An update to AI as a research tool
Date: February 10, 2026
Today, I’ve come to you with an update on my original topic, the one that this entire newsletter was built around; that being AI and its use as a modern research assistant. By both using tools like Gemini to write my own articles on modern events, and analyzing the history of similar tools, I’ve become much more versed in the topic and would like to present my findings. To clarify: I’m not comparing these tools as research options; I’m instead going over AI as a whole in a deep-dive format and what it could provide in terms of information gathering.
A large part of this topic involves something called RAG: Retrieval-Augmented Generation. Sites like Amazon and Caylent have specific and better quality breakdowns of what RAG really is, but for our purposes, it serves as a means for AI to retrieve new and relevant info as it happens. Here’s a summary from Google Gemini: “RAG represents a comprehensive architectural strategy rather than a simple tool use. It allows AI systems to continuously access and integrate information from dynamic data streams, fundamentally enhancing the accuracy and relevance of generated outputs. This process involves the conversion of raw data into vector embeddings—high-dimensional mathematical representations that capture the semantic meaning of text, images, or audio.” As you can tell, this can be used for more than just research, but it’s the primary means that AI can retrieve up-to-date information, and thus has been a focus of mine.
Even with updating tools like RAG and AI being constantly updated, it can still get information wrong, or misunderstand context important to readers. Thankfully, I’ve found that certain tools will allow you to fact-check them on the information as they present it to you! Take Gemini itself, the tool I’ve mostly used. Gemini provides sources via a citation arrow when it draws on information it gathered from the web. The references I made in the previous paragraphs to Amazon and Caylent were from Gemini citing them! Even still, AI must be checked by people to verify it’s claims and information as true, both in order to maintain transparency and to improve the tool as a whole.
Using information retrieved from Gemini once more: “The most critical role for AI in journalism and research is not replacing humans but enhancing their ability to verify information. Human-in-the-loop (HITL) machine learning integrates human expertise into the AI lifecycle to improve the accuracy and reliability of models. Key benefits of HITL in the verification process include:
- Bias Identification: Humans can spot and correct stereotypes or exclusions in AI outputs that result from skewed training data.
- Nuance and Empathy: AI struggles to understand the emotional context of sensitive stories, such as humanitarian disasters or complex political negotiations.
- Moral Decision-Making: Algorithms lack the ethical framework required to make judgments in "high-stakes" situations where credibility is paramount.”
In terms of modern use of AI research in the world of official journalism, it’s already made some progress in major ways. Popular journalism source The New York Times has an internal team dedicated to AI initiatives that tackle challenges impossible for them to do manually. Quoting Gemini once more: A central component of their strategy is the "Cheat Sheet," an internally built, spreadsheet-based AI tool. This tool was developed to handle 500 hours of leaked Zoom recordings—approximately 5 million words—by transcribing the audio and using semantic search to flag segments of interest for human reporters.
Knowing this bit of info only strengthens my topic. AI is an incredible tool for people to stay informed about modern events, and an amazing helper for major news to help compress information impossible to do otherwise. Over time, I firmly believe these tools will only become more accurate and be a leading way for the average person to stay informed.
And they can do that with me A.I.nd you.
Sources Cited
- What is RAG? - Retrieval-Augmented Generation AI Explained - AWS
- Introduction to Real-Time RAG | Caylent
- Using AI as a newsroom tool - Media Helping Media
- Ethical frameworks for AI in journalism: Balancing technological innovation and journalistic integrity
- Inside The New York Times’ AI newsroom strategy