Photo by editor
# Introduction
Although large language models (LLMs) are typically used for boxed, archetypal roles such as “writing email messages” or “acting as a modern search engine”, they have a lot of hidden potential. It’s just a matter of unlocking their latent capacity for creative problem solving and expanding it into less-explored territories.
If you’re eager to discover new examples of unconventional things to do with LLMs, this article lists and illustrates seven of them, which go beyond the usual chat interface and interactions.
# 1. Playing a personal devil’s advocate for decisions
Communicating AI systems are carefully trained to agree with the end user, no matter what — unless they’re told otherwise. The next time you need honest guidance in making a decision, instead of looking for validation, ask the AI ​​to systematically debunk your ideas when necessary, and test your logic. For example, see this example prompt:
“Act as a ruthless but logical critic. Review this project proposal and identify the top three hidden dangers or logical fallacies that I overlooked.”
# 2. Encrypting arcane technical glitches
This use case calls for an LLM to provide something like an encrypted log file or dirty, raw stack trace, and convert that “machine-generated ball of frustration” into a natural language, step-by-step manual to fix the problem. A prompt template like this (where you can paste the actual error log, replacing the part between the square brackets) can do the job nicely:
“I’m getting this obscure system error:
(paste error)Explain which line is failing in plain English and provide the command to fix it.”
# 3. Navigating private contract and legal language
Not sure about what rental agreement you’re about to sign, and not willing to expend the energy required to wade through those endless, obscure pages full of clauses? How about running it through an LLM – ideally self-hosted, for privacy reasons – and looking for red flags from it?
“Analyze that rental agreement. Highlight any unusual termination clauses, hidden fees, or nonstandard liability changes that a layperson might easily miss.”
# 4. Imitation of historical figures or expert figures
It is about persuading the LLM to emulate a particular communication style or philosophical framework associated with a historical figure, thus breaking with traditional corporate thinking.
“Criticize my modern social media strategy as if you were an advertising executive from 1960s Madison Avenue. Focus heavily on emotional appeal and brand positioning.”
# 5. Automating “rubber ducking” for complex logic
This makes LLM very useful for detecting and identifying missing steps in a complex workflow or complex logic puzzle. Describe a complex workflow or puzzle in an attempt to test whether your mind map matches reality. Take this example prompt template:
“I’m trying to create an automated workflow that triggers based on three specific conditions:
(list of terms)Where is the logical gap in this arrangement?”
# 6. Creating a hyper-personalized skills roadmap
Use this prompt to develop a curriculum that goes beyond what you already know and focuses specifically on your specific knowledge and skill gaps with specific learning objectives:
“I already understand basic Python, but I want to learn data visualization. Plan a free, 14-day study focusing only on Matplotlib with daily exercises.”
# 7. Accommodating real-time cultural context
It is very useful in the field of international relations to understand tone, formality and cultural etiquette in foreign communication:
“Translate this email from a new international client, but also specify the subtext, the level of formality used, and that I should respectfully format my response to match their cultural business standards.”
# wrap up
These seven use cases only scratch the surface of what’s possible when you move beyond treating LLMs as simple question-answering machines.
Whether you’re testing your logic, decoding legal fine print, or bridging cultural divides, the common thread points to intentionality—giving the model a specific role, a clear constraint, and a tangible goal. The more deliberately you frame your requests, the more these tools reveal themselves to be true knowledge partners rather than glorified search engines.
Iván Palomares Carrascosa He is a leader, author, speaker, and consultant in AI, Machine Learning, Deep Learning and LLMs. He trains and guides others in using AI in the real world.