We just got back from ETH Denver. Attendance was down to around 8,000 from the 25,000 peak. Sure, Crypto Twitter was complaining about prices, but the people who showed up weren't thinking about charts. They were focused on building. The conversations were sharper, the signal was higher, and the energy reminded us of the early days when nobody was there for the hype.
And then while the conference was still going, Venice.ai's token ripped over 150% in a month on surging usage, processing over 45 billion LLM tokens daily on decentralized, private infrastructure. And just last week, OpenClaw, the fastest-growing open-source AI agent platform whose creator was recently acqui-hired by OpenAI, named Venice as its recommended model provider. Anyone could have bought it. That's the kind of asymmetric upside that only exists in this space. You don't need to wait six months for an $80,000 Blackwell GPU setup to participate in the AI revolution. Blockchain gives you access now, at a fraction of the cost, with no gatekeepers. Venice is one example among many of builders leveraging decentralized compute to compete with the biggest companies in tech.
This month's edition is about what's breaking on the internet (bots are eating the web), what can't be replaced no matter how hard you try (communities), and what history teaches us about the people who show up when everyone else has left.
Convergence is brought to you by 321 Converge Inc. (“321”), a new 501c6 created by the founders of The Medici Network.
Integrations We Want To See 🔮
Now that we own our data, here are some things we’ll do with it
Sparks ✨
Quick, curated Web2 and Web3 insights you need to know
#1 Communities Are Not Fungible
Author: Joan Westenberg (Writer & Technologist, Founder of Signalvs & The Index)
TLDR:
Silicon Valley treats communities as interchangeable assets that can be migrated, replicated, or manufactured on demand. But communities are path-dependent. Their value lives in the specific, unreproducible web of relationships between people, not in the people as discrete units
When platforms die, communities don’t migrate. They fragment. LiveJournal’s users scattered across Dreamwidth and Twitter. Vine’s creators split between Snapchat and TikTok. Twitter’s degradation scattered users to Threads, Bluesky, and Mastodon. No successor ever captures the culture
Each member experiences the community differently. Lose members and you lose entire subjective communities that existed nowhere else. The ambient trust that holds it all together is always the first thing to go
Our Takeaway:
This is the strongest philosophical case for why on-chain social graphs matter. Portable, user-owned social graphs on-chain ensure the web of relationships survives regardless of which app renders it
When your reputation, history, and trust credentials travel with you, the platform becomes replaceable. The community doesn’t
Every platform migration is a forced reset of social capital. Web3 makes that social capital something you own, not something a platform lends you and takes back when it pivots, gets acquired, or dies
#2 You Should Learn from CryptoPunks
Author: Reid Hoffman (Co-Founder of LinkedIn, Co-Founder of Inflection AI)
TLDR:
Most founders get distribution backwards. They build the product first, then scramble to find users. CryptoPunks proved that community-first distribution is a viable and powerful alternative
CryptoPunks didn’t launch with a polished product roadmap. They launched with a shared identity and a culture, and the community itself became the product’s most durable moat
This is a lesson for all founders that the community-first playbook that emerged from web3 has broad applicability across all of tech
Our Takeaway:
When a Web2 titan like Reid Hoffman tells founders to study a Web3-native project, pay attention. This is exactly the kind of signal that precedes mainstream adoption of a playbook
CryptoPunks worked because ownership was real and provided a strong economic alignment that made the community self-sustaining
The next generation of consumer products, across gaming, media, and social, will adopt this hybrid model: Web2 UX with Web3 ownership underneath. Hoffman saying it out loud is a green light for founders who’ve been sitting on the fence
#3 2026 of Crypto = 2002 of Internet, The Darkest Hour before the Dawn
Author: DeFi Cheetah (Investor at Velocity Capital, formerly Binance Research)
TLDR:
High-conviction blockchain investors like Kyle Samani (Multicoin Capital) are leaving for AI, longevity, and robotics. The idealism of 2021 has been replaced by corporate silence, and a crisis of faith is spreading across the industry
Institutions are finally entering (ETFs approved, banks piloting subnets, asset managers tokenizing treasuries) but they’re stripping out the permissionless properties that made blockchain revolutionary. This is the gentrification of blockchain
The parallel to 2002 is striking: after the dot-com bust, the narrative was identical. The internet was useful only for email and buying books. Yet Google, Facebook, and AWS all emerged after the crash
Our Takeaway:
For Web2 decision-makers who think they missed the blockchain window or that the interesting parts are over: that’s exactly what people said about the internet in 2002. The real value creation hadn’t even started. The builders who stuck around during the bust built everything you use today
The gentrification warning is the most important signal. If institutions tokenize assets but lock them inside permissioned wrappers, we've just recreated the old financial system with a new database. The real unlock is permissionless composability where any developer can build new applications on top of existing infrastructure without asking for access or recreating it from scratch
AI sucking capital out of blockchain might actually be the best thing that could happen. It clears out the tourists and leaves the builders. The convergence of AI and blockchain (agents that need payment rails, data markets that need verifiable provenance, AI models that need decentralized infrastructure) will be the next wave
The 3-2-1 🔍
Featured insight breaking down a major story or trend that matters
AI Bots Are Now a Significant Source of Web Traffic
Author: Will Knight (Senior AI Writer at Wired, formerly Senior Editor at MIT Technology Review), with data from Fastly Research, Akamai, and TollBit
TLDR:
Something is quietly breaking on the internet, and most businesses haven’t noticed yet. According to new research from Wired, Fastly, and Akamai, AI bots now account for 1 in every 50 website visits, up from 1 in 200 just three months earlier. AI crawlers make up close to 80% of all AI bot traffic observed. And it’s not evenly distributed. Meta’s AI bots alone generate 52% of all AI crawler traffic, more than double Google (23%) and OpenAI (20%) combined. On the real-time fetching side, ChatGPT drives 98% of fetcher bot requests.
These bots aren’t politely knocking on the front door. Over 13% of bot requests now bypass robots.txt entirely, the basic file that websites use to tell bots what they’re allowed to access. Some bots disguise themselves as normal browsers. Others mimic human browsing behavior so convincingly that they are now nearly indistinguishable from real users. In one case, a fetcher bot hit a single website at 39,000 requests per minute at peak load, creating what researchers described as DDoS-like effects without any malicious intent. The bots aren’t attacking. They’re just hungry.
This is creating an arms race between websites trying to protect their content and AI companies trying to scrape it. Publishers, e-commerce platforms, and media companies are all scrambling to figure out who is visiting and why. The existing toolkit (robots.txt, rate limiting, CAPTCHAs) was designed for an earlier era of simple search crawlers, not for sophisticated AI agents that can impersonate humans at scale. As TollBit’s CEO put it, this isn’t just a copyright problem. There is a new visitor emerging on the internet, and nobody has figured out how to deal with it.
Our Takeaway:
The internet’s trust model is cracking, and the consequences go far beyond copyright disputes. When you can’t tell bots from humans, every metric that runs the digital economy breaks. Ad spend is distorted because impressions aren’t real. Analytics are polluted because traffic isn’t organic. Conversion rates are unreliable because the funnel is full of machines. Content licensing is impossible to enforce because you can’t track who consumed what. The entire measurement layer of Web2, the thing that underpins digital advertising, SaaS pricing, and media monetization, is being undermined by visitors it was never designed to authenticate.
The Web2 answer to this problem is predictably inadequate. Robots.txt is a suggestion, not a wall. Rate limiting is a blunt tool that punishes legitimate high-traffic users alongside bots. CAPTCHAs degrade user experience and are increasingly solvable by AI anyway. Even Cloudflare, which protects nearly 20% of all websites, has had to continuously evolve its bot management tools as AI scrapers find new ways around detection. These are band-aids on a structural problem. The real fix requires something the current web doesn’t have: cryptographic proof of personhood and verifiable identity at the protocol level. When every request carries a credential that proves whether the visitor is human, an authorized agent, or an unauthorized scraper, the arms race ends. Not because bots are banned, but because the rules of engagement are finally enforceable.
But verification is only half the story. The bigger opportunity is economic. If bots are the new visitors, they need a new way to pay for access. Right now, AI companies scrape content for free, train models on it, and capture all the value. Reddit figured this out the hard way and started charging for API access after years of its data being used to train models for free — no more free lunch. That’s the Web2 playbook: put up a toll booth. But it’s clunky, bilateral, and doesn’t scale. Stablecoin micropayments at the infrastructure layer could replace the arms race with a marketplace, one where agents pay for what they consume and content creators get compensated in real time. Imagine a world where an AI agent accessing a news article, a product database, or a research paper automatically triggers a payment of fractions of a cent, settled instantly on-chain. The economics of the agentic web don’t have to be adversarial. They can be transactional, transparent, and fair.
The biggest unlock here isn’t just defending against bots. It’s building an internet where every interaction, human or machine, carries a verifiable credential and settles value programmatically. That’s the foundation for agent commerce, data licensing, and a web that works for everyone instead of just whoever scrapes fastest. The companies building this infrastructure layer (verifiable identity, on-chain micropayments, agent authentication) aren’t solving a niche blockchain problem. They’re solving the next internet’s core design flaw before it becomes permanent.
Power Quote:
There is a new visitor emerging on the internet, and nobody has figured out how to deal with it.
What We’re Reading or Downloading 🎧
Podcasts, interviews, discussions, and research reports that are influencing our thinking
Tweet of the Month 🐦
Key takeaways and perspectives from X
Nikita Bier (Head of Product at X and Venture Partner at Lightspeed Venture Partners)
Meme of the Month 📸
Send us your memes for a chance to be featured: hello@321converge.com
See you next month,
Nick Tang
X: 0xtangelo
Kris Jenk
X: 0xjinkys
Follow 321 on X @321Converge
Suggest an article or blog post for next month’s Convergence here.
