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
- 2:11:07
A Discussed Target for El Dabaa Electricity Generation
The discussion stated that the first of four plants or units was expected to begin operating and generating electricity in 2028. This is preserved as a date mentioned in the conversation, not an independently verified official schedule.
- 2:11:57
Temporary Brand Assets and Image Generation for the Website
The current logos are treated as temporary, replaceable brand assets, with GPT Image 2 mentioned as part of the design work.
- 2:13:41
Running an Open Model Locally and Checking Tool Security
The discussion considers running an open-source model locally and offline, while the exact model and hardware requirements remain unclear. It also stresses checking whether an AI tool is safe on a primary computer or should be isolated on another device.
- 2:16:34
Evaluating an AI Trading Bot Idea
The speaker says an AI trading bot could be valuable for personal use or as a product, but the field is crowded and not an easy route to profit. He suggests defining the intended user and performing sensitivity analysis, while emphasizing that both trading and marketing require study, substantial work, and risk. If the tool genuinely proves useful, its performance record could support marketing it, but no test or success is claimed.
- 2:20:05
Preferring the GPT Result While Keeping the Logo Temporary
The comment indicates a preference for the GPT result in the current comparison, while clarifying that the logo is only a temporary choice used to keep the work moving.
- 2:21:44
Cash Flow and Margin Differences Between E-commerce and SaaS
He explains that cash flow must be assessed in the context of the business model: e-commerce typically has greater money movement and lower margins, while SaaS companies may achieve higher gross and profit margins. The figures cited are illustrative and vary by offering.
- 2:23:56
Choosing a Temporary Logo Variant for the Website
He narrows the immediate task to the aspect ratio of the website logo while postponing social-media variants. The current horizontal version is a temporary asset for starting the site, not the official logo.
- 2:26:16
Go-to-Market Strategy with Limited Capital
He discusses how available capital and burn rate affect the speed of gaining visibility and traction, stressing that product quality alone does not guarantee adoption. With a constrained budget, he suggests combining inbound and outbound sales, tightly targeted paid ads, programmatic SEO, and growth-hacking experiments whose effects accumulate over time.
- 2:31:46
Testing Software Ideas Before Building and Entering the Market
The speaker discusses challenging business ideas early through profitability and monetization analysis, arguing that software can offer stronger margins than e-commerce. But building alone—including AI-assisted “vibe coding”—is insufficient: customer acquisition, retention, adoption, and systematic go-to-market work matter, and entering the market is genuinely difficult.
- 2:33:58
Giving Codex Context to Build a Website Diagnostic Element
The speaker assigns Codex to build a diagnostic element for the website and stresses that the agent needs the visual reference and competitive context, including Athena HQ. Once the agent begins working, attention shifts in parallel to another task concerning go-to-market decisions that affect the site.
- 2:39:56
Updating a Go-to-Market Decision After the Brand Voice Changed
The speaker asks for an earlier go-to-market decision and its notes to be reviewed because newer brainstorming, brand philosophy, voice, and messaging angle have made it outdated. The proposed goal is to identify three items to test—the captions do not make their type clear—and then finalize the decision with the agent’s help.
- 2:42:20
Testing Landing-Page Messages for Market Validation
Explains testing different headlines or landing-page variants to identify the most responsive marketing angle and validate the market while the application is still being built, before increasing marketing spend.
- 2:45:57
Lead Capture or an Early Deposit?
Discusses collecting email and store details versus requesting a deposit or early-access reservation. A deposit can signal stronger intent and help validate demand, but it creates a greater delivery commitment and may reduce conversion; a lead form has less friction but provides a weaker signal. It also considers requiring a business email to exclude some personal-email domains and improve lead quality, while leaving whether to adopt that restriction undecided.
- 2:50:24
Using a Target Persona to Focus the Product
Explains that a target persona is a hypothetical model rather than a real individual. It helps focus product and marketing decisions on a defined customer segment so similar real-world prospects recognize that the offer addresses their needs.
- 2:55:59
Designing the Website-Scan and Email-Capture Flow
A review of a proposed flow in which a merchant submits a store URL, receives a free scan and an initial score, then may provide an email and qualification details before the report is revealed, with an optional paid reservation afterward. The placement of the email gate and what appears before it remained under discussion.
- 2:57:49
A Practical Opinion on Claude Versus GPT
Luqman questioned the comparison result and said that, in some previous stream experiences, GPT did not complete requested work while Claude finished it quickly. This was a personal experience-based opinion, not a controlled test.
- 2:59:50
Considering a Locally Run AI Model
Responding to an audience question, Luqman discussed possibly trying a locally run AI model that could work without internet access. He noted that running a very large model was not his main goal and that finishing the website well remained the priority. The model name is unclear in the captions.
- 3:02:29
Partial-Reveal UX and Brand-Aligned Naming
Luqman criticized the experience of revealing most of a download and then requiring an email for the remainder, saying it felt unpleasant. He then asked how the naming could align with the new Burnish brand, without reaching a captured decision.
- 3:11:26
YC Selectivity and Its Value to Startups
In an audience discussion, Luqman described admission to YC as highly selective, comparing its difficulty to entering a prestigious university, and noted that it can be a strong source of investment for startups. This was his assessment, not evidence of an actual application or acceptance.
- 3:14:17
Clarifying Two Paid-Ads Phases
Luqman requested a clear explanation of how phase one and phase two differ in a paid-ad plan because he suspected they might overlap. The excerpt captures the planning question but not the answer or definitions of the phases.
Got a question about something in this stream?
Ask in the group