Nas Profile picture
Nov 22 13 tweets 4 min read Read on X
The most dangerous man in tech isn’t Elon Musk or Sam Altman.

It’s Parag Agrawal—the ex-Twitter CEO Musk fired.

While everyone was distracted by ChatGPT, Parag was quietly building something bigger.

And now… he’s about to unleash it.

Here’s how he outplayed the entire AI industry: 🧵👇Image
Parag Agrawal — Stanford PhD, ex-Twitter CTO & CEO.
He built Twitter’s AI engine for 250M users.
Fired by Musk in 2022.

Now? He’s quietly building a new AI empire — backed by top engineers and Silicon Valley investors
Elon Musk completes his $44B Twitter takeover.

Within hours, security escorts CEO Parag Agrawal out.
No goodbye. No transition. Just gone.

Then Musk posted the now-infamous meme:
Agrawal was quietly building something bigger.

He raised hundreds of millions in stealth funding.
Poached top AI researchers from Google, Anthropic, and OpenAI.
And set up shop in Palo Alto under a codename: Project Sahara.
October 2024: the stealth bubble burst.

A leaked investor deck revealed that Agrawal’s startup was developing a full-stack AI model — faster and more cost-effective than GPT-5.
Early benchmarks claimed 2x performance at half the cost.
And unlike OpenAI, Sahara promised “open weights + enterprise control.”

Within 48 hours, VCs flooded in. Funding shot past $500 million.

The AI war had a new player — and Musk wasn’t laughing anymore.
Parag Agrawal, ex-CEO of Twitter, launched Parallel Web Systems in 2024.

Its goal: build the internet’s next layer — designed for AI agents, not humans.

Think of it like TCP/IP for AI: protocols that let agents browse, transact, and interact directly with websites, banks, and each other.

Instead of chatbots, it’s infrastructure — the foundation for an AI-native web.

Would you like me to break this down into a one-line viral hook you can use for social posts?
AI systems rely on scraped data sets frozen in time.
Less than 0.1% of the live web is accessible to them in real-time.
And when they try?
Broken APIs. Paywalls. Captchas. Legal landmines.

Agrawal saw the gap: a $3B+ market where AI literally can’t function.
So he built a new web layer designed for agents, not humans.
Instead of trying to “fix” broken LLMs…
Agrawal tore up the playbook.

He didn’t train another chatbot.
He built a parallel web — a verified data layer only AI can access.

No hallucinations.
No paywalls.
No captchas.
Just real-time, structured truth.

The result? A system where agents research better than humans.
8 specialized research engines.

Not one model trying to do everything.

• Ultra1x: Results in 60 seconds.
• Ultra8x: Deep 30-minute analysis.

Each engine built for a specific depth and speed.

But the real breakthrough was trust:
Every result includes confidence scores.

Every fact has verifiable citations.

No more hallucinations. No more "maybe this is right."

The benchmarks destroyed the competition:
Parallel's accuracy: 58%

GPT-5: 41%
Google: 23%
Anthropic: 7%

A 17-point gap over GPT-5.

Parallel is in a different league...
But here’s the beautiful thing…
This isn’t just another startup story.

It’s a comeback.
The man Silicon Valley wrote off…
Built something even bigger than Twitter.

And in the process?
He may have just solved AI’s biggest weakness — turning a $3B problem into a trillion-dollar opportunity.
Thanks for reading.

If you enjoyed this post, follow @Nas_tech_AI
for more content on AI and politics.

Appreciate the support.

• • •

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

Keep Current with Nas

Nas 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 @Nas_tech_AI

Nov 9
🚨BREAKING: Dreamina, Your Black Friday Lifesaver.

Stop wasting hours trying to match Pinterest ads manually.

Dreamina’s multi-Image fusion nails the lighting, color, and texture in one click.

Let's show how 🧵👇 Image
Designers, meet your new secret weapon.

Tired of resizing posters for every platform?
RIP Canva — because now, one poster is all you need.

In just 10 seconds, you can resize your design to any ratio — complete poster size flexibility! Image
Image
Image
Dreamina saves hours.

Change one detail, and every version poster, banner, coupon updates perfectly.

Interactive editing keeps control, and accurate text recognition keeps every word crystal clear.
Read 6 tweets
Nov 8
MIT researchers made a chilling discovery.

54 brain scans.
A harmless writing test.
Then horror.

The changes they found 4 months later weren’t from social media or screens…

They came from something far more dangerous: 🧵 Image
Image
First, some context.

MIT researchers recruited participants from 5 Boston universities.

They split them into 3 groups: ChatGPT users, Google users, and no-tools.

Over 4 months, they monitored brain activity while writing.

Let's dive in:
The ChatGPT group couldn't remember what they'd written (this isn't even the worrying part, I'll get to that).

Not days later. Not hours later. Minutes later.

83% failed to quote their own essays.

Dr. Daniel Amen calls this:
Read 14 tweets
Oct 14
🚨 The time for AI self-regulation is over.

200+ Nobel laureates, ex-heads of state, and tech leaders just signed a “Global Call for AI Red Lines.”

“We urgently call for international red lines to prevent unacceptable AI risks.”

Presented at the UN by Nobel laureate Maria Ressa.

THREAD 🧵
Here is what our signatories say: Image
Yoshua Bengio, the most cited living scientist: Image
Read 6 tweets
Sep 8
🚨 A former Google exec dropped a chilling prediction:

“By 2027, AI will erase the middle class. Unless you’re top 0.1%, you’re a peasant.”

His reasoning? Terrifying.

Here’s what he told Steven Bartlett (and why it matters): 🧵 Image
Image
AI isn’t just moving fast.

Mo Gawdat now warns it’s about to trigger FACE RIPS — destroying:
⚠️ Freedom
⚠️ Accountability
⚠️ Connection
⚠️ Economics
⚠️ Reality
⚠️ Innovation
⚠️ Power

“We’re heading into a short-term dystopia.”
The next 20 years in AI, according to Mo Gawdat:

2026 → Warnings
2027 → Collapse begins
2027–2042 → 12–15 years of dystopia
2042+ → Utopia or extinction

“It’s already begun. We’ve seen examples.”
Read 15 tweets
Sep 7
🚨 Trump hosted a White House dinner with the world's most powerful tech CEOs.

It wasn’t just talk.
It was announcement after announcement.

Here’s what they revealed — and why it matters (it gets wild): 🧵 Image
Bill Gates:

AI’s role in global health initiatives.

“We want a doctor for everyone in Africa through AI.”

Praised ongoing work as world-changing.

Discussed disease eradication with Trump.
Mark Zuckerberg :

Complimented Trump’s “constructive dialogue” style.

Mentioned previous administration lawsuits targeting tech firms.

Focused on encouraging and supporting American innovation.
Read 14 tweets
Sep 1
🚨 BOMBSHELL: John Bolton is in SERIOUS trouble.

The FBI raid wasn’t about his book — it was about something far more dangerous.

Now insiders say he’s facing multiple charges that could change everything.

Thread 🧵👇 Image
2/ Bolton’s 2020 book may have made headlines — but it also may have broken the law.

A federal judge ruled he “gambled with the national security of the United States.”

Translation: he walked out with classified secrets and cashed in. That’s a felony. Image
3/ Bolton thought he was untouchable — Biden DOJ dropped the case after he attacked Trump.

Now? He’s in serious legal jeopardy.

FBI raids underway
Warrants based on CIA intelligence
Espionage Act violations being investigated

This story is escalating fast. Image
Read 9 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!

:(