When I was in my early twenties, I moved to New York City after finishing my master's degree. No family nearby. No extended family. Just me, my backpack, and a deep homesickness that hit hardest at mealtimes.
For months, I searched for good Mexican food. Real Mexican food. The kind that reminds you of home. Then I found this little hole-in-the-wall in the East Village, close to where my first job was. The food was incredible.
I got to know the cook, Micaela. Every morning, I would stop in for a breakfast burrito, and over time, we started talking. She would ask about my work, and I would tell her about my day.
One morning, I reached for my wallet and realized I had left it at home. I could not pay. Micaela handed me the burrito anyway, but she had a condition: I would tutor her daughter, who was struggling in school. If I came by after work twice a week for an hour, she would give me breakfast every morning.
I did the math. My time was worth more than a breakfast burrito. But I also loved teaching, I wanted to help Micaela and her daughter, and honestly, I was young and every little bit helped. So I said yes.
Over the next several months, the little hole-in-the-wall started to get noticed. It got featured in Zagat (this was before online ratings were a thing). The restaurant was becoming more popular, and the owner wanted to expand. They started hiring new cooks, a prep team, and more staff.
And that is when Micaela started to struggle.
She had spent years being the one who prepped, cooked, plated, and cleaned. She did everything herself. "Nobody does it like I do" was her line, and she was not wrong. But now her job was not to do the cooking. It was to run the kitchen. Direct the team. Set the standard. Make sure every plate that left the window met her quality standards.
She hated it. Not because the team was bad. Because she did not know who she was if she was not the one holding the knife.
I think about Micaela every time someone tells me they are afraid that AI will make them irrelevant. Because the fear is never really about the technology. It is about identity.
This is what I call the Identity Bottleneck. It is the moment when your sense of professional worth gets tangled up with the tasks you perform, rather than the judgment, creativity, and standards you bring. When AI starts handling tasks you used to own, it can feel like a personal loss. Not because the work disappeared. But because you built your identity around doing it.
Micaela eventually became the best kitchen manager that the restaurant ever had. She set the standard every cook aspired to. She caught mistakes no one else could see. She made the whole operation better than she ever could have on her own.
She did not stop being valuable. She became valuable in a different way.
That is the shift happening right now with AI. And it is happening fast.
This week, OpenAI introduced workspace agents: AI workers that can run long workflows, connect to your apps, remember what they have learned, and keep working while you step away. They also announced they are building a phone designed around AI agents that replace traditional apps entirely. Google Cloud published a report calling 2026 the year of the "agent leap." Salesforce says AI agents are no longer experimental. They are becoming core operating infrastructure.
The era of being a "user" of AI is ending. The era of being an orchestrator has begun.
My mother, as always, had words for this.
El que no cambia con el tiempo, el tiempo lo cambia a él. The one who does not change with the times, the times will change for him.
She said this whenever someone in our family resisted something new. A new neighborhood. A new school. A new way of doing things. Her point was never that change is easy. It was that the cost of standing still is always higher than the cost of adapting.
That is where we are with AI right now. The professionals who will thrive are not the ones who can do the most tasks. They are the ones who can direct AI to do the tasks while they focus on what only a human can bring: judgment, context, relationships, creativity, and standards.
At FlipWork, we call this Agentic-Human Reinvention™. It is the idea that your value is not in what you produce. It is in the intelligence, both authentic and artificial, that you bring together to produce it. The People² Advantage. Your authentic intelligence, amplified by artificial intelligence.
(Human Creativity + Contextual Wisdom) x Agentic Scale = People²
THIS WEEK'S INSIGHTS
The shift from user to orchestrator is no longer theoretical. It is happening inside the tools you are already using.
1. AI agents now work while you sleep. OpenAI's workspace agents run in the cloud, pick up requests from Slack, follow schedules, and continue multi-step workflows even when you are offline. This is not a chatbot. This is a digital team member.
2. OpenAI is building a phone where agents replace apps. Instead of opening five different apps to plan a trip, you would tell one AI agent what you need, and it would handle the rest. The app era made us users. The agent era makes us orchestrators.
3. The most in-demand skill of the next decade is not a technical skill. It is learning how to direct AI. The professionals pulling ahead are not the ones who can do the most. They are the ones who can set the standard, provide the context, and direct AI to execute at their level of quality.
TRENDS
The evidence is everywhere. AI is shifting from tool to teammate, and the professionals who recognize this early will define the next era of work.
Workspace agents are the beginning of AI as a coworker, not a chatbot. OpenAI introduced workspace agents, powered by Codex and run in the cloud. They can write code, prepare reports, respond to messages, use connected apps, and keep working when you are away. Teams can build an agent once, share it across the organization, and improve it over time. This is not "ask a question, get an answer." This is "describe the workflow, and the agent runs it."
Think of it like the difference between texting your team member a question and actually delegating a project to them. One gives you a response. The other gives you a result.
OpenAI is building an AI-first phone. Analyst Ming Chi Kuo reported that OpenAI is developing a smartphone in which AI agents replace traditional apps. Instead of switching between a maps app, a messaging app, and a browser, you would simply tell the phone what you need, and the AI handles the rest. The device would maintain continuous context through a mix of on-device and cloud models. Mass production is targeted for 2028, but the signal is clear: the future interface is not an app you open. It is an agent you direct.
Gartner predicts 40% of enterprise apps will embed AI agents by the end of 2026, up from less than 5% in 2025. That is a colossal shift. And it means the apps your team already uses (your CRM, your project management tool, your communication platform) are about to have AI agents built directly into them. The question is not whether agents will show up in your workflow. They are already arriving. The question is whether you will know how to direct them.
The core challenge has shifted. OpenAI said it plainly in their Codex documentation: "The core challenge has shifted from what agents can do to how people can direct, supervise, and collaborate with them at scale." Read that again. The bottleneck is no longer the AI's capability. The bottleneck is your ability to lead it.
3 MYTHS TO REFRAME
These are the beliefs I hear from professionals who feel the ground shifting but have not yet made the mental leap from doer to director.
Myth #1: "If AI does my work, what's left for me?"
Why we believe it: We built our careers on being the person who delivers. The one who writes the report, builds the spreadsheet, and drafts the proposal. When AI starts handling those tasks, it can feel like the floor just disappeared.
Reframe: This is the Identity Bottleneck talking. Your value was never the task. It was the judgment behind it. The reason your report was good was not that you typed it. It was because you knew what to include, what to leave out, who the audience was, and what they needed to hear. AI cannot replicate that. It can only execute on it. The professionals who break through the Identity Bottleneck are the ones who realize: my worth is not in what I produce. It is in the standard I set and the intelligence I bring.
What to do: Write down three tasks you do every week. For each one, ask: "Is my value in doing this task, or in the judgment I apply to it?" Wherever the answer is judgment, that task is ready to be directed, not done.
Myth #2: "I need to understand how AI works before I can use it."
Why we believe it: Technology has always felt like something you need to learn before you can use it. We think we need a deep understanding of the mechanics.
Reframe: You do not need to understand how a car engine works to drive to work. You need to know where you are going, which vehicle to take, and when to adjust your route. AI is the same. The orchestrator does not need to know how the model was trained. They need to know what it is good at, how to give it context, and when to step in with their own judgment. Last week, I talked about building your AI stack. This week, the message is simpler: learn to direct, not to engineer.
What to do: This week, give AI a task you normally do yourself. But instead of doing it for the AI, tell it what success looks like and let it figure out the steps. See how close it gets. Then refine from there. That is orchestration.
Myth #3: "AI agents are for tech companies, not for me."
Why we believe it: The language around agents sounds technical. Codex, orchestration, multi-step workflows. It feels like something for developers, not for business leaders, educators, or solopreneurs.
Reframe: OpenAI explicitly repositioned Codex for non-technical business users in its April 2026 update. Managers, marketers, finance analysts, and operations leads can drive it entirely through natural language prompts. No coding. No API setup. Claude Cowork already works this way: you point it at a folder, give it a task in plain language, and it reads your context before it starts. The tools are already built for you. The only thing missing is your willingness to shift from doing the work to directing the work.
What to do: Try one agentic workflow this week. In Claude, set up a folder with your context files and give it a real task with success criteria. In ChatGPT, describe a workflow you repeat weekly and ask it to build a plan to automate it. Start small. The shift is not a leap. It is a step.
TOOLS TO EXPLORE
This week, your tools for practice are designed to help you shift from user to orchestrator.
Claude Cowork as your AI team lead
Prompts to steal:
"I want to do [TASK] for [SUCCESS CRITERIA]. But first, read my folder. Then ask me questions to make sure you understand my standards before you start. After you finish, list the assumptions you made and flag anything you were uncertain about."
Power tip: The key word in that prompt is "standards." When you tell AI what good looks like before it starts, you are not using a tool. You are leading a team member. That is the orchestrator mindset.
ChatGPT Workspace Agents: workflows that run without you
Prompts to steal:
"Create an agent that [describes the workflow]. It should run every [frequency], pull data from [source], and deliver the output to [destination]. Flag anything that needs my review before finalizing."
Power tip: Workspace agents are free until May 6, 2026. If you are on a ChatGPT Business, Enterprise, or Edu plan, this is your window to experiment with zero risk. Build one agent this week. Even a simple one teaches you what orchestration feels like.
Perplexity as your fact-checking layer
Prompts to steal:
"I am preparing [deliverable] for [audience]. Here are the key claims I plan to include: [list them]. Verify each claim with the most recent source available. Flag any that are outdated, disputed, or unsupported."
Power tip: Every orchestrator needs a verification layer. AI agents can do the work, but you set the standard for accuracy. Using Perplexity as your fact check before anything ships is how you stay in the loop without doing every task yourself.
TRY IT THIS WEEK (Micro Actions)
These are designed to take you from knowing to doing. Share these with the leaders in your circle.
1. Run the Identity Bottleneck audit. Write down the five tasks that take up most of your week. For each one, ask: "Is my value in doing this, or in the judgment I bring to it?" Circle the ones where the answer is judgment. Those are your first candidates for AI orchestration.
2. Direct instead of doing. Pick one task from your audit. Instead of doing it yourself, give it to AI with clear success criteria and your context. Let the AI execute. Then review the output and refine. Notice how different it feels to direct versus do.
3. Try a workspace agent. If you are on a ChatGPT Business or Enterprise plan, workspace agents are free right now. Build one simple agent for a workflow you repeat every week. Even if it is not perfect, the experience of building it will change how you think about your role.
POWER TIP
The most important sentence from OpenAI this week was not about GPT 5.5 or image generation. It was this:
"The core challenge has shifted from what agents can do to how people can direct, supervise, and collaborate with them at scale."
Read that again.
The AI is no longer the bottleneck. You are. Not because you are doing something wrong. But because the game has changed. Your job is no longer to be the best at every task. Your job is to be the best at directing the work, setting the standard, and knowing when your human judgment matters most.
That is what it means to be an Agentic-Human. That is becoming People². Your authentic intelligence (your creativity and contextual wisdom), amplified by artificial intelligence.
👉 Where is your Identity Bottleneck? What task are you holding onto because it feels like "you," even though AI could handle it? Reply and tell me.
Closing Thought
Micaela spent years being the best cook in that kitchen. She thought her hands were her value. It turned out her standards were.
The same is true for you.
AI agents are not here to replace you. They are here to carry the work while you carry the vision. But that only works if you are willing to let go of the knife and pick up the standard.
Two weeks ago, I told you to keep your skills sharp. Last week, I told you to build your stack. This week, I am telling you the hardest part: it is time to change how you see your role.
You are not a user anymore. You are an orchestrator. And the sooner you make that shift, the further ahead you will be.
♻️ Share this with a leader who is still holding onto every task because it feels like "their job."
¡Hasta la próxima, un abrazo fuerte! (Until next week, a big hug!)
Monica Marquez
🔔 Follow me for more on LinkedIn
How did you like today's newsletter?
Your opinion matters!
🔥 Loved it! It was insightful.
Responses