scripod.com
How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO)

Overview

Shownote

Highlights

Transcript

Chapters

Pins

How Devin replaces your junior engineers with infinite AI interns that never sleep | Scott Wu (Cognition CEO)

How I AI

2025/09/08
How I AI

How I AI

2025/09/08
OverviewShownoteHighlightsTranscriptChaptersPins

Shownote

Scott Wu is the co-founder and CEO of Cognition Labs, the creators of Devin, an AI agent designed to function as a junior engineer on software development teams. In this conversation, Scott demonstrates how his team uses their own product to accelerate dev...

Highlights

In this episode, Claire Vo sits down with Scott Wu, co-founder and CEO of Cognition Labs, to explore how AI is reshaping software engineering workflows. The discussion centers around Devin, an AI agent developed by Cognition Labs, which functions as a capable junior engineer capable of handling routine development tasks. Scott walks through how his team integrates AI into their daily operations, using tools like Slack, DeepWiki, and GitHub to delegate, monitor, and collaborate with AI in a seamless and asynchronous manner.
02:55
Devin acts like an engineer on the team, different from other tools
03:53
Devin excels at executing clear tasks rather than solving complex architectural problems.
08:55
Constructing a prompt for Devin using existing MCP server details
10:27
Refining prompts contextually can make interactions more efficient
11:24
Devin AI works asynchronously, reducing user frustration with latency
13:38
Devin can make pull requests on GitHub, show diffs, and send screenshots for changes.
14:45
Devin streamlines tasks like research and documentation updates asynchronously.
18:39
Devin collaborates with human engineers on tasks like standardizing fonts or adding features.
23:27
Public use of AI agents fosters team learning and agent improvement.
25:50
Devin handles front-end fixes, version upgrades, and writes complete documentation.
34:32
ChatGPT Voice reduces search time from 10 seconds to 1-2 seconds.
38:38
Reviewing AI agent history to identify errors and re-instructing them is key to effective use.

Chapters

Introduction to Scott Wu and Devin
00:00
Where Devin excels
03:53
Using DeepWiki to research codebases and create better prompts
06:08
Prompting tips
10:27
The asynchronous nature of working with Devin
11:24
Multithreading tasks
13:38
Using Devin to implement an MCP server integration
14:43
Setting up workflows in Slack for first-line responses
18:38
Encouraging AI adoption in public Slack channels
23:22
Top five engineering tasks for Devin
25:50
Using ChatGPT voice as a meeting participant
32:17
Lightning round
35:57

Transcript

Scott Wu: Devin is async. Once you kick off the Devin session, Devin's gonna start working and looking through the code, but you're not expected to be there with it. It's just as if you gave your intern a project and your intern is going and working on it....