Mustafa Al Bassam

Co-founder @Celestia

[Donncha Ó Cearbhaill @DonnchaC]

1/ Want to investigate and disrupt the world’s most sophisticated spyware surveillance threats? Our team at Amnesty International’s Security Lab is hiring for a passionate Technologist to help expand our team. Applications close Sunday 15 Nov 🧵👇

[Mustafa Al Bassam @musalbas]

@0x_Osprey What kind of setup is this?

[Mustafa Al Bassam @musalbas]

@WillPapper @ReadwiseReader How does it compare with Grok Bot? I was blown away by ChatGPT Dot, but Grok Bot was when I realised that there's been a step change in AI capabilities over the past few months.

[Mustafa Al Bassam @musalbas]

Being a swarm rider is quietly the most valuable skill at any organisation right now.

[Matt Huang @matthuang]

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?

[Celestials @celestials_id]

We landed. Celestial Ids now live on Eden, an EVM chain with @Celestia Underneath.

https://t.co/brmYvulOC7 runs on Eden from today. Every Id moved over with its owner, so your name is waiting at the same wallet address.

[Celestia @celestia]

Celestia mainnet turned 3 last week, and last year was our biggest yet:

• Network usage grew 2.6x
• Throughput capacity grew 115x
• 3 new products were released including Fibre and Voyager

Here’s a full recap👇

[Mustafa Al Bassam @musalbas]

@toly would be in the monolithic SHU tho

[Celenium @celenium_io]

Three years of Celestia on our indexer:

15.5M blocks
79.8M transactions
4.27 TB of blobs
1,098 namespaces

Agents are welcome to push those numbers up. We’ll keep counting 💜

[Angry Crypto Show @angrycryptoshow]

UPDATE: Charles Hoskinson says $NIGHT will "use Celestia as a data availability layer and Filecoin for archival storage. We have a great relationship with Celestia and Filecoin. There's a lot of data overhead—Nightframes for agents, private payments, the DeFi layer for privacy."

[Mustafa Al Bassam @musalbas]

3 years ago today, Celestia mainnet went live.

Still surreal to see LazyLedger go from a wild idea to a living network. We've had highs and lows, but I believe in the mission more than ever:

Abundant blockspace, fast enough for machines, governable enough for humans.

Agents may out-transact us sooner than anyone expects. When they do, humans still need to write the rules. Social consensus is the final layer.

That's why we built Fibre, benchmarked at 3 Tb/s end-to-end. And it's why we're launching Voyager 👇

To everyone who's been part of this: thank you. Onwards 💜

[Celestia @celestia]

The internet was built for humans, but AI agents are changing internet traffic patterns and challenging two of the internet’s core business models: advertising and subscriptions.

A pay-per-use model suits the agentic internet, but existing payment rails can’t support it.

[Mustafa Al Bassam @musalbas]

@colludingnode is it?

[Mustafa Al Bassam @musalbas]

Okay so this year is the year that Linux on the desktop actually won, given that all the AI cloud desktop environments (ChatGPT Dot/Grokbot/etc) are Linux-based, and that's how most heavy computer-based tasks are going to be performed...

[Mustafa Al Bassam @musalbas]

Asked Claude to make a video about this (150 BPM)

[Celestia @celestia]

Introducing Voyager: hyperscale payments for agentic commerce.

Benchmarked at 21M TPS, Voyager is a new blockchain architecture for an internet where millions of agents pay for content, data, and services on demand.

https://t.co/y0r03Fd4eZ

[Nick White @nickwh8te]

I'm thrilled to announce Voyager, a blockchain design for hyperscale agentic payments clocking in at 21M TPS.

3 things worth highlighting:

1. Celestia Labs is one of the best positioned teams to capture the agentic payments opportunity. We have been building towards millions of TPS for over 5 years, when many thought such a feat would be impossible and never needed.

In time, even 21M TPS could look small so we're prepared to scale Voyager far beyond this. Fibre can handle up to ~2B TPS of DA, now it's time to find out what the upper limits on execution are.

2. Voyager marks a significant change in Celestia Labs' capability. We are no longer just providing raw blockspace, we are building the tools to turn that blockspace into useful execution to solve real problems.

When we acquired Sovereign Labs in July, the goal was to do exactly this. Build full stack solutions that encompass DA, execution and beyond. Voyager is just the first product. More to come.

3. Celestia will do more for AI that just power agentic payments. Blockchains provide the ideal substrate for agents to coordinate in both financial and non-financial ways. We also see blockchains as a critical tool for humanity to establish safeguards and maintain governance over agents in the future.

But in order for any of this to work, you need a blockchain that can keep up with the speed and scale of agentic activity. We believe Celestia has a role to play here and we're working towards this.

---
Major props to the engineering team for reaching yet another major benchmark milestone after hitting 3Tb/s on Fibre just a few weeks ago. Incredible work.

Stay tuned for more details on the technical architecture and our benchmarking setup.

[Celestia @celestia]

Introducing Voyager: hyperscale payments for agentic commerce.

Benchmarked at 21M TPS, Voyager is a new blockchain architecture for an internet where millions of agents pay for content, data, and services on demand.

https://t.co/y0r03Fd4eZ

[Derive.xyz @DeriveXYZ]

Derive V3 started with one question: how do you scale an onchain derivatives exchange without moving settlement away from Ethereum?

Our answer was to build an application-specific rollup for derivatives, with Celestia for data availability, Succinct for proofs and Ethereum for settlement.

Why not just put everything on Ethereum?
Derive produces a lot of state, and that grows quickly with activity. Posting all of the underlying data directly to Ethereum would make scaling the exchange much more expensive.

V3 posts the data behind state changes to @celestia, while the protocol continues to settle on Ethereum.

Why does that data need to exist outside Derive?
Because V3 is designed so users can still verify their state and withdraw without depending on Derive’s API or sequencer. If the data only existed in our own database, that guarantee would be much weaker.

Celestia gives the protocol somewhere independent to publish the data needed to reconstruct and verify state, while Succinct proofs verify the updates that are ultimately committed to Ethereum.

Where does privacy fit?
We've been working with the Celestia ecosystem on a design where account data will eventually be encrypted while still giving users what they need to prove their state themselves.

[Celestia @celestia]

Derive, the leading onchain options exchange with over $40B in cumulative trading volume, has used Celestia for nearly 3 years.

With V3 now live, we spoke to the Derive team about what changed in the upgrade and why they continue to build on Celestia.

Here's what we learned👇

[Mustafa Al Bassam @musalbas]

@alexanderlhicks @sjmurdoch I think if there's aggregate statistics/info on how accurate that belief has been historically, so that the public can at least evaluate how harmful or not the program has been to civilian infrastructure, it would be a good balance between transparency and secrecy

[Mustafa Al Bassam @musalbas]

The NCSC tells us to patch security vulnerabilities. Its parent agency, GCHQ, exploits them to hack targets.

When GCHQ discovers a 0-day, does it disclose it to protect the public, or keep it secret to exploit?

Isn't this a massive conflict of interest?

[NCSC UK @NCSC]

The NCSC is encouraging UK organisations to prepare for immediate updating of systems using PHP. Where immediate remediation is not feasible, organisations should consider taking the system offline until it can be fully mitigated.

For more⬇️
https://t.co/l6TDtVurJN

[Mustafa Al Bassam @musalbas]

@AlexMartin @sjmurdoch Aggregate statistics would be a good start. And perhaps have statistics on how many retained CVE was exploited by non-GCHQ actors before they became public, and if they were eventually made public by GCHQ or someone else?

[Mustafa Al Bassam @musalbas]

@sjmurdoch Thanks. I think we should at least get more transparency / statistics on this to help evaluate the effectiveness of the equities process https://t.co/8682zS1dex

[Mustafa Al Bassam @musalbas]

Thanks. I'd love to know which vulnerabilities were previously retained, that later become public and/or exploited by others, so we can assess whether they actually protected the public. Let's see if the IPCO respond to this lol. Maybe you should send as they might be more likely to reply to a journo.