One AI Department Done. 3 to Go.
Want to back the stream? The «داعم» membership on YouTube is EGP 60 a month: badges, emoji, and I answer your questions first on stream. Become a supporter
The gist
The stream centers on a work-in-progress automation system that organizes AI agents like connected departments for research, coordination, content production, publishing, and analytics. Luqman explains a possible progression from scheduled tasks and role-specific instructions toward shared goals and indicators, with research handed to agents that adapt posts, articles, and short videos for each platform. Practical work includes reviewing the post-stream pipeline, trimming silence and adjusting clips, and preparing operations across LinkedIn, X, Facebook, and Instagram while retaining human review initially. Speed, quality, resource use, queues, measurement, and agent autonomy still require testing and refinement; the stream does not establish a successful end-to-end system. It also discusses distribution and building in public, comparing direct outreach with SEO, social content, and advertising, alongside the speaker’s view that understanding an audience and solving a real problem matter more than selling a generic chatbot. Audience questions broaden the discussion to changing developer roles, learning and working with AI, game development, video-editing tools, background or local agent execution, community platforms, tools, and subscriptions. Other cautious, unresolved topics include AI jobs and pricing, proposed Egyptian data centers, model and hardware performance, YouTube Shorts growth, and ideas for competitions and products; these are generally presented as opinions, proposals, or open questions rather than verified findings.
171 moments from the stream
- 4:36:23
Finding Energy for Personal Projects After Work
An audience discussion considers whether time and energy remain for personal projects after a workday, with the view that passion can help someone continue despite feeling drained.
- 4:37:02
Defining a Platform-Specific Content Strategy
The speaker discusses turning an ambiguous output range into a clearer plan, distinguishing content by account and platform, and determining the required story volume before setting a publishing cadence.
- 4:40:31
Using AI to Assess Skills and Improve LinkedIn
A participant describes giving an AI their experience and asking it to test each skill while preparing a LinkedIn profile. The speaker supports the approach and mentions using Claude to help improve his own profile.
- 4:41:20
Testing a Connected Prototype During the Day
A participant describes testing a prototype during the workday so its logs collect information that can support later development, with the mobile interface connected to a backend on another machine.
- 4:43:59
Questioning Publishing Cadence Instead of Accepting It
The speaker reviews a proposed publishing cadence and asks for the reasoning behind differences across newsletters, platforms, and accounts rather than accepting suggested frequencies automatically.
- 4:45:30
Is Obsidian Worth Using as a “Second Brain”?
A qualified response to an audience question: the speaker did not claim hands-on experience, but said he had heard good things about Obsidian and that it could be useful for someone with time to get through its learning curve.
- 4:48:59
A Daily Publishing Cadence for an AI News Outlet
A proposal to cover three distinct stories or updates across the morning, midday, and afternoon, resulting in three publishing slots per day.
- 4:49:23
Learning New Skills to Build Your Project
Advice to assess one’s willingness to learn new things for a project, then ask for the possibilities that would become available after gaining those skills.
- 4:54:11
What an “Extremely Aggressive” Media Growth Strategy Means
The aggressive strategy is interpreted as running a daily media operation with high publishing volume, fast coverage, original material, and active distribution, including social-platform publishing branches such as Instagram and Facebook.
- 4:55:56
Balancing Content Volume, Quality, and Operating Cost
The speaker argues that publishing more may drive additional traffic, but it also risks weaker originality and quality plus higher token consumption. The goal is a scalable production system that preserves content standards rather than relying on volume alone.
- 4:57:41
Using Engagement on Viral Content to Attract Traffic
An agent is proposed to monitor viral content on X and consistently post selected comments intended to attract likes and traffic, treating comment-section engagement as a possible traffic channel. It remains an idea; no test results are presented.
- 5:01:56
Publishing Through an Intermediary Tool Instead of the X API
Responding to a question about the X API, the speaker said at the time that he did not need a direct integration because an intermediary publishing tool connected to the required platforms. He also described certain link features as having pricing and usage limits that did not seem worthwhile for his situation.
- 5:03:08
Defining the Post Mix and Format for Each Platform
The discussion stresses that a steady flow of valuable content is essential to keep the rest of the marketing funnel working, then calls for a platform-by-platform breakdown of post type, whether each item includes video or imagery or is text-only, and how often it appears. The aim is to turn broad targets into clear post specifications.
- 5:05:31
Traffic Channels and Automated Content Distribution
A discussion of diversifying traffic acquisition through outbound links, SEO, listings, analytics, and ads, alongside a proposal to automate short-form content production and distribution while reviewing the expected output before adopting the workflow.
- 5:06:26
Is a MacBook Air Enough for Studying AI?
A brief answer says the MacBook Air configuration mentioned in the question should be more than adequate, followed by a note that the meaning and curriculum of an “AI major” need clearer definition.
- 5:09:46
When Does a Task Need an AI Agent?
The advice is to first look for an API or MCP interface for the target software, then consider computer control when no direct integration exists. In the speaker’s view, agents are most appropriate for highly repetitive work such as content-posting workflows, rather than every task indiscriminately.
- 5:11:50
An Opinion on Computer Use in Codex and Claude
The speaker expressed a personal preference for computer use in Codex and criticized the Claude experience for taking control of the mouse pointer during execution. This was an anecdotal opinion, not a documented comparative test.
- 5:12:15
Understanding a Gaming-Stream Editing Workflow
An exploratory discussion with a gaming-stream editor about the role, including finding standout moments and turning them into content, with StreamLadder mentioned as an example tool in this space. The page does not contain a complete workflow explanation.
- 5:17:07
Simplifying Media Downloads and Conversion with AI
The discussion says a tool within Claude can handle a stream link, downloading or transcription, and MKV-to-MP4 conversion, reducing the need to configure separate encoding tools. It also notes that human editing may produce cleaner results but takes substantially more time.
- 5:19:01
Training an Agent on a Creative Workflow Step by Step
The speaker suggests training an agent on creative tasks incrementally with a capable model, feeding it instructions and screenshots one step at a time. Suno Studio is used as an example of tuning a workflow toward a desired musical style, while noting that training takes time and full editing remains difficult for AI.
Got a question about something in this stream?
Ask in the group