Output of the same prompt in two different LLMs. This post is a personal reference to check the differences between LLMs. For this post I used Gemini and Mistral, below you get the outcome:
Original Prompt:
corrige este post para linkedin y X que acabo de hacer:
The more I'm walking this agentic rabbit hole, the more I feel the depth of the AI revolution.
TLDR: AI is forcing us to think through and avoid instant gratification, that will transpire in all corners of your life.
What I enjoy the most is throwing bold "dumb" questions ( as it is part of my growth strategy, for most of those queries I already know the answer others I just made a link between 2 unrelated topics or I simply don't know - I just like to do it because is my way to test and discover new learning vectors) without having to read the peers body language with the judgement of .. you should know that already. The AI is already rewiring our brains for the better, I have no other proof that my own brain, and so far the agentic flow is already having an impact on how I approach work, an example of it.
Before 2026 - get a problem description, identify/isolate/plan then google it, get the hit, wrap up the solution and deliver it ... hit of dopamine and enjoy the high.
After 2026 - sit down, get your first prompt in your mother tonge (write from your heart) then change to your work/business language and reevaluate if what you asked is what you wanted (write from your head), get a small evaluation on which llm will be the best fit for the task (or be lazy and dry your tokens). Then execute on your favorite agent and environment (I love CLIs), see the agent going through the thinking process, evaluate if that was the path you wanted it to go or not, wait for the plan. Evaluate the plan, even challenge some parts of the plan, then accept it or ask for changes. Wait for it ... still wait for it ... now check the artifact created, evaluate the artifact and propose changes, if that works for you then you move to the next stage. Iterate as many times as you need up until you get something refine. Then publish for the world to see ... Then you get the dopamine and enjoy the high.
PD: If you read until here, thank you .. and congrats for rewiring your brain.
Gemini response:
LinkedIn Post
The deeper I go into the agentic AI rabbit hole, the more profound the AI revolution feels.
TL;DR: AI is forcing us to delay instant gratification, a shift that will rewrite how we approach problem-solving in every area of life.
One of the greatest benefits of interacting with AI is the freedom to ask bold, seemingly "dumb" questions. I use this as a growth strategy: some questions are to test boundaries, others connect unrelated concepts, and some stem from genuine curiosity. The absence of human judgment or body language creates a safe space to discover new learning vectors.
AI is rewiring our cognitive frameworks. My own workflow has completely transformed over the last year:
- Before 2026 (The Linear Era): Receive problem $\rightarrow$ Isolate/Plan $\rightarrow$ Search/Google $\rightarrow$ Implement $\rightarrow$ Instant dopamine hit.
- After 2026 (The Agentic Era):
- Drafting: Prompt in native language to capture the core intent (from the heart).
- Refining: Translate to professional language to audit logic (from the head).
- Routing: Select the optimal LLM infrastructure or agent framework.
- Execution: Run inside the preferred environment (CLI).
- Auditing: Observe the agent's thought process, challenge the plan, and iterate on the generated artifacts. The dopamine loop has shifted from instant search-and-find to deep architectural curation. If you made it this far, thank you—and congratulations on rewiring your brain for the agentic shift. --- ## X (Twitter) Post The agentic AI rabbit hole is shifting our cognitive frameworks from instant gratification to deep curation. A comparison of how workflows have changed: Before 2026: Problem $\rightarrow$ Google $\rightarrow$ Quick fix $\rightarrow$ Dopamine hit. After 2026:
- Prompt in native language (heart)
- Shift to business language (head)
- Select optimal LLM/Agent framework
- Execute via CLI
- Audit execution thoughts & modify artifacts AI environments create a judgment-free zone to ask bold questions, test boundaries, and discover new learning vectors without human bias. The reward is no longer the speed of the answer, but the depth of the iteration. Have you rewired your workflow yet?
Mistral response:
Versión LinkedIn:
The deeper I dive into the agentic AI rabbit hole, the more I grasp the depth of this revolution.
TL;DR: AI is pushing us beyond instant gratification, and this mindset shift will reshape every corner of your life.
What I enjoy most is asking bold, seemingly naive questions. It is part of my growth strategy—sometimes I already know the answer, other times I am connecting unrelated concepts, and occasionally I simply do not know. I do it to test boundaries and discover new learning vectors. The best part: no judgment, no body language to decipher, no "you should already know this."
AI is already rewiring our brains for the better. My only proof is my own experience, and the agentic workflow has changed how I approach work.
Before 2026: Receive a problem, identify, isolate, and plan, Google it, find the solution, wrap it up, deliver. Dopamine hit. Enjoy the high.
After 2026: Sit down. Draft your first prompt in your native language (write from the heart). Then translate it to your work language and verify if it captures what you truly want (write from the head). Assess which LLM fits best—or just burn tokens. Execute in your preferred agent and environment (I love CLIs). Watch it work through the thinking process. Evaluate whether it is on the right path. Wait for the plan. Review it. Challenge parts of it. Accept or request changes. Wait... still waiting. Now examine the artifact. Evaluate it. Propose refinements. Iterate as needed until it is polished. Publish for the world to see. Then, finally, the dopamine hit.
P.S. If you read this far, thank you. And congratulations—you are already rewiring your brain.
---
Versión X:
The deeper I go into agentic AI, the clearer I see this revolution’s depth.
AI forces us beyond instant gratification—this mindset will transform everything.
I love asking bold "dumb" questions. Sometimes I know the answer. Other times I connect unrelated dots. Often I just do not know. No judgment, no body language to read. Just discovery.
Before 2026: Problem → Google → Solution → Dopamine.
After 2026: Native prompt → Work language → LLM → Agent → Think → Evaluate → Challenge → Iterate → Refine → Publish → Dopamine.
Read this far? Congrats—you are rewiring your brain.
Top comments (0)