# Collaborating with LLMs: Level Up from Centaur to Cyborg > How centaur and cyborg ways of working can turn large language models into useful collaborators, with practical prompts for deeper thinking. - **Author:** [Mark Calleja](https://markcalleja.info/about.html) - **Published:** 22 September 2026 - **Canonical URL:** https://markcalleja.info/centaur-to-cyborg.html In today’s work environment, integrating Large Language Models (LLMs) isn’t just a passing trend; it’s essential for staying competitive. Yet, many of us still feel a bit hesitant to use LLMs in our daily tasks. Often, they’re viewed as a threat to our jobs or a shortcut that undermines our skills. But here’s the truth: when we learn how to work alongside LLMs—especially the difference between **centaur tasks** and **cyborg tasks**—we can transform these powerful tools into valuable partners in our work. This article serves as a follow-up to my previous OCEAN article, which outlines effective prompting strategies for educators. If you’re looking for more information on what ‘objectives’ and ‘context’ mean in this discussion, be sure to check that article out on [pp32-33 of Hello World #25](https://assets.ctfassets.net/oshmmv7kdjgm/7libv7GBDRNVQXHLkWuPaY/4baa0ebe7825791bb63421ae2e026902/HelloWorld25.pdf). ## Understanding Centaur vs. Cyborg Tasks So, what’s the difference between centaur and cyborg tasks when it comes to using LLMs? ### Centaur Tasks Think of **centaur tasks** as a classic collaboration. In these situations, you’re in the driver’s seat while the LLM takes care of the more repetitive or data-heavy parts of the job. Like the mythical half man, half horse - there’s a clear division between the two parts. (But still, nobody knows how they wear trousers.) ![Four-panel diagram asking how a centaur would wear trousers, with four possible arrangements.](https://markcalleja.info/centaur-trousers.jpg) For example, consider a team member at the Raspberry Pi Foundation working on a curriculum for a new developing market. They might brainstorm key learning objectives and structure while using an LLM to help generate lesson plans or suggest coding activities. The LLM serves as a helpful assistant, allowing the educator to focus on tailoring the content to their audience. Here are some example centaur personas: - **Supportive Curriculum Developer**: - *Prompt*: "Behave as a supportive curriculum developer. Help me generate lesson plans and coding challenges for a new workshop on [topic]." - **Insightful Community Outreach Coordinator**: - *Prompt*: "Behave as an insightful community outreach coordinator. Assist me in compiling feedback from workshop participants and summarising key insights." - **Engaging Workshop Facilitator**: - *Prompt*: "Behave as an engaging workshop facilitator. Generate interactive activities that can be included in our Code Clubs on [topic] to enhance student participation." - **Resourceful Technical Writer**: - *Prompt*: "Behave as a resourceful technical writer. Help me draft clear and concise documentation for this Github repo. [URL]" ### Cyborg Tasks Now, let’s talk about **cyborg tasks**. Here, the LLM isn’t just an assistant; it becomes an extension of your capabilities, actively enhancing your work process. Consider a community outreach coordinator preparing a report on the effectiveness of Raspberry Pi workshops in local schools. They could feed the LLM data from participant surveys and workshop feedback. The LLM can then analyse this information, summarising key trends and insights that the coordinator can use to inform future workshops. Taking that analysis, the coordinator writes their report and asks the LLM to check it over to make sure it is accurate and that the language is effective and appropriate, before sharing it with the team. Additionally, think about a team working on a new educational resource. They could use an LLM to brainstorm lesson ideas based on specific topics or age groups. The LLM can generate multiple options for interactive activities or projects, giving the team a diverse range of ideas to choose from. The team can discuss and iterate these options in collaboration with the LLM, using it to weigh in on ideas and their possible outcomes and implementation. This collaborative brainstorming not only saves time but also sparks creativity, making the development of educational materials more efficient and engaging. In these examples, LLMs become essential partners in both the creative and analytical processes, empowering team members to achieve more than they could alone. The LLM enhances what you can accomplish, making collaboration seamless and effective. Cyborg tasks blur the lines between human and AI work, resulting in a dynamic partnership where both contribute equally. ## Cyborg Prompts: Sparking Creativity and Innovation One of the most exciting benefits of working with LLMs is their potential to boost creativity and innovation through structured engagement. ### Critical Thinking Mode Let’s explore **Critical Thinking Mode**. In this mode, the LLM acts like a mentor, asking questions that help you pull out your best ideas. Instead of giving you answers, it prompts deeper thinking. - **Prompt**: “I want your help to create [content] for [Context]. We are going into Critical Thinking Mode now. Please ask me only questions that challenge my thinking, without giving me the answers. Your role is to gather information without providing any answers. From now on, ask me only one question at a time and wait for my response before proceeding.” ### Interview Mode Another effective approach is **Interview Mode**, where the LLM interviews you to gather insights and create content based on your responses. This personalised approach can lead to richer results. - **Prompt**: “I want your help to create [content] for [context]. We are going into interview mode now. Please interview me with relevant, insightful questions that will help me articulate my thoughts and ideas. Your role is to gather information without providing any answers. From now on, ask me only one question at a time and wait for my response before proceeding. When I am ready I will instruct you to create some content from my answers.” ## Tackling Common Misconceptions As LLMs become more widespread in the workplace, it’s crucial to address some common misconceptions. ### LLMs as Cheating One frequent concern (I have sensed, more than heard directly) is that using LLMs is like cheating. This is an unnecessarily limiting view. (If I gave you an umbrella to stop the rain but you insisted on not using it, that would be a bit mad, right?) Instead of seeing LLMs as a shortcut, let’s embrace them as tools that enhance our skills. The goal isn’t to replace us but to amplify our contributions. ### LLMs as Risky Business Another worry is that LLMs are inherently risky or dangerous. While these concerns can be valid, responsible usage revolves around principles of fairness, transparency, and privacy. When we understand how to use LLMs ethically, they transform from sources of anxiety into reliable partners in our work. By addressing these misconceptions, we can shift our mindset from fear to empowerment. ## Embrace Experimentation The path to effective LLM integration starts with a willingness to experiment. You should definitely explore both centaur and cyborg tasks in your daily routines. Start small—maybe use an LLM to help draft emails or analyse data. As you become more comfortable, you might discover innovative ways to leverage LLMs in your projects. To wrap things up, integrating LLMs into our work lives presents a fantastic opportunity for innovation and productivity. By understanding the dynamics of centaur and cyborg tasks, we can harness the power of LLMs to enhance our workflows and create a more creative work environment. Let’s embrace LLMs not as a threat but as collaborators in our pursuit of success. Together, we can redefine what’s possible in the workplace, one innovative idea at a time.