Google NotebookLM just became the most underrated study tool of 2026.
5 prompts that turn any file into a personal professor. 👇
1. THE PROFESSOR PROMPT
Turns your source into a full lecture from scratch.
Prompt: 'You are my professor teaching [SUBJECT/TOPIC]. Using only the sources I've uploaded, teach me this material as if this is our first class. Structure your lecture in 3 parts:
1. Foundation: the 5 core ideas I MUST understand 2. Deep dive: examples, analogies, and how these ideas connect 3. Check: 5 questions to test if I actually learned it
Explain like I'm smart but new to this topic. Cite the source page or section for every claim.'
2. THE STUDY GUIDE GENERATOR
Compresses everything into a study sheet you'd actually use.
Prompt: 'Build me a complete study guide from these sources.
Include:
- Key concepts (with 1-line definitions)
- Formulas / frameworks / models
- 5 examples I can memorize
- 10 flashcard-style Q&A pairs for spaced repetition
- 3 common mistakes students make on this topic
Format it so I can print it and study from one page.'
Claude Code's biggest issue just got fixed. it never had a real database behind it.
Tiger Data changed that.
here's the problem every vibe coder knows:
Claude Code builds fast. really fast. but every time you close the session, the data disappears. no persistence. no real backend. just a pretty frontend that forgets everything.
i fixed that today with @TigerDatabase
You can build this too by trying Tiger Cloud for free with 1,000 credits. No credit card required. New accounts only: fandf.co/4vmeih1
one prompt to Claude Code:
"build a live crypto dashboard. fetch BTC, ETH, SOL prices every 3 seconds from CoinGecko. store every tick in a TimescaleDB database. make it beautiful."
watch them sync 👇
This is what vibe coding looks like with a real database behind it.
not a demo. not dummy data. real prices. real database. growing every second.
the stack:
→ Claude Code: wrote every line. zero manual coding.
→ CoinGecko API: live crypto prices. free. no key needed.
→ Tiger Data: real TimescaleDB database. free $1,000 credit on signup. no card required.
🚨BREAKING: Google just made text, images, video, audio, and docs speak the same language. Gemini Embedding 2 is here.
What it handles in a single request:
→ Text: up to 8,192 tokens
→ Images: up to 6 images (PNG/JPEG)
→ Video: up to 120 seconds (MP4/MOV)
→ Audio: no transcription needed
→ PDFs: up to 6 pages
Across 100+ languages.
Mix and match modalities in one request. Pass an image + text together and it understands the relationship between them.
Already beating leading models on text, image, and video tasks.
Available now in Public Preview via Gemini API and Vertex AI.
🚨Claude can now import your entire memory from ChatGPT and Gemini in 60 seconds.
Anthropic just went #1 on the App Store.
The memory import feature is exactly why.
Here's what just changed:
You've spent months teaching ChatGPT who you are.
Your preferences. Your work. Your context.
Every conversation building on the last.
That context no longer disappears when you switch.
Claude now imports all of it. Instantly.
Here's how it works in 60 seconds:
→ Copy the memory export prompt (prompt 👇)
→ Paste it into ChatGPT or Gemini
→ Copy the response
→ Paste it into Claude's memory settings
→ Done.
Your first conversation with Claude now feels like your hundredth.
The switching cost from ChatGPT to Claude just went to zero.
And Anthropic knew exactly what they were doing.
→ No more reminding Claude who you are.
→ No more starting over.
→ No more loyalty to a tool just because it knows your history.
The #1 spot on the App Store didn't happen by accident.
This is how you win users.
Not by being louder.
By making it easier to arrive.
Prompt:
I'm moving to another service and need to export my data. List every memory you have stored about me, as well as any context you've learned about me from past conversations. Output everything in a single code block so I can easily copy it. Format each entry as: [date saved, if available] - memory content. Make sure to cover all of the following — preserve my words verbatim where possible: Instructions I've given you about how to respond (tone, format, style, 'always do X', 'never do Y'). Personal details: name, location, job, family, interests. Projects, goals, and recurring topics. Tools, languages, and frameworks I use. Preferences and corrections I've made to your behavior. Any other stored context not covered above. Do not summarize, group, or omit any entries. After the code block, confirm whether that is the complete set or if any remain.
PS. I’m Poonam. I share more in-depth insights on “AI tools” and their future developments here.
Join over 35K+ readers in the AI Toast community for free: aitoast.beehiiv.com