I’ve been exploring a bunch of AI tools over the past few months and had put together some notes just to make sense of things for myself. I wasn’t planning to share them—especially not six months later—but a friend happened to ask me about them today, and I figured… why not? Maybe someone else will find them useful too. Or maybe not. Either way, here they are.
Perplexity.ai — Real-Time, Source-Cited Search
What it does:
Perplexity is a conversational search engine. Unlike traditional search engines that list links, it gives direct answers and cites the sources it pulls from.
Core strengths:
- Provides real-time information
- Transparent with linked citations
- Handles multi-step questions in conversation form
Limitations:
- May lack depth for complex or academic topics
- Dependent on web data quality
- Doesn’t replace deep literature review tools
Best used for:
Quick research, fact-checking, comparing options, staying current with news or trends.
Claude.ai — Safe, Contextual, Conversational AI
What it does:
Claude is a large language model developed by Anthropic with a strong focus on ethics, empathy, and conversation flow. It handles long-form dialogue while minimizing harmful or biased outputs.
Core strengths:
- Maintains context in long conversations
- Prioritizes ethical, safe, and respectful interaction
- Good for writing, explanation, and creative assistance
Limitations:
- Can be overly cautious in creative tasks
Best used for:
Writing assistance, thoughtful conversations, educational support, brainstorming, and sensitive contexts.
Explainpaper — Decoding Complex Research Papers
What it does:
Upload a PDF of a research paper, highlight the hard parts, and Explainpaper simplifies the language for you. It’s designed to help you understand technical and academic content without needing deep subject knowledge.
Core strengths:
- Breaks down jargon in context
- Interactive explanations tailored to what you highlight
- Great for interdisciplinary understanding
Limitations:
- Struggles with unclear or poorly formatted PDFs
- Not ideal for humanities or abstract subjects
- Doesn’t provide citations from external sources
Best used for:
Reading and understanding academic papers in science, engineering, AI, and medical fields.
Elicit — Literature Review and Evidence Synthesis
What it does:
Elicit helps researchers automate parts of the literature review process. Enter a research question, and it finds, extracts, and organizes information from academic papers.
Core strengths:
- Semantic search (understands meaning, not just keywords)
- Extracts relevant data into structured tables
- Supports workflows for systematic reviews and evidence analysis
Limitations:
- Limited to available academic databases
- May require refinement for complex research queries
- Interface may feel unfamiliar to non-researchers
Best used for:
Systematic reviews, academic research, policy analysis, and data-driven decision-making.
NotebookLM — Personal AI Trained on Your Notes
What it does:
NotebookLM is an AI assistant that works only with your uploaded documents. It summarizes, answers questions, and helps you draft ideas based on your own content.
Core strengths:
- Tailored responses based only on your materials
- Cross-references multiple documents
- Provides citations from your uploaded content
Limitations:
- Doesn’t search or retrieve new data from the internet
- Limited by the quality and quantity of uploaded documents
- Some advanced features still under development
Best used for:
Summarizing and organizing your research, thesis writing, content creation, and collaborative document analysis.
Choosing the Right Tool
Each of these tools excels in a specific niche. Here’s a simple way to decide what to use when:
| Task | Recommended Tool |
|---|---|
| Quick, cited answers from the web | Perplexity.ai |
| Conversational help or writing feedback | Claude.ai |
| Understanding dense academic papers | Explainpaper |
| Synthesizing findings from research | Elicit |
| Summarizing and working from your own documents | NotebookLM |
I’ve been playing around with NotebookLM quite a bit lately—and I have to say, it’s helped me zip through subjects in just a couple of hours, things that would usually take half a day (and a lot more caffeine). If you’re looking to get a quick grip on a new topic without drowning in information, it’s definitely worth exploring.
Currently, I’m experimenting with Claude—trying to finally finish The Order of Time, a book I’ve been circling back to for the last four years. I’m also using it to create study worksheets for my niece, hoping to make learning a bit more engaging for her.
More updates soon as this little AI-assisted adventure unfolds.
Urvi goes experimenting 