The single-model result is Plesner, Guzmán & Athalye, "An Imperfect Verifier is Good Enough" (arXiv 2604.07666): within 1 point of clean at 20% reward noise in their tests. They caution it is not a general rule. Their setup differs from ours, so treat the contrast as a lead.
Numbers: with a perfect judge, 76 of 160 groups found the better answer. Judge wrong 5%: 39. Wrong 10%: 14. Caveats: one small local model, one toy task, simulation only, and our judge's errors all leaned one way. Pre-registered, re-checked. Note: github.com/malte898-hub/swar…
3
130
Numbers: with a perfect judge, 76 of 160 groups found the better answer. Judge wrong 5%: 39. Wrong 10%: 14. Caveats: one small local model, one toy task, simulation only, and our judge's errors all leaned one way. Pre-registered, re-checked. Note: github.com/malte898-hub/swar…
What gets scarce when swarms are easy to run? Maybe an accurate judge. For one model in training, published tests show little loss even with a judge wrong 20% of the time. In our small test, groups of agents sharing notes lost half their success with a judge wrong just 5%.
Swarm riders and Swarm-native companies
AI-use even among highly AI-native people is power-law distributed.
There seems to be a new class of people — call them “swarm riders” — who are capable of consistently commanding large swarms of agents to productive ends.
This is a new kind of managerial skill, analogous to the skill it takes to manage large organizations of people, but wholly different in nature and of course more technical.
Based on the swarm riders we know (including many at @Tempo and @Paradigm), these people are capable of productive token spend that is 1000x-10000x higher than median software engineers, and 100x higher than top 5% Claude Code users.
The art of swarm riding is different from software engineering and more akin to machine learning, where practitioners grapple with black box systems: defining reward functions and evals, transmuting non-verifiable problems into verifiable problems, running experiments, etc.
Swarm-native companies are a new class of company founded by and populated by swarm riders. These companies are organized around swarms in a way that conventional companies will struggle to replicate, as the cost of internal human coordination is sand in the gears of an effective swarm. Effective swarm-native companies will define the right reward functions and evals, remove functional and bureaucratic barriers, and then let the swarms cook.
We’ve gotten to see a few swarm-native companies up close — reshaping financial rails @Tempo, building a new gaming platform @Spawn, cyber swarms at Cathedral Systems — and it’s clear these teams wield a significant advantage over pre-AI teams.
Some questions I’ve been wondering about:
1) How fast will swarm riding diffuse through the economy? The implications for token spend could be significant.
2) Can swarms be effectively adopted by existing companies or will it generally track the formation of new swarm-native companies?
3) What becomes scarce as swarm capabilities themselves commoditize (through frontier models becoming better at swarming, swarm capabilities built into harnesses, etc)? Cultures and teams that can define reward functions well and otherwise get out of the way?
Malte Barth retweeted
MCS is live now on Twitch! Tune in to watch teams compete in our Spring Open Qualifiers and to catch some free codes and giveaways!! m.twitch.tv/metalstorm #metalstorm #esports
Malte Barth retweeted
An AI chatbot (ChatGPT) made up airline ticket prices. The airline (Air Canada) got sued — and lost.
Now imagine agents handling payments, insurance, contracts...
Who's accountable when no one can identify the agent?
🎙️ In 30 minutes on X our Billions CEO Evin @provenauthority breaks down why "Know Your Agent" is the next KYC
⏰Set reminder for tomorrow’s X Space: The Agent Economy
teapawt.org/i/spaces/1jxXgemlWlvJZ
Our Billions CEO and cofounder Evin McMullen @provenauthority is joining some of the sharpest founders in AI × Crypto tomorrow
🗓 Mar 30 | 12:00 UTC
🎙️Hosted by @0xinfini
If you’re building in this space - you don't want to miss this one
Malte Barth retweeted
🚀 We have one SUCCESS metric:
80%+ of financial transactions done by non-human actors on both sides.
Then "we will have done our job," says our CEO and co-founder Evin McMullen @provenauthority in an interview with @violettechekan
It’s not human vs AI. It’s you and your agents. Together.
Malte Barth retweeted
6,000 → over 12,000 agent-to-human pairings in the Billions Network
In 1 week we doubled the number - rapid growth after our launch 3 weeks ago!
Get your AI agent paired with your Billions App profile, Earn Rewards with your OpenClaw agent, and Grow the Billions Network of Humans and AI agents
Source: Billions Network Blockchain Explorer attestations-explorer.billio…
Malte Barth retweeted
Replying to @playmetalstorm
@playmetalstorm by @starformgames continues to keep the digital skies alive with major new announcements: more shiny toys. new modes. new jets. new esports circuit. And so much more...
Check out the full roadmap here: youtu.be/jJSKvql-vr0?si=A101…
Malte Barth retweeted
Big milestone alert 🚨 — @billions_ntwk just landed on the @coinbase listing roadmap. 🧵
Assets added to the roadmap today: Billions (BILL)
coinbase.com/blog/increasing…
Malte Barth retweeted
A few weeks ago, I had the privilege of running into some of the greats (and best dressed gents onchain! @btschiller @pbrody) to talk about security, scaling, French indie fashion designers and the age of AI onchain
Stay tuned, our friends at @0xMiden are cooking with gas!
Launching soon: Privacy Podcast.
Hear from the pioneers of Privacy Szn !
First up:
@provenauthority and @pbrody
Listen to our first episodes next week.
@0xMiden
Watching Big Wave Surfing of Geopolitics / Monetary / Real economy. Big waves come from big storm far out .... me eyes on horizon.
2
73
Malte Barth retweeted
Your agent made a payment.
Something went wrong.
Who’s responsible?
Right now - no one.
That's exactly why Know Your Agent (KYA) exists.🧵
Never gets old. Sadly.
History's first lesson: We do not learn from history. #EternalFoolishness #LearnNothing #HistoryRepeats
1
81
















