Can 150K Emails + Custom Studies Drive Installs?
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 the hardest challenge after building a product: reaching the market and attracting qualified traffic with limited resources. The speaker and audience compare paid ads, inbound marketing, and cold outreach, discussing Shopify-store filtering, email validation, personalized pages, open-and-click measurement, and domain-reputation protection. Viability and ROI remain unresolved; the prevailing direction is to run small staged tests, improve the offer and product before scaling, and avoid treating AI estimates or promised outcomes as proof. Practical work also covers indexing and searching stream transcripts, improving short-form content, and keeping human review in automated publishing workflows. Audience questions range across Codex versus Claude and usage limits, hands-on AI learning, knowledge bases, geographic targeting, and community organization. On product engineering, the discussion emphasizes clear requirements, UX/UI and ERDs, phased delivery, Playwright tests before merge, post-release monitoring, and authentication concerns including login protections, MFA, and SSO. These are presented as evolving discussions, plans, and experiments—not completed implementations or guaranteed successes.
144 moments from the stream
- 4:17:11
How Did Tadween Acquire Its Users?
The speaker discusses verifying whether Tadween’s growth came from paid advertising, organic traffic, or social media, while noting that building for the founder’s own audience may aid distribution. The actual acquisition channel remains unconfirmed in the captions.
- 4:19:09
Extracting Key Points from Video Transcripts
A brief suggestion proposes analyzing a video through its transcript to extract content and the key points relevant to a specific objective.
- 4:20:26
Evaluating an Expert Before Booking a Consultation
The discussion considers seeking advice from an experienced domain expert and treating a long history of publishing specialized content as a useful signal when evaluating a consultation, without establishing its value or effectiveness.
- 4:26:01
An Onboarding Phase for Contractor Information
A proposed onboarding phase would collect the contractor’s necessary documents, providing enough information to compare against the referenced requirements and reach an assessment. The requirements acronym is unclear in the captions.
- 4:26:59
AI Skills, Contextual Triggers, and Natural Language
The speaker prefers AI to retain a natural voice, with specific skills triggered when needed instead of comprehensively rewriting its language. Simplification is useful when requested, while stronger language may sometimes help clarify particular ideas.
- 4:29:51
French in Algerian Education
The discussion says French is taught from an early age in Algeria and is widely used across education through university, offered as an explanation for some Algerians’ fluency.
- 4:32:01
Keeping the Workflow In-App with an Intuitive UI
The direction is to defer using Telegram and keep everything inside the current system, implementing an intuitive, easy, polished interface that respects the existing UI. The system’s name is unclear in the captions.
- 4:35:26
Turning a Transcript into Topics and Timestamps with Codex
A workflow is described in which a recording is transcribed and Codex identifies the main topics with timestamps, helping locate possible edit points and reducing review time.
- 4:36:49
Can Skills Reduce Token and Context Usage?
In response to an audience question, the speaker says some skills may help reduce token or context usage, but tighter limits can affect AI performance. He does not identify a specific skill and suggests asking the model itself or consulting an experienced community.
- 4:39:15
Preparing an App Before Traffic Arrives
The speaker says that if incoming traffic were expected, he would focus heavily on the application first because parts of it still need changes or improvement.
- 4:41:03
Choosing a Codex Usage Tier and Comparing Tool Consumption
The speaker addresses upgrading after weekly usage limits are reached. In his opinion, a paid tier can be a meaningful step up, while higher tiers are not currently advisable; he also finds Claude and Codex broadly similar in consumption efficiency and recommends staying with the familiar tool or trying Codex before deciding.
- 4:42:16
Does Vibe Coding Need a Learning Roadmap?
The speaker argues that vibe coding is not a formal discipline with a fixed course or roadmap, and clarifies that he uses Codex and Claude alongside each other rather than replacing one with the other.
- 4:42:44
Use AI to Build on Skills You Already Have
The speaker argues that AI is most useful when it strengthens an existing skill, such as music, video editing, or website building. He frames AI-assisted coding as a practical journey that opens new areas and teaches users along the way, rather than as a fixed standalone subject.
- 4:44:53
Start Small: Codex, Claude, and Interface Design
The speaker advises beginners to build a personal tool or very small project and learn through experimentation. In his current experience, Codex and Claude are broadly similar in capability, though their interaction style and usage limits differ; if one tool already delivers what is needed, he suggests staying with it. For UI/UX work, he stresses communicating the idea and visual direction clearly, while noting that he had not tried some of the tools mentioned. The discussion also touched on local-currency price display and differences between speed and usage tiers, with some answers explicitly qualified as uncertain.
- 4:51:45
A Workflow for Turning Video into Short Sports Content
A participant describes a program that analyzes video, generates a script and description, creates sports-commentator-style voice-over, adds crowd sound and event-dependent excitement, and produces a vertical captioned version for platform publishing. The speaker understands it as a workflow for reshaping a stream or long video into engaging short-form content intended to be shareable.
- 4:54:54
Codex vs. Claude: Quality, Efficiency, and Usage Limits
An experience-based comparison of Codex and Claude across token consumption, output quality, and usage limits. The speaker considered their practical differences small, suggested Claude may be slightly stronger for some creative work, and viewed Codex as potentially more generous with limit resets. He noted that results depend on the model and measurement method, while an audience member disputed the limits and efficiency claims, prompting a request for comparative evidence rather than treating the opinion as settled.
- 4:59:31
How to Learn to Use AI for Your Work
The speaker recommends starting with the work itself: explain what you do, ask the AI how it can help, and gradually improve the interaction through practice instead of waiting for prior theoretical mastery.
- 5:01:07
Can Fine-Tuning Improve a Local Vision Model’s Accuracy?
Asked whether moving a vision task to a fine-tuned local model would improve accuracy, the speaker doubted that it would and suggested accuracy could fall relative to frontier models. He explicitly said his experience with local models and fine-tuning was limited, so this remained a qualified opinion rather than a tested result.
- 5:02:17
Why Some Learners Need a Course Before Starting
The speaker observes that some people will not begin a new subject without a structured course involving listening, note-taking, and memorization before practice. The excerpt ends while he is developing the point, so its conclusion or proposed alternative is not available.
- 5:02:46
Practical Learning Versus Relying on Courses
The speaker, while acknowledging that his perspective may be limited, argues that starting with hands-on practice and available videos may be more useful than relying on a course, while allowing that a course can still help and withholding a firm judgment.
Got a question about something in this stream?
Ask in the group