KIMI K3 + TTFD?
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 combines work on a Shopify-focused tool for measuring traffic, sales, and visibility from AI platforms—and improving product pages and FAQs—with practical development of an interactive website that explains the product before installation. Reviews covered page inventory and reusable templates, visual hierarchy, effects, footer details, and cross-device behavior, while acknowledging that some pages, simulations, and fixes remained incomplete or unverified. Luqman argues that AI web-design tools need clear references, a design system, and repeated human direction; his impressions of Kimi K3, Claude, Codex, and OpenCode varied in speed, quality, and instruction-following, but did not amount to a conclusive systematic comparison. The group also discussed the cost and privacy of discounted API providers, key and configuration management, and local model requirements. Explanations for unusually low prices, subscription-versus-API differences, and possible caching benefits remained hypotheses requiring verification and clear data-retention policies. Audience questions covered AEO versus SEO, B2B versus B2C marketing, and evaluating business ideas before building. A recurring conclusion was to prepare the conversion funnel and track metrics such as CAC, LTV, churn, and trial-to-paid conversion before scaling spend, while carefully costing incentives and free periods; unfinished plans and experiments were not presented as proven outcomes.
72 moments from the stream
- 2:52:11
A Subjective Assessment of Kimi K3’s Design Output
The speaker is dissatisfied with a displayed design result despite generally regarding Kimi K3 as strong at design. He says another tool—whose name is unclear in the captions—started the work and Kimi completed it. This is a subjective impression, not evidence of a systematic comparison or formal test.
- 2:53:51
Why a Reference Pricing-Page Recreation Failed
A reference pricing page was to be recreated with limited adjustments, but the result was judged unrelated to the reference and inconsistent with the requested pricing after a long wait. The cause remained uncertain between the handoff prompt, execution, or model. The requirement was then clarified as a side-by-side match covering wording, typography, colors, spacing, and different states, with the tool asked to restate its understanding before composing a new prompt.
- 3:07:38
Reference-Driven Web Design with AI
The speaker describes a workflow that starts with strong reference pages or sections, asks an AI tool to reproduce them, and then customizes colors, motion, layout, and added ideas. He emphasizes checking whether the tool will transfer code or merely imitate the design from scratch, while copyright objections also arise; these are the speaker’s views and workflow experience, not evidence that copying any site is legally permissible.
- 3:13:51
Capturing a Website Can Be Faster Than Editing It Directly
The speaker explains that capturing an entire website may take less time than having the tool work directly on the site itself.
- 3:15:23
Choosing an API Provider for Privacy and Data-Retention Requirements
The speaker discusses splitting API usage by task and prefers an established provider for practical work until alternatives are tested. The decision depends on a clear privacy policy and documented zero-data-retention support, requirements that may complicate approving some models for platforms such as Shopify.
- 3:19:48
Checking Site Content Structure and Missing Pages
A quick site review covers sections such as blogs, case studies, benchmarks, privacy, and the changelog, followed by questions about where the help center is and whether any pages are missing.
- 3:20:56
Trying Kimi K3 and Using a Command Center to Start a Project
The speaker says he has started using Kimi K3 and plans to keep experimenting with it. He adds that a tool he describes as a command center gave the project a strong initial push, without presenting a final evaluation.
- 3:21:57
A Debugging Workflow Mismatch and the Need for a Marketing Dashboard
After the application was completed, progress stalled because the debugging workflow differed from the build workflow, preventing the speaker from continuing. The discussion ends by identifying a marketing dashboard as a future requirement.
- 3:27:58
Competitor Comparison and Pricing Positioning
A general comparison of competing services by website strength and premium positioning, with the qualification that pricing decisions may change in response to post-marketing data.
- 3:31:16
Growth Metrics That Guide Marketing Spend
An explanation of tracking ARPU, cost per install, trial-to-paid conversion, customer acquisition cost, lifetime value, and churn. These indicators help determine when spending can scale and when it should pause so creatives or funnels can be fixed first.
- 3:36:17
A Shopify App for Measuring and Improving AI Visibility
Responding to an audience question, the speaker explains that he is building an app for Shopify merchants that measures their visibility on AI platforms and aims to improve it through a plugin and optimization of FAQs and product pages.
- 3:37:23
Tracking Sales and Traffic from AI Platforms
Explains the idea of attributing sales and traffic referred by AI platforms, describing it as analogous to search optimization but focused on the AI ecosystem.
- 3:39:35
Turning a Sitemap into Agent-Ready Design Templates
Outlines a workflow that begins with studying reference pages and documenting their patterns, then grouping sitemap pages and approving one reusable template per category. A coding agent can subsequently apply each template and connect existing content, with design decisions identified as the most time-consuming part.
- 3:47:51
Starting a Template Design Brainstorm
The group is invited to choose a starting point, propose ideas, and help lead a discussion about template designs.
- 3:51:58
Critiquing Visual Hierarchy in an AI-Generated Interface
A promising but incomplete interface frame is reviewed: the AI-engine marks should follow a deliberate structure, while the box around the shoe feels too intrusive; a simpler floating-product presentation is suggested.
- 3:54:24
GEO Tool Scope and Cross-Platform Expansion
Asked whether the GEO tool supports regular websites and WooCommerce, the speaker says it is currently Shopify-only, while later expansion to other platforms or a standalone web app is possible. The approach is presented as phased: focus on one platform before broadening the product.
- 3:58:26
How to Evaluate a Business Idea Before Building It
The speaker argues that gross margin, profitability structure, and monetization strategy should be examined before building a product. He also recommends studying demand trends and competition, seeking a middle ground between an overcrowded market and one with no proven demand rather than relying on an appealing idea alone. An emerging answer- and generative-engine optimization field is used as an example, but its expected growth is presented as the speaker’s and cited experts’ view, not a confirmed outcome.
- 4:03:25
Turning Interface Feedback into a Visual Sketch
The interface review gives concrete feedback on moving a visual element out of a frame, defining a case around visibility, content, and revenue, and requesting a visual sketch before final approval.
- 4:04:48
Reviewing a Page Built with an AI Coding Tool
The speaker reviews a page requested to match an existing reference and expresses satisfaction with the result after iterative prompting and investigation of problems. He notes that the code is now in the repository, allowing the page to be modified later.
- 4:12:33
Visually Improving a Dashboard Frame
A practical interface review: frame 03 had improved, while frame 02 needed a lighter background and a more convincing dashboard with rounded corners and its own space away from the page grid lines.
Got a question about something in this stream?
Ask in the group