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
- 7:22:16
Lower-Cost Alternatives for Coding Agents
An audience member asked about using Codex and Claude with no available budget. The speaker suggested trying less expensive models and referring the detailed question to the community for additional guidance.
- 7:24:05
Choosing a Coding-Assistant Plan by Usage
Beginners were advised to start with a suitable plan and monitor recurring limit usage before upgrading. The speaker also expressed the view that Codex may offer better efficiency, while stressing that the choice depends on the project and workload.
- 7:26:07
Giving Clear UI and UX Improvement Directions
A practical instruction to make an interface clear, straightforward, visually appealing, and inviting, while asking the assistant to review and improve its proposal before implementation.
- 7:27:18
Website Security Checks and AI on Both Sides of Cybersecurity
Participants discussed the value of having a specialist assess website security, especially for systems with sensitive functions. The speaker clarified that he had not tested his own site’s security, then argued that AI will be used by both attackers and defenders, increasing the speed and complexity of the contest.
- 7:34:09
Comparing Codex and Claude Through Cost and Token Usage
The speaker compared the assistants using listed pricing and average output tokens per task, concluding that GPT models appeared more efficient to him than Claude. He framed this as his reading of the data rather than a definitive judgment of quality.
- 7:37:36
The Speed of AI-Assisted Cyberattacks
The group discussed how AI can accelerate attacks, data collection, and exploitation, describing the increased speed as a growing risk that requires ongoing security attention.
- 7:38:25
The Difficulty of Building an Audience on LinkedIn
The speaker considered gaining followers and building an influential presence on LinkedIn harder than on several other social platforms, presenting it as a personal impression rather than a formal measurement.
- 7:42:43
Verbose Code Comments Affect Model Evaluation
A participant observed that Claude can sometimes add so many comments that explanation outweighs the code. The speaker responded that the evaluator’s experience strongly affects judgments about a model’s output.
- 7:47:25
Estimating Success from Evidence
The assistant said the direction may be viable, but confidence should remain conditional until there is sufficient evidence from paying customers and complete, reconciled revenue and cost data; it therefore characterized the likelihood as moderate rather than high.
- 7:51:27
Aligning Pricing and Analytics with Traffic Acquisition
Luqman said the offers and feature tiers were designed for merchants of different sizes. He said he was adding an analytics section to examine traffic in detail and planned to connect it to GA4. In his view, the main remaining challenge was driving traffic to the application.
- 7:57:09
When to Use OpenCode with OpenRouter
Luqman explained that he connects OpenCode to OpenRouter when he wants to try a model available there, and also uses it for command-line work. He added that improvements to the Codex and Claude graphical interfaces reduced his reliance on CLI tools.
- 7:58:59
Risks of Unofficial Subscriptions
The discussion warned against buying accounts or subscriptions through unofficial markets, citing anecdotal experiences from the speaker and audience in which access was terminated quickly. The conclusion was that large providers can detect such accounts, so they are not a safe option.
- 7:59:32
Reducing Long-Session Strain with a Standing Desk
After an audience comment about his chair, Luqman said his chair had caused neck discomfort. Because he works for long periods, he recommended a standing desk that allows switching to standing when sitting becomes tiring, even if the chair is not ideal.
- 8:00:19
A Qualified Comparison Between Warp and an Untried Tool
Asked to compare Warp with another tool, Luqman said he had used Warp but had not tried the alternative, so he could not claim either was better. He suggested testing both directly and said he was also considering a partial return to IDEs.
- 8:04:31
Standing Desks and Walking Pads for Computer Work
The discussion considers combining a standing desk with a walking pad while working at a computer. The walking pad is presented as a healthy way to add movement and burn calories, though its suitability depends on personal goals.
- 8:06:24
Learning AI Tools Practically Without Abandoning School
In response to a student asking about investing in AI tools and building something practical, the speaker encourages hands-on experimentation while explicitly declining to advise leaving school and recommends continuing both paths.
- 8:09:08
Measuring the Direct Contribution of Acquisition Channels
The speaker asks for a simpler customer-acquisition table that shows each channel’s direct contribution to sales, traffic, or revenue and highlights activities without a clear direct contribution.
- 8:10:23
Livestreaming as an Accountability Tool
The speaker explains that the livestream schedule was not fixed and suggests notifications and advance announcements for absences. He also describes streaming as a way to create external pressure that supports accountability and work.
- 8:12:49
Classifying Growth Activities as Inbound or Outbound
The review reframes marketing roles in inbound and outbound terms, asking whether social activity draws interest inward or relies on direct outreach. The captions do not provide a complete final allocation.
- 8:19:26
An Operating Dashboard for Analytics, Goals, and Calendars
The speaker reviews an operating dashboard containing results, analytics, company goals, a calendar, and internal employee workspaces. He says the shared goal should be strong and understood across the team, may use traffic or another suitable metric rather than money alone, and can be supported by a calendar inside each employee workspace.
Got a question about something in this stream?
Ask in the group