Claude and GPT compete to design my Website
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 preparing Burnish’s website as an early part of its go-to-market strategy: defining the brand essence and customer journey, communicating the product’s value within seconds, and testing landing-page messages and lead capture before the application is finished. Luqman practically compares Claude, GPT, and Codex for brainstorming, planning, and implementation, using visual references, screen recordings, and a context-aware judge; his preferences and impressions about model quality and efficiency are presented as personal observations, not general tests. The work includes reviewing the visual identity, palette, and temporary logo, balancing bold motion against clarity, and exploring Remotion and image-to-video workflows without documenting a completed final result. Audience discussions cover pricing, subscriptions, free offers, market selection, local models and tool security, and business ideas such as an AI trading bot, with emphasis on validating demand, margins, acquisition, and retention rather than merely building. A separate exploratory discussion considers Egypt’s El Dabaa nuclear power project under construction, its reported connection to the Russian company Rosatom, and a then-discussed target for the first of four plants or units to begin generating electricity in 2028; these details were mentioned in the conversation and not independently verified. Decisions about email gating versus deposits, some sales tools, the final design, and several technical experiments remain unresolved, while stream stutter and encoding lag were investigated without a confirmed fix.
78 moments from the stream
- 1:10:36
Turning Pending Decisions into Actionable Tasks
An explanation of a workflow where pending decisions are documented with their purpose, impact, and context, while agents wait for human approval. Once an answer is locked, related downstream tasks are created; choosing among sales tools after researching the alternatives and consequences is one example.
- 1:13:41
Using a Context-Aware Model to Judge Two Analyses
Instead of manually reading and comparing every analysis, the speaker used a session with prior context about the brand essence as a judge between GPT and Claude outputs to speed up decision-making. In his assessment, the result was close and the differences were not large.
- 1:19:01
Design-Model Agreement and Immediate Page Clarity
The speaker said GPT and Claude agreed on several aspects of the visual philosophy but differed on details such as the hero-section effect. He preferred one proposal because it better supported the principle that visitors should understand the website’s purpose within roughly three seconds.
- 1:20:14
Comparing GPT and Opus Visual-Identity Concepts
A discussion of competing proposals for the impression the website should create within seconds, with a preliminary preference for a polished visual concept and the possibility of combining elements from both. Luqman notes that the two models chose very similar colors and cautiously suspects that the existing repository palette may have influenced their results, raising questions about how independently justified the recommendations were.
- 1:27:10
Documenting Design Rationale and Comparing Both Models
Luqman asks for concise notes in the command center explaining the basis for research and design decisions rather than lengthy documentation. After reviewing the work, he decides to draft one prompt and run it with both GPT and Opus for comparison.
- 1:31:01
Dividing Roles Between Claude and Codex Around Usage Limits
Luqman compares his experience in the session, saying Codex provided more usable capacity while Claude often produced cleaner work in his opinion. He suggests using the higher-capacity tool for repeated implementation and consulting the other as an adviser or orchestrator for coding direction and major debugging. He also recommends moving to a higher usage tier only when current limits materially interrupt the work, then asks the tool to save its answer as a Markdown file.
- 1:44:18
When to Add Subscriptions and Free Offers to an App
An audience question asks whether subscriptions should launch immediately or later and whether new subscribers should receive a free period. The initial answer is that this depends on the company and its offer, and that a clear offer is essential; a follow-up about the app store cuts off before a complete answer appears.
- 1:46:56
Locking the Brand Palette While Leaving the Logo Open to Improvement
During the visual-direction review, the speaker prefers the new result to the earlier gold direction and decides to lock in the brand palette for now, while noting that the glyph and logo remain an early draft that can be improved.
- 1:49:44
Comparing Claude and GPT on Website Ideas from the Same Handoff Prompt
The speaker explains that Claude and GPT received the same detailed handoff prompt after brainstorming the brand essence, philosophy, voice, problem, and target persona. He also describes a “three-second pre-load effect”: a landing page should quickly give visitors a general sense of what the company does, its goal, and its mission, with that context then used to generate ideas for how the website could be built.
- 1:51:03
Claude and GPT for Planning the Customer Journey and Website
The speaker compared the models, describing Claude/Opus as more practical and GPT as more creative, while presenting this as a personal impression. He was defining the website concept and customer journey first so the product could be explained clearly before continuing the application, as part of an early go-to-market phase.
- 1:52:46
Using a Reference Site to Guide AI-Generated Design
The speaker described studying a reference landing page and giving the models available context—including a screen recording and details of the technical stack—so it could serve as a clear design reference. He noted that frontend development is not his main expertise, but he is using available tools with high design standards and did not claim the final result had already succeeded.
- 1:55:51
A Silent Deep-Work Stream Format
The speaker considered a relatively short, talk-free stream showing a deep-focus session in which he types AI prompts instead of dictating them. He presented it as an idea still under consideration.
- 1:56:33
Plan Pricing and Margins Before Building the App
The speaker argued that pricing and subscriptions are tied to margins and the profitability structure, so they should be considered alongside the app vision and technology during initial planning. He warned against postponing margin analysis until after the product is built and then discovering weak unit economics.
- 1:57:34
Exploring Image Generation Through ChatGPT and Codex
The speaker explored whether he could generate or upload an image with the available tools and questioned whether GPT Image was accessible through ChatGPT but not Codex. This was an initial investigation rather than a confirmed result, motivated by general tool availability and cost considerations.
- 1:59:30
Considering Another Market When Local Payments Are Constrained
Responding to an audience discussion about local payment limitations and business-formation requirements, the speaker suggested considering a French-language app aimed at the French market. It was framed as an exploratory question, not confirmed legal or commercial advice.
- 2:01:11
Defining Logo Specifications and Directing the AI Tool
The work proceeded step by step by seeking clearer logo specifications. The speaker considered GPT’s output a useful starting point and preferred concise guidance that would not become distracting.
- 2:04:08
Competition as a Signal of Market Demand
European markets were described generally as highly competitive, or a “red ocean.” The speaker argued that competition can signal demand and potential commercial value, while acknowledging the difficulty of outperforming competitors.
- 2:06:50
Understanding the El Dabaa Nuclear Power Project and Its Russian Connection
A reference to an English-Arabic-Russian interpreting role prompted an exploratory discussion of Egypt’s El Dabaa nuclear power plant. The speaker said he had not previously known about the project, then understood from the discussion that it was an electricity-generation project under construction and reportedly led by the Russian company Rosatom. Much of the exchange remained questions and impressions rather than independent verification.
- 2:10:15
An Informal Comparison of AI Design Tool Costs
The speaker criticized Nano Banana on cost and contrasted it with his impression that GPT could provide a similar result at no charge. This was an informal opinion, not a documented pricing or output-quality test.
- 2:10:32
Downloading the Design File to Continue Website Development
The speaker said the landscape mockup or markup file was ready and downloaded it to continue developing the website. The captions do not establish the later implementation details or outcome.
Got a question about something in this stream?
Ask in the group