Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Guri Singh Profile picture
@heygurisingh
Jul 23, 2025 • 20 tweets • 11 min read • Read on X
Scrolly
🚨 BREAKING: Microsoft just dropped an 18-episode series called "Generative AI for Beginners".

Ideal for beginners, developers, and AI enthusiasts looking to build a solid foundation.

Here’s a breakdown (Save this👇):🧵 Image
Episode 1: Introduction to Generative AI and LLMs

Carlotta Castelluccio discusses Generative AI and large language models, explaining their mechanisms and their transformative impact on various industries, particularly education.

learn.microsoft.com/en-us/shows/ge…Image
Episode 2: Exploring and Comparing Different LLMs

Carlotta Castelluccio and Pablo Lopes talk about different big language models, how they are used in different industries, and how businesses can take advantage of them.

learn.microsoft.com/en-us/shows/ge…Image
Episode 3: Using Generative AI Responsibly

Talks about why it's important to use AI responsibly in apps that create content.

It explains how to make sure AI outputs are fair and don't cause harm and gives steps to use AI responsibly.

learn.microsoft.com/en-us/shows/ge…Image
Episode 4: Understanding Prompt Engineering Fundamentals

Nitya Narasimhan explains the basics of prompt engineering, showing how to create good prompts and improve them to get better results from generative AI.

learn.microsoft.com/en-us/shows/ge…Image
Episode 5: Creating Advanced Prompts

Chris Noring explores advanced techniques for creating prompts to make AI responses better. He shows how these methods work with real-life examples.

learn.microsoft.com/en-us/shows/ge…Image
Episode 6: Building Text Generation Applications

–Explains the main ideas of creating text with OpenAI and teaches you how to make a text-generation app step by step, changing settings like prompt, temperature, and tokens to get the result you want.

learn.microsoft.com/en-us/shows/ge…Image
Episode 7: Building Chat Applications

– Jasmine Greenaway shows how to build and add smart chat apps into systems we already use. She covers how to customize them, make them work better, and keep an eye on how they perform.

learn.microsoft.com/en-us/shows/ge…Image
Episode 8: Building Search Apps (Vector Databases)

– Dave Glover talks about making smart search apps using vector embeddings. In this episode, he shows how to build a search app for an education startup's video library. The app helps find the right video parts by using and searching through text embeddings.

learn.microsoft.com/en-us/shows/ge…Image
Episode 9: Building Image Generation Applications

–Chris Noring and Pablo Lopes show how to create pictures from text descriptions using models like DALL-E and MidJourney, and guide you through making an image-generation app step by step.

learn.microsoft.com/en-us/shows/ge…Image
Episode 10: Building Low-Code AI Applications

– Someleze Diko explains how to use low-code tools like Microsoft Power Platform with AI. This episode talks about using AI features such as Copilot and AI Builder to make apps and workflows with AI, without needing a lot of coding skills.

learn.microsoft.com/en-us/shows/ge…Image
Episode 11: Integrating External Applications with Function Calling

– Korey Stegared-Pace talks about using function calling with large language models (LLMs). He explains what function calling is, shows how to set up an OpenAI function call, and how to use it in an app to make AI do more things.

learn.microsoft.com/en-us/shows/ge…Image
Episode 12: Designing UX for AI Applications

– Bethany Jepchumba talks about how to create user-friendly designs for apps that use AI. She highlights the need to build trust and be clear with users to make sure AI is used responsibly. The focus is on knowing what users need and designing AI tools that allow for teamwork and feedback.

learn.microsoft.com/en-us/shows/ge…Image
Episode 13: Securing Your Generative AI Applications

– Explains the security issues for AI systems. Korey Stegared-Pace talks about common risks and dangers to AI programs and shares ways and tips to keep AI systems safe from these dangers.

learn.microsoft.com/en-us/shows/ge…Image
Episode 14: The Generative AI Application Lifecycle

– Pablo Lopes dives into the dynamic world of the generative AI lifecycle—where innovation never rests. Discover the shift from traditional machine learning to cutting-edge large language model operations. This session unveils the essential tools, metrics, and strategies to elevate and sustain AI applications.

learn.microsoft.com/en-us/shows/ge…Image
Episode 15: Retrieval Augmented Generation (RAG) and Vector Databases

– Bethany Jepchumba explores how to connect large language models (LLMs) with your own data using Retrieval Augmented Generation (RAG). The lesson covers what RAG is, why it's helpful, how to make and use a vector database for storing information, and how to add RAG to an app.

learn.microsoft.com/en-us/shows/ge…Image
Episode 16: Open Source Models and Hugging Face

– Explores the world of open-source large language models. Korey Stegared-Pace talks about why open-source LLMs can be better than private ones and shows how to find and use open models, like on Hugging Face or Azure AI Studio. Also includes tips on how to fine-tune these open models.

learn.microsoft.com/en-us/shows/ge…Image
Episode 17: AI Agents – This section introduces AI Agents, systems where large language models (LLMs) perform tasks using tools or frameworks.

Korey Stegared-Pace explains what AI agents are, examines four distinct agent frameworks, highlights what makes each one unique, and discusses the optimal scenarios for using these agent-based methods in developing AI applications.

learn.microsoft.com/en-us/shows/ge…Image
Episode 18: Fine-Tuning LLMs

– Nitya Narasimhan talks about how to make pre-trained language models work better for certain tasks. In this last part, she explains what fine-tuning is, when it's helpful, and how to fine-tune a large language model, while also mentioning the limits of fine-tuning.

learn.microsoft.com/en-us/shows/ge…Image
I hope you've found this thread helpful.

Follow me @heygurisingh for more.

Like/Repost the quote below if you can:

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Guri Singh

Guri Singh Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @heygurisingh

Guri Singh Profile picture
Feb 10
BREAKING: AI can now create and sell digital products while you sleep.

Here are 18 insane Grok 4 prompts to build passive income streams in 2026: (Save for later): Image
Prompt 1: The Profitable Product Idea Generator

I have skills in [your expertise areas] and audience interested in [your niche].

Analyze current market gaps and generate 10 digital product ideas that:
→ Solve expensive problems people pay to fix
→ Can be created in under 10 hours
→ Have proven demand on Gumroad or Etsy
→ Require minimal ongoing maintenance
→ Price between $15-$97 for impulse purchases

For each idea, estimate market size and competition level.
Prompt 2: The Customer Pain Point Researcher

I want to create [type of product] for [target audience].

Research their biggest pain points by analyzing:
→ Reddit complaints in relevant subreddits
→ Twitter threads about frustrations
→ Amazon reviews of competing products
→ YouTube comments on tutorial videos
→ Common questions in Facebook groups

List the top 15 pain points ranked by frequency and intensity. Quote real examples.
Read 20 tweets
Guri Singh Profile picture
Feb 10
BREAKING: AI can now script, voice, and animate viral documentaries

Here are 15 insane Claude 4.5 prompts to build a faceless YouTube brand in 2026: (Save for later): Image
Prompt 1: The Viral Topic Miner

I want to build a faceless YouTube channel creating short documentaries in [niche].

Find 10 trending topics that:
→ Have underexploited keyword potential
→ Have 100K–2M monthly searches
→ Are emotional, shocking, or curiosity-driven
→ Can be told in under 10 minutes
→ Have strong comment section engagement

Rank them by viral potential, storytelling strength, and emotional payoff.
Prompt 2: The Netflix-Style Scriptwriter

Write a YouTube documentary script about [topic].
Format as narration + scene direction.

Include:
→ Cold open that hooks in 7 seconds
→ Emotional arc (curiosity, conflict, payoff)
→ 3-act structure like a Netflix episode
→ Voiceover pacing at 140 words per minute
→ Visual timing for 16:9 Sora animations

Each paragraph should be a new visual scene.
Read 17 tweets
Guri Singh Profile picture
Feb 9
STOP SAYING “MAKE A PRESENTATION FOR ME.”

AI isn’t confused.
Your prompt is.

Use these smart prompts instead:
1. THE ANTI-CORPORATE DECK

Build a presentation on [TOPIC] for people who hate presentations.

Rules:
- 1 idea per slide, max 15 words
- Every claim needs a number (not "many" but "47")
- No jargon unless you define it immediately
- Suggest specific visuals ("line graph: 23%→67% in 90 days" not "growth image")

Structure: Hook slide → 6-8 insights with proof → "What this means" → Next step

Tone: Explaining to a smart friend, not reading a script.
2. THE RESEARCH BEAST

Create a presentation on [TOPIC] that looks like 40 hours of research.

Include:
- Real case studies (company names, exact outcomes)
- Stats with sources (institution, year)
- 3-5 "most people don't know" insights
- Competitor comparisons with numbers

Each slide: Surprising claim → Data proof → Why it matters

Make it Stanford-level credible, not Wikipedia summary.
Read 8 tweets
Guri Singh Profile picture
Feb 8
Holy shit... my AI apps were eating 47GB of storage.

ChatGPT, Claude, Midjourney, and voice cloning apps cache entire models locally.

I used Claude to audit and clean it in 5 minutes.

Here are 8 prompts that freed up massive space: ↓
1. Storage Audit Prompt

"Analyze my iPhone storage breakdown and identify which AI apps are using the most space. Focus on apps with cached models, generated content, and offline data. Rank them by storage impact."

Most people have no idea ChatGPT caches 8GB+ locally.
2. Model Cache Clearing

"Write step-by-step instructions to clear ChatGPT's local model cache and conversation history backups without losing my actual chat threads. Include iOS-specific settings paths."

This alone freed 12GB for me.
Read 11 tweets
Guri Singh Profile picture
Feb 7
I ASKED CHATGPT TO FIX MY LINKEDIN PROFILE.

It turned into a recruiter magnet in seconds.
No expensive tools, no paid boost.

USE THESE 7 PROMPTS TO GET HEADHUNTED ON LINKEDIN:
1. Profile Perception Audit

"Analyze my LinkedIn profile as if you're a potential client in [YOUR INDUSTRY]. Tell me exactly how I'm being perceived, what red flags exist, and what messaging is confusing or weak."

Why this works: Most profiles are written for YOU, not your buyer.
2. Headline Transformation

"Rewrite my LinkedIn headline using this framework: [WHO I HELP] get [SPECIFIC OUTCOME] without [COMMON PAIN POINT]. Make it one clear sentence. No buzzwords. No job titles."

90% of headlines are forgettable job titles. This makes you findable.
Read 9 tweets
Guri Singh Profile picture
Feb 5
If Google is trying to make you pay for Gmail storage, do this before you upgrade.

I cleared almost 9GB in one afternoon
(14.9/15GB → 6GB).

No apps. No hacks.
Just a few settings most people never check.

Hope this saves you money like it did for me 👇
1. Find the real storage hogs inside Gmail

Go to your Gmail storage breakdown first:
- Visit drive.google.com/settings/stora…
- Check what's using space across Gmail, Drive, and Photos
- Click on "Gmail" to see your biggest culprits

Then in Gmail search bar, type:
- `has:attachment larger:10M` (finds emails over 10MB)
- Sort by size, select multiple, and delete what you don't need
- Try `larger:5M` or `larger:3M` for a more aggressive cleanup

Just clearing big attachments can free gigabytes instantly.

(FOLLOW ME FOR MORE LIKE THIS)
2. Nuke entire categories in bulk

Gmail sorts emails automatically use this to delete fast:
- Search: `category:promotions` (all promo emails)
- Search: `category:social` (Facebook, LinkedIn notifications)
- Search: `category:updates` (receipts, confirmations)
- Click the checkbox → "Select all conversations that match this search"
- Delete in one shot

These categories pile up over years most are safe to delete.
Read 11 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

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

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

Send Email!

:(