Post

@HeyAbhishekk

More from @HeyAbhishek

Abhishek

@HeyAbhishek

May 19

this AI workflow just ended interior design 🔥

what used to cost $100,000 and months of work now costs pennies and few minutes.

here’s exactly how to do it:👇

Step 1:

go to Vidfield.ai → select Nano Banana Pro → upload your floor plan

prompt:

“convert this 2D architectural floor plan into a photorealistic modern home interior while preserving the exact dimensions, wall placements, doors, windows, and room layout from the blueprint. generate a cohesive luxury contemporary design across all spaces with realistic furniture scale, natural lighting, premium materials, soft shadows, architectural visualization quality, ultra detailed, cinematic interior photography, 4k”

it generates your first 3D render.

Step 2:

do this for every room.

living room. kitchen. bedroom.

Living room prompt:
“generate a photorealistic render of the living area from this floor plan. camera positioned near the screen porch entrance looking inward toward the sofa seating and dining area. preserve the exact room dimensions and furniture placement from the blueprint. modern warm interior design, soft natural daylight, realistic shadows, wooden textures, cozy atmosphere, architectural digest style, ultra detailed, cinematic 4k”

Kitchen prompt:
“ultra realistic modern kitchen render generated from this floor plan. camera positioned at the kitchen entrance facing the sink and countertop area. preserve exact wall positions, kitchen proportions, and appliance layout from the blueprint. warm ambient lighting, realistic materials, matte cabinets, soft reflections, architectural photography style, highly detailed, cinematic 4k”

Bedroom prompt:
“generate a luxury master bedroom render from this floor plan. camera positioned near the entrance door facing the bed and side windows. preserve the exact room dimensions, bed placement, and circulation space from the blueprint. warm cinematic lighting, premium wood and fabric materials, soft shadows, realistic interior styling, ultra detailed architectural visualization, 4k”

this will take only few minutes.

Read 5 tweets

Abhishek

@HeyAbhishek

May 7

This AI workflow is wild 🔥

I used ChatGPT Image 2.0 to create a perfect animated scene.

Then Seedance 2.0 turned it into a full cinematic animated video in minutes

Super simple workflow below 👇

Step 1:

Go to the ChatGPT app and generate your image with GPT Image 2.

You just need one strong frame.

Use this prompt:

“Create a cinematic 3D animated scene image based on this idea: [YOUR IDEA]

Style: high-end animated film still, realistic 3D animation look, cinematic lighting, polished textures, expressive characters, strong motion, dynamic composition, shallow depth of field, high production quality.
Scene: Show the main subject in the middle of an action moment, with the environment helping tell the story. Add secondary objects, motion details, and small visual chaos that make the scene feel alive and believable.
Composition: Use a wide, dramatic camera angle. Make the frame feel energetic and immersive. Capture one strong action moment instead of a static pose.
Environment: Build a detailed setting that matches the idea. Include props, background depth, texture, and atmosphere.
Character/Object Design: Keep the main subject visually appealing, expressive, and instantly readable. If there are multiple elements, make sure they feel stylistically consistent.
Effects: Add motion blur, splashes, particles, flying objects, smoke, sparks, dust, trails, or other effects only if they fit the idea.
Color: Use a strong, appealing color palette that supports the mood and helps the subject stand out.
Final Output: A production-quality animated scene image that looks like a frame from a premium 3D animated short film.”

3/ Step 2:

Go to

Then:
→ Go inside canvas
→ Upload that image
→ choose Seedance 2
→ generate the video

That’s it.

check Seedance 2.0 prompt below 👇 vidfield.ai

Read 5 tweets

Abhishek

@HeyAbhishek

May 5

Stop asking ChatGPT Image 2.0 to summarize PDFs.

Text summaries are boring and don't stick.

Use this prompt framework to create visual summaries instead:

Step 1: Upload Your Content to Flinote

Go to Flinote. AI

Upload:
• PDFs
• YouTube videos
• Articles
• Audio files

Flinote can process anything in seconds.

No copy-paste. No manual work.

Step 2: Let Flinote Generate Study Materials

Flinote automatically creates:
• Flashcards for memorization
• Quizzes to test yourself
• Mindmaps for visual structure
• Clean notes organized by topic

Everything is ready in under a minute.

No prompting required.

Read 6 tweets

Abhishek

@HeyAbhishek

Apr 27

I searched every NotebookLM prompt that went viral on Reddit and X.

Most writers still use it like a fancy summarizer.

That's insane.

15 prompts that turn messy drafts into published work:

1. The 5 Essential Questions

This went viral for a reason.
Forces structure instead of generic summaries.

Prompt:

Analyze all inputs and generate 5 essential questions that, when answered, capture the main points and core meaning of all inputs.

When formulating questions:
• Address the central theme or argument
• Identify key supporting ideas
• Highlight important facts or evidence
• Reveal the author's purpose or perspective
• Explore significant implications or conclusions

Answer all generated questions one by one in detail.

2. Find The Interesting Bits

Steven Johnson (NotebookLM director) used this on 500,000 words of NASA transcripts.

Did 10 hours of work in 20 seconds.

Prompt:

From all uploaded sources, extract the most surprising, unexpected, or especially interesting insights.

For each insight:
• Explain why it's surprising
• Include a direct quote from the source
• Focus specifically on non-obvious patterns

Avoid generic or obvious points.

Read 18 tweets

Abhishek

@HeyAbhishek

Apr 24

Stop asking ChatGPT Image 2.0 to summarize PDFs.

Text summaries are boring and don't stick.

Use this prompt framework to create visual summaries instead:

Step 1: Upload Your PDF

Open ChatGPT.

Upload your PDF directly.

Ask:
"Break down the key concepts from this PDF into simple points I can turn into a visual summary."

Copy that output.

Step 2: Generate the Visual

Click "Create Image" in ChatGPT.

Paste the summary from Step 1.

Use this exact prompt:

"Create a hand-drawn sketchnote visual summary of these concepts. Use a clean white background. Style: graphic recording with black fine-liner outlines. Add colored markers (teal, orange, muted red) for accents. Put the main title in a 3D box at the center. Add simple doodles, icons, stick figures, and arrows connecting ideas. Text should be handwritten, all-caps, clear and legible. Layout: A4."

Read 6 tweets

Abhishek

@HeyAbhishek

Apr 17

Google has world's most powerful free study tool.

But most people are still using it wrong.

I give you ten uses of NotebookLM that save you many hours.

Save it, you'll thank me.

1. Meeting Notes Organizer

You record every meeting but never review them.

Upload all your meeting transcripts.

Prompt:
"Organize all action items by person. What decisions were made? What's still pending? Create a priority list."

Your scattered notes become a clear action plan.

2. Content Repurposing Engine

You wrote one long article and need 10 social posts.

Upload your blog post or article.

Prompt:
"Extract 10 standalone insights that work as social posts. Make each one complete on its own. Include the hook and key takeaway."

One piece of content becomes a month of posts.

Read 11 tweets

Share this page!

Enter URL or ID to Unroll

Abhishek

Try unrolling a thread yourself!

More from @HeyAbhishek

Abhishek

Abhishek

Abhishek

Abhishek

Abhishek

Abhishek

Did Thread Reader help you today?

Don't want to be a Premium member but still want to support us?

Send Email!