Post

https://twitter.com/1662123349148073984/status/1955246864103887214

More from @heygurisingh

Guri Singh

@heygurisingh

Mar 21

BREAKING: Meta's AI team uses a prompting method internally that they never talk about publicly.

It's called "Negative Prompting." You tell the AI what NOT to do.

My output relevance: 5/10 → 9.4/10

Here's how it works:

Most people prompt like this:

"Write me a professional LinkedIn post"

"Give me a meal plan for weight loss"

"Summarize this article for me"

You're telling the AI what to do. But you're never telling it what to avoid.

So it defaults to every generic pattern it learned during training.

You get bland, predictable, forgettable output.

Negative Prompting flips this.

Instead of only describing what you want, you explicitly define what you don't want.

LLMs are pattern matching machines. Without boundaries, they match the most common patterns.

When you add constraints, you force the model to search for less obvious, higher quality outputs.

Boundaries create better thinking.

Read 11 tweets

Guri Singh

@heygurisingh

Mar 20

BREAKING: OpenAI and Anthropic engineers leaked a prompting technique that separates beginners from experts.

It's called "Socratic prompting" and it's insanely simple.

Instead of telling the AI what to do, you ask it questions.

My output quality: 6.2/10 → 9.1/10

Here's how it works:

Most people prompt like this:

"Write a blog post about AI productivity tools"
"Create a marketing strategy for my SaaS"
"Analyze this data and give me insights"

LLMs treat these like tasks to complete.
They optimize for speed, not depth.

You get surface-level garbage.

Socratic prompting flips this.

Instead of telling the AI what to produce, you ask questions that force it to think through the problem.

LLMs are trained on billions of reasoning examples.
Questions activate that reasoning mode.

Instructions don't.

Read 12 tweets

Guri Singh

@heygurisingh

Mar 19

I gave Perplexity the same task every day for 90 days straight.

By day 30, I had replaced 3 software subscriptions.
By day 60, I automated half my workflow.
By day 90, I was earning $2K/month from systems Claude built me.

Here are the 12 prompts that made it all possible:

1. The "Second Brain" Strategy Prompt

"You are a senior business strategist. I'm going to describe my current workflow, tools, and recurring tasks. Analyze everything and give me:

- 5 tasks I should automate immediately
- 3 tools I'm paying for that you can replace
- A weekly system I can follow using only you

My workflow: [paste your daily/weekly routine]"

This one prompt saved me $147/month in software.

2. The "Content Machine" Prompt

"Act as a viral content strategist who has grown 10+ accounts past 100K followers. I'm going to give you my niche, audience, and voice.

Create a 30-day content calendar with:
- Daily post hooks
- Thread ideas (1/week)
- Engagement-bait tweets (2/week)
- A CTA strategy that builds my email list

Niche: [your niche]
Audience: [your audience]
Voice: [casual/authoritative/provocative]"

I stopped spending 2 hours/day on content. Claude does it in 4 minutes.

Read 14 tweets

Guri Singh

@heygurisingh

Mar 17

Holy shit... researchers just proved that vibe coding is destroying the internet's visual diversity.

University of Washington studied AI-generated apps and found something terrifying:

The title? "Interrogating Design Homogenization in Web Vibe Coding."

And the findings are devastating.

What they found is simple:

Vibe coding isn't just making it easier to build apps.

It's making every app look exactly the same.

Not similar. Identical.

The web is losing its visual diversity faster than at any point in internet history.

To understand why, you need to know about something called the "fixation effect."

When an LLM generates your first design -- with its clean layout, rounded corners, and Tailwind defaults -- it looks SO polished that your brain stops pushing back.

You accept it. You ship it.

Read 14 tweets

Guri Singh

@heygurisingh

Mar 16

MY RESUME GOT REJECTED 14 TIMES IN A ROW.

So I gave Claude my resume + the job descriptions.

3 hours later, interview callbacks from 4 companies.

No career coach. No $500 resume service. Just 7 prompts that completely rewrote my job search:

1. Resume Surgeon

Prompt: "Here's my resume and the job description I'm applying for. Rewrite my resume to match this role's exact keywords, tone, and requirements without lying about my experience. Make every bullet point prove impact with numbers."

2. ATS Killer

Prompt: "Analyze this job posting and extract every keyword, skill, and qualification mentioned. Now compare it to my resume and tell me exactly what's missing, what to add, and what to rephrase to beat the ATS filter."

Read 10 tweets

Guri Singh

@heygurisingh

Mar 15

2 Billion people use Google Maps every month.

But 90% of them are using it like it's 2015.

Here are 15 hidden features that'll make you feel like you just unlocked a cheat code: 👇

1. HistoryPin Integration

Google partnered with HistoryPin to let you overlay historical photos onto real-world locations.

Pin old photos, explore how neighborhoods looked decades ago, and share stories tied to specific places on a global map.

2. AI-Powered Conversational Search

You can now chat with Google Maps like you'd chat with an assistant.

Ask "best rooftop cafes near me with Wi-Fi" and it gives you tailored results -- not just a list of pins.