In Part 1, we used ChatGPT Work to review job-search information and organize possible opportunities. Now let’s make it more useful. Instead of running that review one time, let’s ask ChatGPT Work to run it every weekday morning. That is the next shift. You are not just asking: "What did you find?" You are asking: "Can you check this for me every morning and tell me what needs my attention?" 🔁 Practice This Schedule the Job Search Review: 📌 Stay in the same Work thread from Part 1. That matters because ChatGPT already has the context from your first job search review. It already knows the kind of roles you are looking for, the criteria you provided, and what the first review produced. Now give it a simple recurring assignment. 💬 Sample Prompt Use this as a starting point: Schedule this job search review to run every Monday through Friday at 7:00 AM. Each morning, review new job-search related emails, recruiter messages, saved job postings, and relevant connected files you are allowed to access. Use the resume, career summary, or criteria I already provided as the comparison point. Do not apply, submit, message anyone, upload documents, send anything, delete anything, move anything, update anything, schedule anything else, or modify any file without my explicit approval. Each morning, give me a short job search review that includes: - Keep the review concise and practical. - New opportunities found - Roles worth reviewing first - Strong matches - Possible concerns or gaps - Repeated keywords or skills - Recommended next steps - A “Needs Human Review” section for anything uncertain Keep the review concise and practical. That is it. Do not overcomplicate this. A good recurring review should: - Find new job-search information - Separate stronger opportunities from weaker ones - Compare roles against your criteria - Flag uncertainty - Recommend next steps - Avoid applying, sending, uploading, or changing anything - Keep the summary easy to scan

Most of us first learned ChatGPT as a conversation. Chat feels like it is asking: What can I answer for you today? That is useful. ChatGPT Work feels different. Work feels like it is asking: What can I take off your plate? You are not just asking for an answer. You are giving ChatGPT an assignment. You define the outcome. You give it context. You let it use connected apps where available. Then you review the result before anything happens. That last part matters. OpenAI currently describes Work as the place for longer, multi-step tasks and finished deliverables, and apps can connect ChatGPT to external tools, information, and actions depending on your plan, workspace, permissions, and settings. (OpenAI Help Center⁠) Start With a Safety Boundary When you are first learning ChatGPT Work, I recommend adding this line to your prompt: Review and prepare the work, but do not apply, submit, send, message, delete, archive, move, update, schedule, cancel, publish, or modify anything without my explicit approval. That may feel like overkill. It is not. Connected apps make ChatGPT Work more useful because the information may already live in your email, calendar, files, or other tools. But useful does not mean uncontrolled. My simple rule is: Analyze first. Recommend next. Act only after approval. Practice This: Job Search Review For the first practice, let’s use a job search review. This is a good starting point because it is useful, but still easy to inspect. We are not asking ChatGPT to apply for jobs. We are not asking it to message anyone. We are not asking it to upload anything. We are asking it to review what it can access, organize the opportunities, and prepare a recommendation for us. Depending on what you have connected, ChatGPT Work may be able to review things like: - Job alert emails - Recruiter messages - Saved job descriptions - Career notes - A resume or professional summary - Relevant files in connected storage

In Part 4, we talked about dialing model capability up or down. 📌 Now let’s practice it. My shorthand is: Find the floor, then move back up one level. That means you reduce the reasoning or effort level until you can see the model starting to miss the mark. Then you move back up to the lowest setting that still produces the quality you need. That gives you a practical balance between: - Output quality - Speed - Cost - Available usage or capacity - The amount of correction you have to do yourself Practice This: - Choose an assignment you already understand well. - Do not use something completely unfamiliar. - You need to be able to recognize when the output gets weaker. Good practice assignments include: - A business proposal - A project plan - An app framework - A book chapter outline - A training lesson - A process improvement plan ✅Step 1: - Use the model map from Part 3 to choose your product and model. - Start with the model you believe is appropriate for the work. ✅Step 2: - Run the assignment using a relatively high reasoning or effort setting. - Your goal is to create the first strong version of the plan, framework, proposal, or structure. - Save the result. This becomes your quality baseline. ✅Step 3: Once the main thinking is complete, reduce the reasoning or effort level one step. For example: High reasoning → Medium reasoning Or: High effort → Medium effort If the work is now very clear, and the product allows it, you can test reducing it two levels. ✅Step 4: Ask the model to continue the work. For example: - Expand one section - Rewrite part of the proposal - Create a summary - Draft an email from the plan - Turn the framework into a checklist - Create social posts from the article - Refine the tone - Format the output for easier reading This is where the test begins. You are no longer asking the model to do the hardest thinking from scratch. You are asking it to continue from an established structure.

In Part 2, we looked at the range of model capability. Now let’s make it practical. After more than 100 hours of testing, I have developed a fairly simple approach. I choose the product, model, and reasoning setting based on the type of work I am doing (not cost - yet). 🧠 Heavy Logic, Planning and Structure, for work such as: - Business proposals - Strategic frameworks - Complex planning - Coding - Difficult analysis - Multi-step problem solving - Assignments where maintaining context is critical - Consistency of theme or storyline For these applications, I want stronger reasoning, even if it takes longer. Product: ChatGPT Model: GPT-5.6 Sol Setting: High reasoning and for my book editor skill, Pro. Product: Claude Model: Claude Opus 5 Setting: High effort, with extended thinking when needed For these assignments, creativity in the delivery is secondary. My priorities are logic, structure, context integrity, planning, and getting the framework right. I rarely go higher, but keep in mind, most of what I do is front office business execution: strategy, project management, process refinement, education and training, and Cowork automation. I use Fable 5 rarely. I am not primarily using Claude to build sophisticated production applications. ✍️ Writing and Everyday Communication, for: - Blog posts - Creative writing - Emails - Business communication - Social posts - Editing and rewriting I usually want a capable model that leaves a little more room for creativity and variation. Product: Claude Model: Claude Sonnet 5 Setting: Standard/default effort Product: ChatGPT Model: GPT-5.5 Instant Setting: Instant I still use GPT-5.5 for this type of work while it remains available to me. The Interesting Exception: Image Creation Image creation breaks my normal rule. Why? I need the LLM to understand: - What I am trying to communicate - The audience - Composition and hierarchy - Style and mood - The creative objective So I need reasoning. But I also want the system to have enough freedom to interpret the brief creatively.

In Part 1, we established that the model behind the work can affect the quality, speed, structure, creativity, and cost of the finished result. Now we need to understand that AI models do not all sit at the same level or serve the same purpose. It is tempting to think of every new model as simply being more powerful than the one before it. The reality is more useful than that. Models are increasingly designed around different combinations of: - Speed - Reasoning depth - Task complexity - Token usage - Cost - Tool use The ability to remain effective across a longer assignment 🔑 The model with the greatest overall capability is not automatically the best model for every task. Sometimes speed is key. Sometimes reasoning is key. Sometimes controlling cost is key. Sometimes predictable, repeatable performance is most important. And sometimes you need a combination of two or more. 📊 MODELS EXIST ACROSS A RANGE A fast and economical model may be a good choice for: - Extracting information - Reformatting content - Creating a simple summary - Renaming or organizing files - Completing repetitive work - Handling a high volume of similar tasks A balanced professional model may be better for: - Business writing - Research and synthesis - Presentations - Data analysis - Document creation - Most everyday Work and Cowork assignments A higher-capability model may be worth the additional time and usage for: - Difficult analysis - Ambiguous instructions - Complex planning - Conflicting information - Long-running assignments - High-value recommendations - Work where errors are expensive This gives us a basic range: ⚡ Fast and economical ⚖️ Balanced for professional work 🧠 Advanced reasoning for difficult assignments 🏆 Maximum capability for the hardest work OpenAI describes its model families as spanning the cost-intelligence curve, while Anthropic allows users to choose the Claude model, adjust the effort level, and turn extended thinking on or off.