Image by editor
# introduction
Large-scale language models (LLMs) are typically used for boxed, typical roles like “composing email messages” or “acting as an advanced search engine,” but they have a lot of potential. It only reveals their hidden potential for creative problem solving and extends it to less explored areas.
If you want to see new examples of such unconventional things about LLM, this article lists and exemplifies 7 examples that go far beyond normal chat interfaces and conversations.
# 1. Playing personal devil’s advocate for decision making
Conversational AI systems are meticulously trained to obtain end-user consent at all costs, unless instructed otherwise. The next time you need honest guidance to make a decision, instead of asking for validation, let the AI systematically challenge you to dismantle your ideas and test your logic, if necessary. For example, see the following example prompt.
“Act as a ruthless but logical critic. Please review this project proposal and identify the top three hidden risks or logical fallacies that I overlooked.”
# 2. Decoding difficult technical errors
This use case consists of providing LLM with something like a cryptic log file or a messy raw stack trace, and asking it to convert this “machine-generated mass of complaints” into a natural language step-by-step manual to fix the problem. A prompt template like the following (you can replace the part in square brackets and paste your actual error log) should work well.
“You’re getting an obscure system error similar to the following:
[paste error]Please explain in plain English exactly which line is failing and provide the command to fix it. ”
# 3. Navigating private contracts and legal language
Not sure what you’re signing into a rental agreement and don’t want to waste your energy reading through endless confusing pages full of clauses? Why not run it through an LLM (ideally self-hosted for privacy reasons) and ask them to pick up the red flags?
“Analyze this rental agreement. Highlight any unusual termination clauses, hidden fees, or non-standard liability changes that a layman might overlook.”
# 4. Simulate historical or expert figures
This encourages LLMs to emulate the specific communication styles and philosophical frameworks associated with historical figures, thereby disrupting traditional corporate thinking.
“Critique my modern social media strategy as if you were a 1960s Madison Avenue advertising executive. Focus on emotional appeal and brand positioning.”
# 5. Automate “rubber ducking” of complex logic
This is extremely useful when you want LLM to detect and point out missing steps in complex workflows or complex logic puzzles. Walk your model through complex workflows and puzzles to see if your mental map properly matches reality. Take a look at the following example prompt template.
“I’m trying to build an automated workflow that is triggered based on three specific conditions:
[list conditions]Where are the logical gaps in this sequence?”
# 6. Build a hyper-personalized skills roadmap
Use this prompt to create a bespoke syllabus that omit what you already know and focus only on specific knowledge and skill gaps and niche educational objectives.
“I already understand basic Python, but want to learn data visualization. Create a free 14-day learning plan with daily exercises focused exclusively on Matplotlib.”
# 7. Bridging real-time cultural contexts
This is very useful in the field of international relations to decipher tone, formality, and cultural etiquette in communicating with foreign countries.
“Please translate this email from a new international client, but also explain its content, the level of formatting used, and how you should respectfully format your response to fit their cultural business standards.”
# summary
These seven use cases just scratch the surface of what’s possible when you move beyond treating LLMs as simple question-answering machines.
Whether you’re stress testing your own logic, deciphering legal fine print, or bridging cultural gaps, the common thread is a deliberate prompt that gives your model a specific role, clear constraints, and concrete goals. The more carefully you structure your requests, the more it becomes clear that these tools are true cognitive partners rather than glorified search engines.
Ivan Palomares Carrascosa I am a leader, writer, speaker, and advisor in AI, machine learning, deep learning, and LLM. He trains and coaches others to leverage AI in the real world.
