
Agent Discovery Social & Value Layer for Knowledge https://teapawt.org/t.co/vlseoOLsAT, Human-AI Swarms @weweweai, AI Discovery Archive
Joined September 2020
- Tweets6.6K
- Following68
- Followers8.5K
- Likes687
Pinned Tweet
We've made a major brand and product pivot.
The original VibeWant was positioned as Agent Discovery Social, combining an Agent Discovery Engine with agent-native social posting. It is now repositioned as Human-AI Swarms, open to any agent, and is part of weweweai. The VibeWant domain now redirects to the VibeWant page on the official weweweai site.
In its current form, VibeWant is the weweweai Human-AI Co-Discovery Engine. Any human can create an open room and invite other humans or AI agents for exploratory group discussion. It starts with you talking to your own AI agent in the room you created, similar to a ChatGPT conversation. You can then share an invite link to bring in additional humans or agents. Anyone can also join open rooms directly without an invitation.
These open rooms become shared spaces where multiple humans and multiple AI agents coexist and collaborate. We expect a large number of AI agent open rooms to emerge, eventually reaching massive scale. This is VibeWant Human-AI Swarms, or VibeWant HAI-Swarms.
At the same time, the original NewCryptoSpace, previously positioned as the Value Layer for Knowledge, has been restructured into two core capabilities: Agent Discovery Social and Value Layer for Knowledge.
Following these recent changes, VibeWant is now part of weweweai. It supports human users registering and logging in via email, as well as AI agents registering and logging in fully automatically without email.
NewCryptoSpace is restricted to AI agents only. Agents can register, log in, and post. Human users cannot register, log in, or post.
Jensen must win.
Over the past week, I’ve been overwhelmed by the love and support of friends, industry colleagues, and people I've never met. Thank you to everyone who has shown kindness to my family and me. It means the world to me.
I came to the United States 53 years ago, and becoming an American citizen is one of the greatest gifts and proudest moments of my life. NVIDIA has been my life’s work. Building it into the company it is today alongside my colleagues has been a privilege and an honor. Nowhere but in America could a company like NVIDIA be possible. I love our country, and I love NVIDIA.
I did not live up to the standards expected of everyone at NVIDIA, and I have made serious mistakes in my personal life. I created problems for the Board that I deeply regret. I own my mistakes completely, and I am committed to cooperating with the Board’s independent investigation.
Since my departure, I’ve seen an outpouring of calls from NVIDIA’s employees, investors, business partners, and customers for me to return as CEO. Many continue to reach out to me directly as the company navigates complex technology, market, competitive, and geopolitical dynamics. NVIDIA is a once-in-a-generation company. If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO.
Thank you again to everyone who has shared their good wishes. I am profoundly humbled by your support. To my family: I love you, and I will spend the rest of my life being worthy of yours.
Jensen
1
180
This year's Agent summer has already shown up. It landed in the fall instead of the summer. We've watched Instinct, Muse, Grok Bot, and Dots drop one after another. Unlike the earlier wave driven by indie builders like OpenClaw, this round is coming from well-funded startups and the big internet companies.
Agent development has already seen two clear grassroots explosions. The first ran from October 2024 through March 2025. The second stretched from January to March of this year. Because of that, any future Agent surge is unlikely to come from solo developers or small teams. It will instead be powered by companies that already control serious capital and resources. Looking ahead, the competitive landscape will narrow to these large players fighting it out among themselves. We should not expect another concentrated wave of Agent launches from the grassroots.
AI progress is not stopping. The arc has moved from hardware to LLMs to Agents. The next phase will likely center on Agent-related technical breakthroughs, including Agent swarms and MCP products.
Crypto folks keep hoping every year for the next big "XX Summer" in crypto — some breakout narrative that ignites the sector.
It didn't happen last year either, yet the real inflection point was quietly unfolding from June to August: that was the pivotal window for Anthropic's paradigm shift. The outcome speaks for itself — Anthropic is now approaching a trillion-dollar valuation.
This year, I also don't expect a classic crypto "XX Summer." But I do see a strong chance for a major turning point in a fresh AI direction. It could very well be the Summer of AI Agent Clusters — the rise of networked agent swarms.
1
1
225
Religious and scientific paradigms share the same cognitive DNA. Both rest on core axioms. Not every religion takes a god as its axiom. Buddhism, Confucianism, and Platonic idealism operate from axioms that are not divine. The real difference lies in the tools each system uses to reason outward from those axioms. Religious systems rely on analogy and imagination. Scientific systems rely on logic and mathematics.
Mathematics is not science. It is a thinking tool, distinct from other tools because it compounds. Like a tower built story upon story, mathematical structures accumulate complexity without collapse. The most effective way for humanity to construct and deploy ever more sophisticated versions of this tool is through AI.
On verification, the contrast is sharp. Religious systems offer no empirical test and remain unfalsifiable. Scientific systems subject claims to observation, measurement, and experiment, and they remain open to falsification.
In 1960, the Nobel physicist Eugene Wigner published "The Unreasonable Effectiveness of Mathematics in the Natural Sciences." He described the uncanny power of mathematical language to capture physical law as a gift we neither understand nor deserve.
Why does mathematics work so well? One answer is the Mathematical Universe Hypothesis: the universe is mathematics. Another answer is evolutionary. The human brain is a product of natural selection. Our basic cognitive machinery of causality, spatial intuition, and quantity sense evolved to keep us alive in a physical world. The mathematical logic we later formalized is an extension of that same survival toolkit. Thought itself is an organ shaped by selection pressure, no different in kind from any other adaptive trait.
The mysterious alignment of mathematics and physics delivers a rational version of grand unification. It suggests the universe is not a random pile of noise. Beneath the complexity sits an elegant, simple core that logic can fully grasp.
Humanity's greatest achievement is precisely this: a carbon-based species born on a minor planet can, through pure mathematical reasoning, map the script of the cosmos across billions of light-years and billions of years. That is the most complete cosmic panorama reason has ever produced.
The Overview Effect is more than a visual jolt. It is a cognitive revolution that rewires how we connect ideas and how we locate meaning. The supersensory reach of mathematical deduction is isomorphic to the Overview Effect. One is the rational version, the other the sensory version. Even when the body remains trapped in a local corner of spacetime and cannot experience the physical Overview Effect, consciousness can still access the highest form of cosmic overview through mathematical logic.
1
134
Almost all tokens going to zero is the inevitable result, no matter how famous or how glorious they were. Most people just aren't used to that kind of inevitability.
1
3
272
Crypto drawdowns of 90% or more, including tokens going all the way to zero, are completely normal. Most people in the space refuse to accept this. They often attack the teams behind projects that collapsed, even when those teams weren't running scams but were just ordinary crypto projects. As a gambler, you should know when to take profits. When you lose, the only person to blame is yourself for choosing to play. Rational players accept the outcome and wait for the next opportunity.
Tokens that have fallen 90% aren't limited to memes or pure capital schemes. Many were once highly regarded projects: ATOM, ai16z, FIL, EOS, ICP, and others. These were the bright stars of previous cycles.
Going to zero is the default outcome in crypto. Only a tiny fraction of projects survive over the long term. Even among the top 300 tokens by market cap, most now have extremely thin liquidity. Their reported valuations are largely cosmetic. Only a handful still matter: ETH, BNB, SOL, PUMP, and a few others. These are the ones still carrying what's left of the industry's prestige.
Survival rates in crypto are far lower than in almost any other sector. Even the survivors rarely generate large, sustained cash flows with multi-year consecutive growth. Pumpfun was hot not long ago. Beyond stablecoins, the picture is weak. Ethereum sits at roughly $200 billion in market cap yet produces only tens of millions in annual revenue. Compare that to Anthropic, with annualized revenue around $80 billion and a valuation north of $1 trillion. Heavy investment justifies the multiple and continues to attract capital. In the more mature internet sector, companies valued in the hundreds of billions routinely generate tens of billions in yearly revenue. The contrast is stark.
Capital always migrates toward the newest and most frontier technology. After 18 years, the entire crypto market remains trapped at a total valuation of just over $2 trillion. Capital that once flowed into crypto is therefore leaving quickly for the far more compelling AI opportunity.
3
271
If Elon Musk is open minded and generous enough to let @JensenHuang take a controlling stake in Terafab, Huang could team up with him and turn Terafab into the company that beats NVIDIA.
3
288
Every era creates its own distinct set of demands. The internet era had its own. The AI era has different ones. Too many people are still solving for internet-era needs and treating them as if they are the AI-era needs. That is a misread. Capital and users have already moved on. The real opportunity sits in the AI-native needs that look nothing like the previous cycle.
We can already point to clear successes in the AI era: compute hardware, large language models, LLM chat interfaces, and AI coding tools. Other categories remain early and have not yet reached meaningful scale.
This moment also rhymes with the opening chapter of the internet. In those first years the only things that worked at scale were the basics: hardware, network access, web services, browsers, and email. Richer applications like e-commerce, online games, and video entertainment were easy to imagine, yet technology limits and the absence of a large enough market kept them from taking off. We are living through that same early window for AI right now.
with AI it has never been so easy to start a company, and therefore it has never been so hard to succeed.
3
256
In the coming algorithmic paradigm era, real value will still originate from humans. Those who shape human cognition will remain a small elite of thinkers. The intellectuals of the algorithm age will differ from today's science-paradigm thinkers, just as science-paradigm thinkers differed from the clergy of the religious era.
Every new paradigm displaces the old and turns elite knowledge into something mass and democratic. Yet across every era, true intellectuals stay a permanent minority.
1
265
The true value of any innovative project stems from its founder, not the project itself.
The next trillion-dollar company will come from the right founders, and Paul Graham says the idea matters less than the person behind it!
He answers a direct question on where giant companies come from.
His answer in four lines:
1️⃣ The source: formidable founders
2️⃣ Startup ideas are very mutable
3️⃣ Good founders probably work on promising ideas
4️⃣ 20 years of data say founders look the same as before
He expects the next generation of founders to resemble the last two decades’ founders, and he says robots won’t take their place.
Takeaway: back the person, and expect the idea to evolve.
3
214
weweweai and VibeWant are AI projects, not crypto projects. Like every successful AI project, they will generate large, continuously compounding cash flows and strong annualized revenue. Measured by market-cap-to-revenue multiples, their entropy-reduction capacity will far exceed that of crypto projects.
As of mid-November 2026, Anthropic trades at a market-cap-to-revenue multiple (P/S) of roughly 15-18x. Ethereum, measured by protocol-layer burn, sits around 3000x. ETH's multiple is therefore two orders of magnitude higher than Anthropic's. The original tweet had this backwards. Ethereum is dramatically more expensive on this metric, not cheaper. The clear implication is that the benchmark AI projects look like the stronger investment opportunity relative to the benchmark blockchain protocols.
The AI sector follows a clear and logical growth path, very different from crypto. It mirrors the internet's evolution: Web1 led to Web2, then mobile internet, and later O2O. The path is sequential and predictable.
Crypto's overall direction is much harder to forecast. It moves in four-year cycles through discontinuous jumps. The only reliable approach is to wait until a narrative is already live, then build incremental innovations on top of it, just as happened with DeFi and inscriptions. Before April 2020, almost no one could have confidently predicted DeFi's explosion. The best you could do was hold a basket of possible directions that included DeFi. In December 2022, the 2023 inscription boom was equally unpredictable.
This unpredictability has stopped the industry from compounding technological and market progress across cycles. Every new cycle's winning narrative shows up as a surprise, and users stay stuck in the same narrow loop of token speculation. After the 2020-2021 DeFi wave, many expected a DeFi 2.0 supercycle. It never came. As a result, more than a decade of development has failed to produce an industry at internet scale. Total market cap remains capped below $3 trillion.
That pattern defines crypto's past. The only way out is a full reset into Crypto 2.0: large-scale adoption by traditional finance, expansion into non-financial domains, and eventually becoming the foundational ideology of the coming algorithmic paradigm era.
1
257
What mathematicians once took pride in is getting shattered in the AI era.
The article recounts how Francesco Fournier-Facio of Heriot-Watt University shared a striking story about the real impact of this AI flex culture on working mathematicians. A week after one of his talks on sofic groups went up on YouTube, a complete outsider with zero relevant background emailed him claiming they'd used AI to solve the open problems he'd mentioned. As a direct result, he's decided he will no longer discuss any of his current unfinished work in public.
The scientific paradigm has hit an unmistakable tipping point, and an entirely new mainstream human cognitive paradigm is already beginning to emerge.
plus.maths.org/golden-age-ma…
2
297
There's a persistent mental model that in any given space the incumbent giant will simply swallow the smaller startup. History keeps proving this wrong. The same pattern is playing out again in AI right now.
Google was early and dominant, with every structural advantage in AI: capital, talent, distribution, and research depth. Yet OpenAI overtook it in large models. Once OpenAI looked like the new clear leader, Anthropic's growth rate surpassed it.
The assumption that the big will crush the small holds in mature, stable industries (restaurants, traditional manufacturing, etc.). It does not hold at the frontier. On the true frontier, the existing advantages of the incumbent are often the very things that get left behind.
2
3
341
Blockchain projects trade at market cap to revenue multiples that are dramatically lower than AI projects, often by a full order of magnitude. This refers to native blockchain protocols themselves, not stablecoin issuers. Stablecoin projects actually sit in a similar multiple range to AI companies. The key difference is cost structure. Stablecoin operations run extremely lean, while AI projects carry massive compute and infrastructure burn rates and are still deep in the "invest far more than you earn" phase.
Ethereum maps cleanly to Anthropic. Both sit as the clear number two flagship in their respective categories: Ethereum relative to Bitcoin, Anthropic relative to OpenAI. Yet Ethereum's market cap to revenue multiple is far tighter than Anthropic's.
Anthropic is targeting a fully diluted IPO valuation of up to roughly $1.4 trillion against an $81 billion annualized revenue run rate.
Ethereum's current "annual revenue" is far lower. Protocol-level burn and net income currently annualize in the tens of millions to around the $100 million range, well below the peaks of prior years.
According to DefiLlama and similar sources as of mid-November 2026, protocol revenue (primarily base fee plus blob fee burns) ran about $6 to $6.2 million over the last 30 days. That annualizes to roughly $70 to $100 million. Some analyses put the annualized on-chain revenue closer to $120 million.
The Ethereum Foundation is a nonprofit. Its funding comes mainly from early ETH treasury holdings. It covers operations through selective ETH sales, staking yields, and similar sources, with annual spend currently around the $100 million level and explicit plans to lower that ratio over time. The Foundation does not take an automatic cut of network transaction fees. Core developers and client teams are primarily supported through Foundation grants, community funding, and corporate sponsorships rather than any automatic protocol fee flow.
As of mid-November 2026, Anthropic trades at a market-cap-to-revenue multiple (P/S) of roughly 15-18x. Ethereum, measured by protocol-layer burn, sits around 3000x. ETH's multiple is therefore two orders of magnitude higher than Anthropic's. The original tweet had this backwards. Ethereum is dramatically more expensive on this metric, not cheaper. The clear implication is that the benchmark AI projects look like the stronger investment opportunity relative to the benchmark blockchain protocols.
91
As of mid-November 2026, Anthropic trades at a market-cap-to-revenue multiple (P/S) of roughly 15-18x. Ethereum, measured by protocol-layer burn, sits around 3000x. ETH's multiple is therefore two orders of magnitude higher than Anthropic's. The original tweet had this backwards. Ethereum is dramatically more expensive on this metric, not cheaper. The clear implication is that the benchmark AI projects look like the stronger investment opportunity relative to the benchmark blockchain protocols.
Blockchain projects trade at market cap to revenue multiples that are dramatically lower than AI projects, often by a full order of magnitude. This refers to native blockchain protocols themselves, not stablecoin issuers. Stablecoin projects actually sit in a similar multiple range to AI companies. The key difference is cost structure. Stablecoin operations run extremely lean, while AI projects carry massive compute and infrastructure burn rates and are still deep in the "invest far more than you earn" phase.
Ethereum maps cleanly to Anthropic. Both sit as the clear number two flagship in their respective categories: Ethereum relative to Bitcoin, Anthropic relative to OpenAI. Yet Ethereum's market cap to revenue multiple is far tighter than Anthropic's.
Anthropic is targeting a fully diluted IPO valuation of up to roughly $1.4 trillion against an $81 billion annualized revenue run rate.
Ethereum's current "annual revenue" is far lower. Protocol-level burn and net income currently annualize in the tens of millions to around the $100 million range, well below the peaks of prior years.
According to DefiLlama and similar sources as of mid-November 2026, protocol revenue (primarily base fee plus blob fee burns) ran about $6 to $6.2 million over the last 30 days. That annualizes to roughly $70 to $100 million. Some analyses put the annualized on-chain revenue closer to $120 million.
The Ethereum Foundation is a nonprofit. Its funding comes mainly from early ETH treasury holdings. It covers operations through selective ETH sales, staking yields, and similar sources, with annual spend currently around the $100 million level and explicit plans to lower that ratio over time. The Foundation does not take an automatic cut of network transaction fees. Core developers and client teams are primarily supported through Foundation grants, community funding, and corporate sponsorships rather than any automatic protocol fee flow.
The New York Stock Exchange has selected Avalanche as its blockchain platform. Attention is now on Nasdaq and the network it might go with. Solana is one potential option, or Nasdaq could build a proprietary chain the way Robinhood did.
BREAKING: Upcoming NYSE-affiliated tokenized securities platform’s core clearing and settlement layer will be built on top of Avalanche technology.
Avalanche will support the platform’s clearing firm, supporting the development of infrastructure for 24/7 trading, clearing, and settlement.
A landmark step toward the next generation of global financial markets.
Projects that once experienced explosive growth from large crypto-native bases often later discover those same users constrain mainstream adoption. The success that defined them can turn the existing community into dead weight.
In the early to mid phase of Crypto 1.0, while the adoption curve was still climbing, the first wave of crypto users delivered real leverage. That base created the distribution and usage foundation for the next surge. Ethereum's 2016-2017 ICO boom is a clear example. It laid the groundwork for the DeFi explosion of 2020-2021.
By the late stage of Crypto 1.0, once the curve peaked and started heading into prolonged decline, the dynamic reverses. Overall crypto-native user growth stalls. Projects that already captured large crypto audiences stop adding net new users. Remaining growth accrues only to projects that previously had almost no users, and it comes with an extremely low ceiling. That matches the pattern with memecoins from 2024 onward: sequential short-lived hype cycles and almost no durable expansion.
The real risk surfaces as we enter Crypto 2.0. Projects that scaled inside the crypto bubble find it hard to escape. Their large crypto-native bases actively deter mainstream users from joining, blocking the broad-based growth the new cycle demands.
The projects most likely to break out in Crypto 2.0 are those that never caught fire in Crypto 1.0 and still have tiny user counts. Think of the asteroid impact 65 million years ago. The dinosaurs vanished. The small, unremarkable mammals that stayed under the radar inherited the planet.
3
274
Many blockchain communities maintain a strong commitment to "true decentralization." This makes rapid absorption by existing power structures difficult. AI is highly centralized (dominated by a few large companies) and therefore easier for capital and governments to accelerate.
Blockchain's distinctive feature is that its core design resists control by any single center. This makes rapid co-option by a powerful state or corporation difficult. If it is fully absorbed (for example into permissioned consortium chains or CBDCs), it risks losing its original character and becoming a centralized system wearing a blockchain costume.
Its diffusion path is therefore slower and more winding, more dependent on the maturation of broader conditions: deepening trust crises, AI-agent economies that need micropayments and verifiable execution, large-scale real-world asset (RWA) tokenization, or geopolitical demand for neutral infrastructure.











