26 AI Employees → 9 Stronger Ones
Stream summary 74This stream in minutesWant 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 documents broad practical work and discussion around building, marketing, and measuring AI-assisted products. Its strongest working thread was a Shopify app: reconsidering free-plan versus trial pricing, reviewing onboarding, code, and testing, and developing GA4 analytics that connect visibility, traffic, and orders. Deployment and test outcomes and the pricing impact were not settled, while Luqman viewed acquiring enough traffic as the main remaining challenge. Growth discussions covered inbound and outbound channels, demos, partnerships, advertising, separating traffic problems from conversion problems, demand validation and distribution, and possible later expansion into more languages, platforms, and markets such as Salla. They also included general opinions on bootstrapping, investment, when to raise funding, avoiding excessive early equity dilution, the trust barrier in government sales, and market-dependent YouTube monetization. The stream also explored smaller AI-agent teams with defined roles, dependencies, and human oversight; a permission-limited central executive agent; and voice-assistant tests that exposed latency, interrupted conversations, conflicting records, and incomplete integrations. Luqman compared coding tools and models including Codex, Claude, Gemini, and Grok, arguing that token prices and published benchmarks are insufficient without task-specific efficiency measurements and consistently framing his assessments as personal observations. Codex slowdowns and context degradation remained unresolved. The attempt to locate Gemini 4 was specifically a historical snapshot dated October 1, 2026 and did not establish its availability or final identity. Further discussion covered goal engineering, data and website security and AI’s use by both attackers and defenders, hardware constraints on local models alongside the opinion that local compute may matter long term, practical AI learning while continuing computer-science study, and whether AI weakens thinking or improves efficiency. Other topics included remote work and the possible advantage of lower-cost locations; general distinctions among work or study migration, asylum, and visits; non-personal opinions about financial independence, investment funds, and dollar-cost averaging; and workspace ergonomics through standing desks or walking pads. Luqman also discussed content and livestream strategy, including consistency versus repetition, streaming for accountability, useful standalone videos, and background-music copyright risk. The music work went beyond a general focus-music experiment: it covered tempo, layering, and Arabic maqamat, plus a Suno Studio workflow in which a musical scale is created with Python, uploaded, arranged in layers, and supplemented with sounds. This work and the assistant’s visual and vocal identity remained experiments rather than demonstrated successes.
180 moments from the stream
- 2:23:22
An AI-Employee Workflow for Sales and Partnerships
The speaker explains a sales workflow that identifies agencies serving Shopify stores, gathers contact information, and sends partnership outreach so agencies can introduce the product to their clients. He also describes tracking employee activity through calendars and asks to receive copies of sent emails for review.
- 2:26:57
An Opinion on Emotional Companionship with AI
In a concerned, speculative comment, the speaker predicts that some future relationships may involve AI companionship or even attachment to robots; this is presented as personal opinion, not an established outcome.
- 2:27:29
End-to-End Smoke Testing and Error Cleanup After a Software Task
The speaker describes a general post-implementation approach: run an end-to-end smoke test, review the limited errors that surface, and polish the workflow while preparing a quick prototype. The assessment of the result remains the speaker’s initial impression.
- 2:28:35
Title Fields When Posting to YouTube, TikTok, and Instagram
Responding to a report of clips appearing without titles, the speaker says—based on his understanding—that YouTube has a separate title field, while TikTok and Instagram rely on accompanying text rather than the same kind of title field in the posting workflow being discussed. He then asks for the blank title input to be checked.
- 2:32:19
Title and Description Differences Across Publishing Platforms
A review of a cross-platform publishing interface notes that YouTube uses a title field, while TikTok and Instagram rely on descriptions or captions without an equivalent title field. The discussion points to adapting the system to each platform’s metadata structure and improving the publishing experience.
- 2:34:16
Working on a System to Support Traffic Acquisition
A brief explanation of work on an application or system intended to support traffic acquisition.
- 2:36:06
Free Plans, Trials, and Promotions as Pricing Options
An opinion on pricing strategy combines a free plan for users who want to start without paying, direct purchase for ready customers, and an introductory promotion. Free trials are discussed as a common user expectation rather than a proven solution.
- 2:37:17
Distinguishing Installs From Traffic Acquisition
The discussion generally distinguishes installs and early testing from acquiring enough traffic before paid conversion can be assessed. The project’s actual metrics are omitted.
- 2:41:24
Building an AI-Assisted App Starts with a Clear Vision
Answering how to create a similar app, the speaker said he built it himself and that AI can help create this kind of application, provided the builder has a specific vision of what they want to make.
- 2:43:01
Selling to Governments: Lower Competition, Higher Trust Barriers
The discussion suggested that government-focused projects may face less competition, but entering the market requires overcoming a trust barrier and making an appropriate investment; the speaker also cautioned that competition is not entirely absent.
- 2:47:18
Test Multiple Acquisition Channels, Then Double Down on What Works
The discussion covered prospect research and cold outreach alongside user-generated content, inbound traffic, and outbound traffic. The proposed approach was to test several channels in parallel, then intensify the one that begins producing results, while recognizing that landing the first customer may be the hardest stage and that the timeline mentioned was only a conversational estimate.
- 2:48:59
Clear AI Instructions and Estimating Break-Even
The speaker emphasized that creating something similar with AI requires explaining the desired result precisely. In a related discussion of business sustainability, he noted that time to break-even varies with factors including B2C versus B2B, profit margin, monetization strategy, and other conditions.
- 2:52:13
Scoping MVP and Later Features
A discussion of whether deferred features should move to a later release such as 1.5 or 2, or enter the initial product, with the decision left to deeper review and the context of meetings and investors.
- 2:55:19
A Rapid, Presentation-Ready Prototype
The speaker describes a substantial prototype produced in about an hour and a half and suggests it could be turned into a presentation.
- 2:58:49
Requesting a Plain, Concise Workflow Explanation
A request to explain a person's full workflow, including how prospects are received, in concise plain English for a non-specialist audience.
- 3:01:10
Reviewing the Structure of Focus-Music Tracks
A review notes that the first track received most of the effects while later tracks seemed less varied, with a suggestion to keep the voice layer muted and possibly remove it. The music was described as a side project still being adjusted.
- 3:04:35
Human Oversight in Automated Outreach
A workflow review covers researching prospects and published contact details, then applying a quality check before sending and following up. The discussion stresses explaining targeting limits and notifying a supervisor when a reply may require intervention instead of letting automation continue unchecked.
- 3:08:20
How Is Music Designed to Support Focus?
A discussion and listening experiment about background music intended for work or study. Participants argue that slower is not always better, that tempo can match the desired energy level, and that repetition may be intentional so the music fades into the background. They also offer frequency-related opinions that are not tested in this segment, recommending headphone listening and judging the tracks by their actual effect on focus.
- 3:15:50
Combining Suno Studio with Code-Based Composition
The speaker discusses arranging layers in Suno Studio, then describes creating a musical scale with Python, uploading it to Suno, arranging the result, and adding sounds to match a specific idea rather than relying on one-shot generation of an entire track. He notes that exploring this workflow deeply takes time.
- 3:29:51
When Should an AI Employee Be Reactive or Proactive?
A brief principle for AI-employee behavior: it should support both reactive responses and proactive action, while reserving unsolicited interventions for specific, important matters so they do not become disruptive. The available captions do not preserve the full context or details.
Got a question about something in this stream?
Ask in the group