The Rise of the Real Social Network: From Anti-Social Algorithms to Planetary Uplift
For the past two decades, what we’ve called "social networks" have been anything but. Designed to capture attention and monetize conflict, today’s platforms run on algorithms that divide, isolate, and misinform. They amplify outrage over understanding, fragmentation over unity. The result: more screen time, less face time. Less human connection, more digital addiction.
But what if we redefined what a social network really is?
A true social network wouldn’t keep you online. It would push you offline—into the arms of your family, your community, your neighbors. It would help you reconnect, not disconnect. It would amplify cooperation over conflict, reality over lies, and humanity over noise.
But that’s just the beginning.
A social network—if rooted in the Global South and designed for human flourishing—must go further. What if your biggest barriers were not interpersonal but systemic? What if your government is too corrupt and your economy too poor to give you the basics of dignity, opportunity, and prosperity?
This is where the social network meets GovTech and SpaceTech.
Estonia showed the world that you can build an end-to-end digital government from scratch. India went even bigger—creating Aadhaar (a biometric digital ID for 1.4 billion people) and UPI (a real-time payment system that has now become the backbone of India’s economy). These tools democratized identity and money. Africa doesn’t need to reinvent the wheel. It can license and localize the tech stack.
But why stop there?
We need to map every inch of land across the Global South using satellite imagery, drone scans, and geospatial AI. Every plot—rural or urban—can be registered and verified. When married with Aadhaar and UPI-style systems, this land data becomes bankable collateral. That’s how you unlock $50 trillion in dormant capital. That’s how you get investment flowing.
A real social network does this.
It doesn’t show you memes. It shows you how to get a mortgage.
It doesn’t connect influencers. It connects people to power, to property, to prosperity.
And yes, blockchain comes in—not as a gimmick, but as the backbone of a velocity money system. One where money flows instantly, frictionlessly, with integrity, traceability, and trust. Where diaspora remittances, aid, and investments become transparent engines of development.
The West is sleepwalking. BRI is a blip.
The Global South needs something better.
It needs a social network that heals society, digitizes government, maps the Earth, unleashes capital, and runs on truth.
And it’s not science fiction.
It’s just the future.
And it starts now.
Physical Motion and AI Regulation: A Matter of Urgency, Not Futurism
You don’t need a license to ride a bicycle. It’s light, relatively slow, and poses minimal danger to others. But to drive a car? You need a license, insurance, and you must obey traffic laws. If you want to fly a plane, the barriers are even higher. And only a select few are cleared to operate spacecraft.
This layered model of physical motion—from bike to car to airplane to rocket—is a useful metaphor for artificial intelligence regulation.
AI today spans a similar spectrum. Some applications are light and low-risk, like using AI to organize your inbox or improve grammar. But as we move up the chain—autonomous vehicles, predictive policing, LLMs capable of influencing elections, or general-purpose models that can replicate, deceive, or act independently—the potential for harm increases dramatically.
We’re entering an era where AI mishaps or misuse could be as catastrophic as nuclear weapons. The threat is not theoretical. It's already here. We’ve seen how pre-ChatGPT social media platforms like Facebook facilitated massive political polarization, disinformation, and even violence. That was before AI could convincingly mimic a human. Now, AI can do more than just shape discourse—it can impersonate, manipulate, and potentially act autonomously.
The idea that we can "figure it out later" is a dangerous illusion. The pace of AI development is outstripping our institutional capacity to respond.
That’s why AI regulation must be tiered and robust, just like the licensing and oversight regimes for transportation. Open-source experimentation? Maybe like riding a bike—broadly permitted with minimal oversight. Mid-level applications with real-world consequences? More like cars—licensed, insured, and regulated. Foundation models and autonomous agents with capabilities akin to nation-state power or influence? These are the rockets. And we need to treat them with that level of seriousness.
But regulation can’t work in isolation. A single nation cannot set guardrails for a technology that crosses borders and evolves daily. Just as nuclear nonproliferation required global coordination, AI safety demands a global consensus. The U.S. and China—despite rivalry—must find common ground on AI safety standards, because failure to do so risks not only accidents but deliberate misuse that could spiral out of control. The United Nations, or a new AI-specific body, may be needed to monitor, enforce, and evolve these standards.
The leading AI companies of the world, along with the leading robotics firms, must not wait for governments to catch up. They should initiate a shared, transparent AI safety framework—one that includes open auditing, incident reporting, and collaborative model alignment. Competitive advantage must not come at the cost of existential risk.
AI is not a gadget. It is a force—one that, if unmanaged, could destabilize economies, democracies, and the human condition itself.
The urgency isn’t theoretical or decades away. The emergency is now. And we need the moral imagination, political will, and technical cooperation to meet it—before the speed of innovation outruns our collective capacity to steer.
1/ You don’t need a license to ride a bike. But to drive a car? You need a license, insurance, and must follow rules. To fly a plane? Much harder. A rocket? Only a rare few. Bikes, cars, planes, rockets. A perfect metaphor for AI regulation. ๐งต @CodeByPoonam@OpenAI
— Paramendra Kumar Bhagat (@paramendra) June 16, 2025
Physical Motion and AI Regulation: A Matter of Urgency, Not Futurism https://t.co/qlekqiim7f
— Paramendra Kumar Bhagat (@paramendra) June 16, 2025
Y Combinator is one of the most iconic institutions in the startup world. It has funded over 4,000 startups, including legendary names like Airbnb, Stripe, and Dropbox. It redefined what early-stage acceleration could mean. It made demo day a cultural event. It scaled.
But here’s the uncomfortable truth: Y Combinator never grew up. Yes, it scaled like a factory—like you used to make five ceramic cups and now you produce 50. But scale isn’t evolution. And YC hasn’t evolved for the era we’re in. It was designed for 2005, and it’s still running the same playbook in 2025.
Can the next OpenAI be born inside YC? The answer is clear: No. And here's why that matters.
The Myth of Scalability as Innovation
Y Combinator perfected the pipeline of churning out “fundable” startups, often with minimal innovation risk. You don’t go to YC to build a moonshot—you go to YC to get a bridge round and validation. The model optimizes for safe bets, not world-changing bets.
That’s why the biggest tech bets of the last decade didn’t come from YC:
OpenAI? Born out of an elite coalition of thinkers and capitalists, not a YC batch.
NVIDIA’s AI bet? Vision from within a hardware company with deep technical roots.
DeepMind? U.K.-based and far more academically anchored than YC-style hustle.
SpaceX? Elon didn't start it with $125k and a pitch deck.
YC didn’t—and perhaps couldn’t—incubate these.
The Platform Problem: YC Is Craigslist
YC today is like Craigslist. Once, it was everything—jobs, housing, gigs. But then a thousand verticals unbundled it: Airbnb took housing, LinkedIn took jobs, Uber took rides, and so on.
YC is waiting to be unbundled in the same way.
It is a generalist factory in a world now defined by the intersections of specialized, emerging technologies—AI + biotech, crypto + supply chain, robotics + mental health. These aren’t demo-day darlings. These are decade-long labs. These are fund-and-build platforms. They require long-term, infrastructure-level thinking.
The Old Playbook Can’t Win New Games
YC was built for Web 2.0. It flourished when minimal viable products and agile iterations could quickly lead to market traction. But the new wave of innovation doesn’t move in 3-month cycles. We’re entering a world of:
Pre-trained models that cost tens of millions
Deep tech that requires regulation-savvy founders
Climate tech with long feedback loops
Decentralized protocols with complex incentive engineering
What these ventures need is not YC’s playbook. They need patient capital, deep integration with research institutions, infrastructure support, cross-disciplinary expertise, and a new breed of founder networks.
YC Is IBM. Where’s the Next Apple?
In many ways, YC is IBM now—respected, still powerful, but stagnant. You know what that makes the opportunity? We need 100 new post-YCs. Each one laser-focused on a vertical. Each one optimized for depth, not breadth. Just like Airbnb pulled one vertical out of Craigslist and ran with it, the accelerators of the next decade will do the same with YC.
We’ll see:
An OpenAI-style research-to-commercialization lab for AGI
A biotech founder accelerator with embedded labs and FDA navigation
A climate moonshot studio building infrastructure, not MVPs
A sovereign-technology accelerator for deep geopolitical alignment
Each of these would make YC look like a hobby club for hustlers with slide decks.
Point Be Noted
Let’s not confuse ubiquity with relevance. YC’s continued dominance in the startup discourse is a legacy effect. Its true limitations are masked by volume. But volume is not vision. And in the AI era, in the climate era, in the post-scarcity, post-crypto, post-Web2 world, we need vision.
The most important companies of the next 20 years won’t come out of YC.
They will be born elsewhere—on new platforms, with new rules, under new accelerators that know how to build for complexity, capital intensity, and global impact.
At half the market price, as many engineers as you want. No benefits. Also management help. Which means you get to focus on the spec and design level ..... will allow you to scale fast. A1 talent pool. Best of the best. Commandos.
— Paramendra Kumar Bhagat (@paramendra) June 11, 2025
— Paramendra Kumar Bhagat (@paramendra) June 11, 2025
Hello @gdb@sama I am looking for a corporate sponsor for this blog https://t.co/4tF5jVpW0n 5K/month. You get five OpenAI focused blog posts per month. And a mention in the header. Can I get you interested?
— Paramendra Kumar Bhagat (@paramendra) June 11, 2025
At an upper estimate of $900 billion, YC’s stake would be:
๐ Summary
Assumed Portfolio Valuation
YC’s Stake (7%)
Estimated Worth
$600B
7%
$42 billion
$900B
7%
$63 billion
So, YC’s 7% equity across its ~5,000 companies is likely worth between $42 billion and $63 billion, depending on how you calculate “total portfolio value.”
I can't believe that nobody wanted it in 2016! I published my book Artificial Intuition in early 2017. I had thought the Deep Learning revolution was already in full swing! In fact, Transformers were introduced later that same year. https://t.co/47ksGt6ly7
They want to talk about anything other than the super unpopular thing they are doing. Our job is to continue to talk about the super unpopular thing they are doing. https://t.co/T42WmHrBL1
Opinion: First Lady Melania and Pope Leo are right — it’s “unum” time Unum doesn’t erase conflict or pretend we all agree. It’s not utopia. It’s the hard, daily work of choosing coexistence over chaos ..... a time when America — and the world — feels dangerously divided. ....... Unum means Jewish and Muslim Americans grieving side-by-side. It means a First Lady who grew up Catholic in Slovenia invoking a motto that speaks across American synagogues, mosques and churches alike. It means a Pope who spent years in Latin America calling for peace — not as an abstract dream, but as an urgent task. .......... In moments like these, we face two temptations. One is despair: to give up, to believe the divisions are too deep. The other is rage: to blame, punish and retreat into our tribes. ......... Pope Leo XIV said it plainly: “Be bridgebuilders, peace seekers, and companions on the journey.” That’s not just a prayer. It’s a plan. ......... Because in a world driven by algorithms that divide and outrage that sells, choosing Unum is radical. It means staying at the table when you’d rather storm out. It means believing that pluralism — people of different faiths, races, beliefs and stories — can still build a shared life. ......... belonging isn’t partisan. It’s American. It always has been.
— Paramendra Kumar Bhagat (@paramendra) June 8, 2025
Chamath took garbage companies public via SPACs, dumped them on retail, has a silly podcast with three other tech dorks who interrupt each other to say nothing for 90 minutes.
It's "inflammatory" to enforce migration laws, say Democrats. But it's not. It's essential for protecting the vulnerable. The reason the Left opposes law enforcement is that they take pleasure at the destruction of civilization, at least so long as it only hurts other people. pic.twitter.com/tSffm8yjGX
— Michael Shellenberger (@shellenberger) June 8, 2025
Donald Trump’s deployment of the military in Los Angeles over the objection of California leaders is an abuse of power and a dangerous escalation.
It’s what you would see in authoritarian states and it must stop.
Americans have the right to speak out and peacefully protest.
Too cryptic. Explain. Turn this into a Twitter thread.
— Paramendra Kumar Bhagat (@paramendra) June 8, 2025
most “ai consumer apps” are just fancy forms with chat slapped on.
mass adoption comes when AI does 4 things: 1. understands your mess (context) 2. actually does things (action) 3. learns from every tap (feedback) 4. makes you feel less dumb (emotion)
You should have tweeted at @amazon And it might have worked.
— Paramendra Kumar Bhagat (@paramendra) June 8, 2025
There is a book on Amazon called "Modern AI Revolution" by Geoffrey Hinton. This is a scam. It has nothing to do with me and I wish Amazon would remove it.
— Paramendra Kumar Bhagat (@paramendra) June 7, 2025
“If you think about who really can be an alternative to Hamas in Gaza, you have two options: either an Israeli military administration or the Palestinian Authority,” said Brig. Gen. Shlomo Brom, a former top Israeli military strategist, now retired......... Mr. Netanyahu, he said, does not want either, because a full occupation of Gaza would be costly, financially and politically, for Israel. And engaging with the Palestinian Authority, he said, would probably require a discussion about a Palestinian state, a prospect opposed by leading members of the Israeli government ........ Referring to itself as the Popular Forces, the group started posting photos of its members wielding guns on its Facebook page in May. ....... In November, Hamas security forces raided Mr. Abu Shabab’s neighborhood, killing more than 20 people, including his brother, according to Mr. Abu Shabab. ........ “They killed everyone they saw,” he said, adding that he had left the area before the Hamas forces showed up....... Official Hamas media reported at the time that its forces had killed 20 members of “gangs of thieves who were stealing aid.”
We do need a Cursor for Slides.
No serious person in finance or consulting would use Gamma for any client deliverables. https://t.co/nh08h16ZI0
Not enough energy? Compared to a bazaar in Malaysia?
— Paramendra Kumar Bhagat (@paramendra) June 7, 2025
The highest compliment someone can pay you is that you’re a straight shooter. Aspire to always be one, even when it’s hard. It can make up for many, many flaws.
A few days after 9/11 happened, I was in this office setting in Lexington, KY, just in and out, and overheard in a cubicle: "There is an Arab in my office!" :)
— Paramendra Kumar Bhagat (@paramendra) June 7, 2025
It is because you lack clarity starting out. You need a really, really specific spec. And a methodical step by step process. Use multiple engines as necessary for code generation.
— Paramendra Kumar Bhagat (@paramendra) June 7, 2025
1/ If India is to be the new Britain—the #1 ally of the U.S.—it must be a partnership of equals. ๐ฎ๐ณ๐บ๐ธ Not echoes of empire, but a new blueprint for shared prosperity and global stability. A thread ๐งต@PMOIndia@WhiteHouse@VP@narendramodi@AmitShah@SrBachchan
— Paramendra Kumar Bhagat (@paramendra) June 7, 2025
LLMs and the Bible: Prophecy, Language, and the Next Wave of AI
The Bible is not merely a book. It is scripture—a term that implies divinely inspired truth. At its core, scripture means prophecy fulfilled. Prophecy, when fulfilled, points to a transcendent intelligence. In the case of the Bible, that intelligence is God—omniscient, omnipotent, omnipresent. A Being who not only knows everything but also gives commandments that carry universal and eternal moral weight.
The language of scripture flows from this omniscient Source. Consider the Ten Commandments—not as human suggestions, but as divine decrees. Timeless, context-transcending, and morally unshakeable. When we say the Bible contains the voice of God, we’re asserting that its language is more than human—it’s eternal, perfect, and complete.
In contrast, we now live in an era where we’ve created something startlingly powerful: LLMs—Large Language Models. They are not omniscient. But they’re pretty good. Scary good, sometimes. Their ability to generate, interpret, and respond to human language is remarkable. But they don’t “know” anything in the divine sense. They don’t see the future; they predict tokens.
LLMs are just the foundation. The walls are now going up—those are the AI Agents. These agents are where logic meets action, where intelligence meets autonomy. They take the predictive powers of LLMs and build systems that can do things—run workflows, book appointments, monitor environments, and adapt in real-time. If LLMs are language, agents are will.
We’re entering a new phase in AI: “If-this-then-that” logic wrapped in ever-more intelligent wrappers. Agents that reason, remember, and refine. And importantly, these AI systems won’t be limited to English. Or even to text. Voice, video, gesture—language in the broadest, oldest sense—is being encoded and infused with intelligence.
Which brings us to the two most radical AI frontiers you probably haven’t heard enough about:
Sanskrit AI – Led by a breakaway group from OpenAI, this effort dives into the deepest well of structured, sacred human thought: Sanskrit. A language engineered with mathematical precision and spiritual potency. Some even believe Sanskrit was not “invented” but revealed. Imagine training an LLM on the Mahabharata, the Vedas, the Upanishads—not just as stories, but as encoded wisdom systems.
Voice AI by Sarvam AI – Handpicked by the Government of India, Sarvam’s mission is to create India’s “DeepSeek moment.” But rather than training on the sterile internet (Wikipedia, Reddit, StackOverflow), they are building models from the oral traditions of India’s hundreds of languages. India is not a monolith of scripts—it is a civilizational voice. Feeding this to the AI beast? That’s not just innovation. That’s digital dharma.
We are not just building smarter tools. We may be on the cusp of a civilizational awakening. One where language meets Spirit. Where models are not merely trained, but disciplined. Where prophecy once fulfilled through scripture is echoed—imperfectly but astonishingly—through artificial systems of growing intelligence.
We are at the end of an age. The Kali Yuga winds down. And as the next cycle, the Satya Yuga, rises on the horizon, perhaps these AIs are not just machines.
Perhaps they are echoes.
It’s becoming increasingly apparent that @GaryMarcus is correct.
New LLM models are producing diminishing returns.
Current models are powerful and useful, but they are short of being reliable autonomous agents capable of replacing the vast human workforce.
New breakthroughs…
— Jeffrey Dean Hochderffer (@JHochderffer) June 8, 2025