-en/chapter10.md">
|
||
<span class="course-link-title">Agent society and economy cases</span>
|
||
<span class="course-link-path">book-en/chapter10.md</span>
|
||
</a>
|
||
</div>
|
||
|
||
<!-- Presenter cue: Point viewers to the companion paths; do not walk through every extension. -->
|
||
|
||
---
|
||
|
||
# The complete course arc
|
||
|
||
<div class="grid grid-cols-3 gap-5 mt-6">
|
||
<div class="course-card blue">
|
||
<h3>Build · Lessons 01–21</h3>
|
||
<p>Context → memory → tools → code</p>
|
||
</div>
|
||
<div class="course-card green">
|
||
<h3>Improve · Lessons 22–34</h3>
|
||
<p>Evaluation → training → continual evolution</p>
|
||
</div>
|
||
<div class="course-card purple">
|
||
<h3>Expand · Lessons 35–42</h3>
|
||
<p>Voice → embodied action → collaboration</p>
|
||
</div>
|
||
</div>
|
||
|
||
<!-- Presenter cue: Return to the three-part arc from Lesson 1 and connect each stage to evidence viewers saw in the course. -->
|
||
|
||
---
|
||
layout: center
|
||
class: text-center
|
||
---
|
||
|
||
<div class="course-kicker">Pause and apply</div>
|
||
|
||
# Your turn
|
||
|
||
<div class="course-big course-reflection">Which group-level failure cannot be prevented by improving any single Agent in isolation?</div>
|
||
|
||
<!-- Presenter cue: Invite viewers to pause the video. Offer your own answer after a short beat. -->
|
||
|
||
---
|
||
layout: center
|
||
class: text-center
|
||
---
|
||
|
||
<div class="course-kicker">Course synthesis</div>
|
||
|
||
<div class="course-loop mt-8">
|
||
<div class="course-loop-step blue"><span>1</span><strong>Define the failure</strong></div>
|
||
<div class="course-loop-arrow">→</div>
|
||
<div class="course-loop-step green"><span>2</span><strong>Run a controlled experiment</strong></div>
|
||
<div class="course-loop-arrow">→</div>
|
||
<div class="course-loop-step purple"><span>3</span><strong>Interpret the evidence</strong></div>
|
||
<div class="course-loop-arrow">→</div>
|
||
<div class="course-loop-step orange"><span>4</span><strong>Update safely</strong></div>
|
||
</div>
|
||
|
||
<div class="course-loop-return">↺ Repeat when new evidence arrives</div>
|
||
|
||
<!-- Presenter cue: Close by tracing the evidence-driven loop from Lesson 1, then leave viewers with the repeat trigger. -->
|