Sitemap

Let your brain lead (Brain-to-LLM)

3 min readAug 19, 2025
Press enter or click to view image in full size
Three brains side by side with increasing neon blue activity.
Three brains in a row, each with more neon blue activity lines than the last. — Invent

Introduction

Large Language Models (LLMs) are increasingly impacting our daily lives, transforming how we work, learn, and interact with technology.

Adoption is growing rapidly, with LLMs becoming integral to personal knowledge management, productivity, and even entertainment. This widespread adoption is reshaping our lifestyles, from how we access information and manage tasks to how we learn and communicate.

The study

MIT’s 2025 study “Your brain on ChatGPT” used neuroimaging and behavioral data to analyze how people’s brains react to using AI for essay writing.

Press enter or click to view image in full size
A scientific figure showing multiple brain scans with colored neural connection maps, EEG alpha band analysis, and significance scales. Includes side profiles and labels for LLM (red), Search (green), and Brain (blue), representing different brain activity patterns during technology use.
The dynamic Direct Transfer Function (dDTF) EEG analysis of Alpha Band for groups: LLM, Search Engine, Brain-only, including p-values to show significance from moderately significant (*) to highly significant (***).

When participants started writing with the help of ChatGPT, their brains showed lower neural activity, particularly in regions associated with deep thought, memory formation, and critical evaluation.

These participants also remembered less about what they wrote or read compared to those who worked unaided or used AI only later in the process.

Students or writers who began their projects with their own brainstorming, outlining, or drafting (using “brain-only” strategies) and then later brought in AI for refinement or contrasting ideas showed:

  • Higher creativity and memory retention
  • More complex neural connectivity
  • Better long-term understanding of the content

Dive more into the study here.

Key takeaways & how to apply them

After talking with some friends, we agreed that people are growing comfortable, but in a negative sense, they’re becoming increasingly passive instead of actively leading the way.

1. Shift your mindset

Technology isn’t the problem, it’s how you use it. Don’t let LLMs lead your thinking. Do everything with intention and awareness; use AI to review, challenge, and expand your ideas, not to generate everything for you.

2. Sketch first, digitize later

Start every project by mapping your ideas with pen and paper, a notebook, or even a google docs, anything that’s just your thoughts. Give your brain the workout before inviting AI into the process.

3. Always contrast & validate

LLMs can seem magical for their speed, but mistakes and fabrications happen. Do your own research, verify every claim, and don’t assume the AI is infallible. Remember: LLMs can “hallucinate”, creating plausible but false information.

4. Make time for independent thinking

Schedule moments for solo brainstorming and reflection before you bring your ideas to AI. Keep a side notebook (digital or paper) nearby so it’s easy to jot, plan, and reflect away from the chatbot.

5. Be the Leader, not the passenger

You are the author; treat AI as your smart co-editor or co-pilot, not the main driver (It’s not a Tesla). Technology can assist and accelerate, but your judgment and creativity should always set the course.

Conclusion

Keep looking for new studies, learn from their findings, and incorporate those insights into your own process to perform better.

Technology will always be at the forefront.

As humans, we should make sure our brains are at the forefront alongside it.

--

--

Alix Gallardo
Alix Gallardo

Written by Alix Gallardo

Building AI products that work for everyone 🪄 | UX/AI at the intersection of accessibility | @zydeer co-founder, @useinvent co-founder

No responses yet