STOP Using ChatGPT Normally. Use Projects Like This Instead
What this covers
The video walks through how ChatGPT Projects work and their practical application to business writing and marketing. It centers on a single core insight: Projects are not simply folders for organizing chats, but persistent context engines that hold files, custom instructions, and conversation history together so that each new chat inherits the same framing and style without requiring re-explanation. The speaker demonstrates this through setup guidance and two concrete use cases—building a video sales letter with matching landing page copy, and generating a five-email sequence—both showing how projects accelerate creation and improve output quality when properly configured with detailed role, tone, and format instructions.
The video covers the structural mechanics of Projects: how to set one up, where to upload files, how to write custom instructions that reference specific copywriting influences (naming figures like Alex Hormozi and Gary Halbert to shape the AI's style), and when to create separate projects rather than pile unrelated work into one. It distinguishes Projects sharply from custom GPTs—Projects are personal, multi-model workspaces for solo iteration, while custom GPTs are shareable, single-persona tools with fixed capabilities. The video also names practical constraints: Projects cannot be shared with collaborators, have no third-party integrations, and struggle to reference content across separate chats within the same project unless a chat is saved as an uploaded file. AI output is treated as a strong starting point requiring light editing rather than send-ready copy, and the context window of any model means very large file sets may not be fully recalled. The feature itself is locked behind ChatGPT's paid tiers (Plus and Pro).
Using ChatGPT Projects as mere folders wastes their value; their real power for business and marketing comes from loading persistent context (files + custom instructions) so every chat produces on-brand, high-quality output without re-explaining.
- Projects retain files, chats, and custom instructions as shared context across conversations
- Detailed role/tone/format instructions convert generic AI output into on-brand, strategic copy
- Projects differ from custom GPTs: organization/multi-model vs. shareable specialized tools
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Projects and custom GPTs differ fundamentally: Projects organize an ongoing body of work (chats, files, instructions together) with multi-model flexibility for personal solo use, while custom GPTs are shareable, special-purpose AI tools with a fixed persona, possible unique capabilities (web browsing, app connections), and currently run on the 4o model.
“Chat GBD projects are mainly about organizing your stuff for a big task... With a custom GPT, you're essentially creating a specialurpose version of chat GPT... a really big difference is that you can actually share custom GPTs with your team... And GPTs currently use the 40 model as the underlying engine.”
Custom instructions should assign the AI a role or persona, specify preferred response format, and set the desired tone to align outputs with strategic goals and brand voice for consistency and quality.
“it's a really good idea to assign it a role or a persona... You could also tell it your preferred format for the response... And it's also helpful to set the desired tone.”
Naming famous marketers (Alex Hormozi, Dan Kennedy, Gary Halbert) in the instructions gives the AI more information on the desired copywriting style.
“I also mentioned some famous marketers like Alex Horoszi, Dan Kennedy, Gary Halbert. That just gives the AI more information on what style I'm looking for”
Even within projects, AI models have a finite context window, so with very large or numerous files the model may not use every piece of information with perfect recall.
“AI model still has a context window or a limit to the amount of information that it can actively consider at one moment... for very large files or a huge number of files, it may not be able to use all the information, every single piece with perfect recall all the time. It's very good, but it's not infinite.”
Giving a project both uploaded files and detailed custom instructions adds context that takes AI outputs from generic to genuinely useful and on-brand.
“Both of these add a lot of context to the AI to give the best possible output. Now why bother with all this setup? because this is what takes your outputs from generic to genuinely useful and onbrand.”
Within a single project, new but related tasks (like an email sequence) should be done in a new chat to stay organized, and any reference materials (like a lead magnet PDF) should be uploaded to the project files to give the AI the needed context.
“to keep things organized for this new task, we're going to start a new chat within the same project... I'm going to upload a PDF copy of that lead magnet into our project files. Now, this gives AI the context that it needs for the email sequence.”
AI-generated marketing copy is a strong starting point and massive timesaver but typically requires light editing and a personal touch before it is send-ready.
“Now, is it absolutely perfect and ready to send? Well, no. It'll probably require some light editing and your personal touch, but I think it's a really strong start and a massive timesaver.”
Projects cannot be shared or collaborated on with others and lack third-party app integration (unlike custom GPTs), making them best suited for solo work.
“you can't actually share the project environment with somebody else and collaborate with them. So the projects are really good if you're doing solo work... there's no third party app integration unlike custom GPTs”
Projects have no direct cross-chat referencing: chats in the same project cannot see each other's content despite sharing files and instructions; a workaround is to save one chat's content as a file and upload it to the project.
“Chats within the same project cannot directly see or reference the content within other chats... what you can do is you can copy that content from that chat... save it as a file, and then upload that file as a source... so that other chats can access that specific information.”
ChatGPT Projects function as dedicated smart workspaces that keep related chats, uploaded files, and custom instructions organized in one place, so each new chat starts already in context instead of from zero.
“Think of projects as dedicated smart workspaces or intelligent folders within chat GPT... It keeps all your relevant chats, any files you upload... and custom instructions all neatly organized in one place.”
Because the project supplies extensive context, generating a video sales letter and landing page copy within it saves substantial time and produces significantly higher relevance and quality than writing from scratch.
“You can see how this would save you a ton of time compared to writing a video, sales letter, and landing page copy by yourself from scratch. And because we've given it so much context within project, the quality and the relevance of the output is significantly leveled up.”
You should use separate projects when work differs in style or instructions, such as separate content-creation projects for Instagram, LinkedIn, and X, because writing style differs per platform.
“You might want to have a separate file for Instagram, one for LinkedIn, and one for X because maybe the way you write on each of those platforms is a little bit different. So you'd probably want to have different project instructions for each of those.”
Using ChatGPT projects merely as basic folders leaves significant productivity on the table because their real value is in maintaining persistent context across chats.
“just using them as basic folders, and trust me, I made this mistake for a while means you're leaving massive productivity on the table”
The Projects feature is available only on the Plus and Pro paid plans, not on the free plan.
“Right now, this feature is available for those on the plus and pro plans, but it's not available on the free plan.”
Projects now allow using multiple models within the same project, whereas previously users were limited to the 4o model, giving more flexibility to select the best model per task.
“Before you were limited to the 40 model. Now you can use a bunch of different models which is awesome. So now you have way more flexibility and you can select the best model for the task that you have.”
An existing chat can be moved into a project via the three-dots menu (add to project) and later removed the same way.
“go to the left side there. Click on the three dots. Click on add to project and then click on serene spaces. And it's going to move that chat right within the same project.”