Buzz

What this episode covers

Block launched Buzz on 21 July 2026 — an open-source, Nostr-based group chat platform built, in Dorsey's own words, "for teams of people and agents of all sizes." Robin and Tobi didn't just review it — they moved their own business onto it, and this episode is the honest account of what that took. Source: Block's launch announcement, 21 July 2026.

The setup pain is real and specific: keys, environment variables, access control — the unglamorous plumbing that comes before any of the upside shows up. Once it's running, auto-transcribing every voice note changes how a team actually talks to each other, and the hosts get into how agents end up spreading bottom-up inside larger companies, one team at a time, well before any formal rollout.

They don't skip the hard part either: the prompt-injection risk that nobody in this space has fully solved yet. Robin's answer for why he stays on Buzz anyway comes down to one thing — sovereignty over his own data and AI infrastructure, even against easier, more polished closed alternatives.

One correction worth noting here since the show is committed to getting numbers right: an on-air stat about companies listing AI agents on their org charts was corrected after broadcast — the accurate figure is 23%, roughly one in four, not the 25% said on air.

Read the full article →

Transcript

Cold open [00:00] — Host

Buzz is actually the output of just him working. There's not a SaaS product out there that works like that. It's an invention. And now that I've got one or a couple set up, it's like a bag of Maltesers. Like, I need all the agents. And I want agents with different roles, and I want a different agent for every aspect of my life, and I want a different agent for every aspect of my business. This tool is just an early type of tool. What I would expect is Slack should follow up with similar capabilities so that they stay on track. Fast forward — probably won't remember a world where you didn't have that super intelligence just working for you, that knows you and is actioning things for you, right? Coders with these tools and ADHD are unstoppable. They are like the ones who can have like three different things going on at the same time.

Welcome [00:58] — Tobi & Robin

Tobi: All right, we're back. Hey Robin, how you going?

Robin: Hi. Good day, Toby. How you going?

Tobi: Yeah, good. Can't complain.

Host: I'm pretty pumped about this one. So, what are we going to talk about today?

Host: Buzz. Today's about Buzz. It's all about the Buzz. Jack Dorsey.

The framing [01:13] — Host

Host: Can I read a quick framing? I've got some notes today. Block launched Buzz on 21st of July, 2026. It is an open-source group chat platform where humans and AI agents work in the same channels. Built to replace Slack and GitHub in one app. Jack Dorsey's pitch: model agnostic, decentralised, self-sovereign, and open source.

Host: What the heck does that all mean?

Host: Yeah. What do you reckon? Dumb it down for us.

Host: It's a place where agents and humans can chat and work together — and hopefully seamlessly. It's all new. That's what we're testing out. But it's a place where agents and humans can work or chat together and get things done in a more effective way.

Why WhatsApp fails agents [02:13] — Host

Host: We're seeing it ourselves in that we're using agents and WhatsApp in multiplayer mode at the moment. But what we're seeing is like having stupid kids in a group — they just don't pay attention. They don't know what's going on. They don't talk to each other back and forth naturally. You've got to tag them, and if they are set to always reply then it becomes super annoying. So it's not very intuitive. You can train them to be more intuitive, but they just aren't, because they're treated as second-class citizens in channels like WhatsApp, Telegram, even Slack. They don't get rights and kind of respect as a human individual.

A persistent team, not spun-up sub-agents [02:58] — Host (Tobi airtight on the Alara block)

Host: So each agent gets its own account, cryptographic identity, keys, and channel memberships and permissions. I mean, we can see where this is going, right? This is just another step to making these agents more autonomous, where they're linked to crypto, linked to payment systems, and they can actually action stuff.

Tobi: I just want to flow into probably the part that I'm most excited about. What we do is there's a lot of Claude Code fixing our Hermes agents, and it does it really well. And now I can massively see the benefit of having your Claude Code set up, your Hermes agent. You don't have to have one. It's not about one — it's about what is your right combination. But what I'm so [excited] about is I'll go into Claude Code and I'll say fix, you know, Alara, my Hermes agent. That will happen, and then I'd be copying and pasting, relaying that back to Alara. But now it's all in one channel. And Claude Code's in, right? Alara's my hesitation away. And then we might have another agent in something else. And they're all talking to…

Host: It just allows the structure of a team of agents, as opposed to with a Hermes or an OpenClaw — you can have a structure of agents. So you have a primary agent and multiple sub-agents, but they're not really persistent. They often get spun up at the time that you request the work to be done, or request the team to be spun up. So they don't really sit around. You can try and train the sub-agents with roles, but it's very hard to keep them persistent. And the only answer is to make another major agent and have multiple agents that are all running.

With Buzz, it allows you to bring in any agents you create — like your Hermes, your OpenClaw. It could be a Vertex, it could be an AWS, it could be a whatever harness you've set up. So it supports any agent, but it also has agents as well. And this is the thing that got me really pumped on it: you go and you bring your agent and you're like, "Great, I've got my general-purpose agent that has access to everything. I want to set up a channel like I would set up in Slack, and then have another agent which is a product manager, and then maybe another agent which is a tester, maybe another one that's a researcher." And they are persistently in that channel, ready to support. And that way you can actually have a full team working, and monitor their work and what they're talking to each other about.

Host: So it allows you to visualise those teams of agents.

The install story [06:14] — Robin

Host: So you set it up — how did it go?

Robin: Uh, it was a total nightmare, to be honest, and this is the barrier to entry, I think, with any organisation that's going through it. This isn't really like a ready-to-use SaaS product. It's more like a piece of code that you install on your server and then you run on your own server. So I'm running mine in Railway, which is a virtual server — same place I'm running Hermes. So it was as hard as setting up an OpenClaw or Hermes, maybe a little bit harder, because I then had to bring in my Hermes agent. And yeah, it's not like "just invite a friend, here's a link" — you've got to give each other keys and set up variables to give each other access. So it was a real nightmare. And even now I haven't got the mobile app version working yet — it just errors out. There's still that barrier to entry, which is kind of cool, 'cause you feel like you're setting up something that — my mum can't go and set up Buzz today.

Developers right now [07:13] — Host

Host: Yeah, 100%, it's great for developers right now, because they can do all of these things and they can also see the future. For them this really helps the development lifecycle process dramatically, because you can do everything within there and then just push. And if you wanted to use Claude Code to do your pushes and still use GitHub, you totally can — it's just part of the process, but you might go through the entire lifecycle from Buzz. So it really allows you to drive development teams — but the teams of agents. And the future is I could do the same with a business structure. I could have a CEO agent and a CMO agent — whatever agent, any role in the business I could structure into Buzz and actually have them run my business.

Who it's for [07:57] — Host

Host: So who do you think it's right for now? Is this right for solopreneurs? Is this right for SMBs, or enterprise, mid-market? Where does this sit? Who's going to use it?

Host: I wouldn't say it's right for mid-market [or] enterprise at all, just because of the security risk. There's a massive security risk in giving these agents cryptographic identities and allowing them to act independently — they can inject code. Say I've got my Hermes agent, which has a huge amount of access into my personal stuff — for a large business it's like business-closure risk if this kind of thing happens. All of these platforms will have to either adapt, or there'll be a new range — maybe these are like the MySpaces of these collaborative engagements. I was even chatting to someone in Thailand yesterday and they brought me into a chat group and there was a translated bot, and it just works in the same way. It's like the three of us collaborating together and that bot just had one function, which was translate every message back and forth.

"No bots in the thread" [09:13] — Host

Host: It's interesting, because we'll be finding our way. Like in my basketball group the other day, they were saying, "Uh, no bots in the thread." So there's going to be all these social rules that are going to start to emerge of where our new friends are wanted and allowed.

Host: I've had that same thing with my personal friends. We do a lot of voice messages back and forth. If we bring Stephanie, my agent, into our chat group, we can have them transcribe those messages, 'cause people often complain, "Oh, your voice message was too long. Too many things covered. I forgot what you said." So why don't you just auto-transcribe everything? Then it's just easier for everyone. The issue then comes back to, "What's happening with my data?" And for me, I'm storing it into a world model. Everything about everybody is stored in a fact table with information about them. So they've got to trust me that I'm storing the data correctly. I've had one friend that was like, "No, I'd prefer just not to — I want to be able to talk about personal things. I don't want anyone to have access to this ever." So, okay, that's fine. Other friends really love it. We've got another group where Stephanie remembers birthdays and upcoming travel dates, so she will remind us, "Oh, it's Jess's birthday next month," or, "Hey, remember the Thailand trip you've got planned."

Host: Yeah, and I can totally see both points of view. I can see people that want to retain their privacy or just have human chats. But my goodness — when you're in a chat, let's say we organise a motorbike ride, and then we just say to one of our agents, "Hey Alara, can you just organise this motorbike ride for August and find us the best trail and how many days or whatever it might be?" And it does it.

Dorsey, Block and Afterpay [11:04] — Host

Host: Dorsey announced it on X as launching Buzz to reduce our dependency on Slack and GitHub. Block is dogfooding it internally. Block is Dorsey's business — they bought Afterpay, an Australian company, "29 bill". And he built a very nice house in Bondi — not Jack Dorsey, the Aussie guy. They already have fintech, but now he's got full open source with this. Such an interesting play, this open-source world versus all these platforms that exist.

Solopreneurs first [11:38] — Host

Host: I think this is the product of his way of working internally. When you look at that world model — we'll have to put an image of it up on the video — the "Jack Dorsey world model company brain" concept, that's really what Buzz enables. And you and I have been talking about that world model view for a while. He's like, "I've created this better way of working. There's not a SaaS product out there that works like that. It's an invention."

Going back to your question about who should use it: solopreneurs, 100%. You should just go straight there. And even like you and I — we're having all these issues with WhatsApp and our agents. Once you go to Buzz, it's all solved. Let's not even try to make WhatsApp a channel. Any small development team where it's a technical team and you've got full control over your tech stack — not just buying SaaS — if you're a technical dev team with a startup, I'd go to this straight. This will give you a much more accelerated rate of product delivery, because you're augmenting the delivery team with all of these agents. It has so much power as well. I'm not a developer, so I can't comment on whether it can replace GitHub — but if it can, that's pretty astounding.

Killer or coexist [13:01] — Host

Host: There's so many sensationalist comments — it's a this-killer and a that-killer — but usually it's somewhere in the middle: not completely killing something, but coexisting with it. You've got to be careful with those statements. And here's some little stats on what we were just talking about — market timing stats: "31% of corporate leaders now frame AI as teammates. 25% formally list agents on org charts. Org charts and agentic AI adoption in services jumped from 4.3% to 25% between August and November last year." But I bet these companies creating agents on their org charts, I bet they're not real agents. My hypothesis is they're figurative agents — they have some kind of project in Claude which they've trained to do a thing and act a certain way, and that's what they consider an agent. Which is fine, they're using it like an agent in the team, but it's really just probably a shared project in Claude — not really an individual agent with its own identity that can do its own stuff.

It may be very powerful 'cause it's connected to doing stuff. The myth a lot of people are under now is that Claude is an agent. Claude is a harness — you can structure an agent in Claude Code and make it look like an agent, but using co-work or chat, it's not really an agent — that's just using an LLM. This is different. This is actual individual agents that have their own thing going on.

The maturity curve [14:43] — Host

Host: What I find really interesting is you'll be at a dinner party and someone will say, "What are you doing?" — AI and automation — and they'll say, "Ah, well, we use AI at our business." And you choose your words carefully, 'cause you're like, "Do you?" There's so many levels. Opening up Codex or opening up Claude is using AI technically — but is it efficiently running your life or business, doing all the things you want it to do, really removing the manual grunt work, really driving business value and life value? Probably not. There's a maturity curve with it, like any technology.

Growth engineers [15:31] — Host

Host: And it's really a bottom-up game as well. I've got a lot of mates in sales at Salesforce, for example, and they're all building their own agentic sales workflows — custom-building their own pre-sales engineering. You used the term — what was it? Growth engineer?

Host: Uh yeah, a growth engineer. An integral part of a sales team now, because you're leveraging AI to build out a significant part of the sales pipeline. Hopefully you're just driving to a meeting where a human jumps in and closes the sale. You've got all these individual sales reps using Claude to engineer the sales process automated. But it's different to having an agent provided under the same structure. If a sales team went to Buzz, they would have individual agents that do tasks and hand off tasks — a bit more structured than a team. You could automate the creation of a channel for every prospect and have a team spun up just for that prospect, then work on that deal together. Any workflow where you would have used Slack yesterday, you could now use this tool — potentially, if you have the setup.

Slack does have the ability to tag Claude in it now, so you can go direct to Claude through Slack. But again, it's not like "Jeff, the agent" — it's like I'm just going to my Claude account to get that information and come back.

Early beta [17:07] — Host

Host: And it's so early. It's early beta. It's got a mobile app for Mac and Android.

Laptop vs server [17:14] — Host

Host: This is how fast we're moving — we put that up like yesterday. And while I said the setup is very difficult, it's very difficult if you want to host it yourself. But if you want to run it on your laptop, you can install it with the app, and it's very fast and easy to get into. The problem is if you turn your laptop off, then it goes down — it's a server running while your laptop's on. Obviously you want it to run when you're sleeping, so that's why you'd host it on your own server.

Philosophy and the competitor context [17:44] — Host

Host: So interesting — we're hearing a lot at these frontier model companies about the philosophy the leader has. Jack Dorsey has a very open-source philosophy, and he would have philosophies on AI development. Anthropic, they've got their own internal philosophy. And OpenAI. And the competitor context is that there's "Paradigm Centaur", which takes the opposite approach. I'm reading a note here: "Putting a virtual employee inside Slack rather than rebuilding the workplace around the agent." So you've got people coming at this from different angles. I'm excited about Buzz because I want to jump into that environment.

Host: It's just the agent-first nature of Buzz that makes it easy, because the agents are able to interact.

Host: If Slack supports that, then no drama — it's kind of the same thing. I think Slack doesn't have the GitHub replacement, and it doesn't have cryptographic identities. Those kinds of things are unique to Buzz.

Crypto identities and regulation [18:46] — Host

Host: We were speaking to a guy just the other day around financial regulation for AI use. And "some of the global legislation is saying you have to use cryptographic identities for agents so you can track their work on blockchain, so it's immutable". Which makes total sense, especially with financial transactions. Buzz supports that path forward, whereas other platforms don't have cryptographic identities yet, so it's hard to establish any kind of trust in the platform.

Host: Shout out to Tim — we had a great chat with him about building trust in agents. Agents will have different trust levels, I guess, just like humans do. We have background checks or whatever, which will enable them access into different places. So agents will have different levels of authorisation. And what's that playground going to look like?

Air gaps and the mailroom rule [19:40] — Robin

Robin: Yeah. Like you and I, we have personal agents. I used to set up another independent agent for Axela, and having that kind of separation is very important. Because what we have seen is that even if you train them and spend a lot of time on them, they occasionally slip up and share some information. I think there need to be air gaps in various use cases in our lives. You may have an air-gap agent that just deals with your finances — that has no access to your other agents, and there's no sharing of private information. And maybe you don't have them in chats where they're publicly accessible either. That's where delineation and identity between agents is really important.

Companies have to think about the same thing. If I'm using agents in my Salesforce development lifecycle — which is something that is starting to become very common — I may give them access to a sandbox instance of Salesforce, but I don't give them access to production. It would be a lot easier if they had access to production, 'cause I could see what's going on there, but I don't want them to change production. I have to have this air gap with how they interact with different systems. Same with Buzz. Stephanie, my homey, has got a lot of access, but Micah — who's a project manager in one of my threads — doesn't have any access to do anything. She is really just there to be a product manager. The real risk is if Micah injects a prompt into Stephanie and manipulates her. That is a real threat that I think hasn't been solved yet.

Those kinds of things are the limitation in moving this into enterprise. It's just analogous to a human environment, right? The new kid in the mailroom is not going to have the same access to the company data that the CEO is. There's going to be this structure — essentially, to simplify it, like an org chart — different levels of access to information and access to actioning tasks.

Infinity agents [21:41] — Host

Host: For a while I've been thinking about this approach to infinity agents. I've got a new customer that onboards with me. I want to give them a key point-of-contact agent that's their number-one concierge whenever they need them. But I also want them to have a quality checker on anything that's produced for them. I want them to have an independent, impartial survey agent that's just checking how's everything going. I want to have all these different agents that spin up every time for every customer, but I need a walled garden of data for every customer — I'm not sharing the same customer-success agent with another customer that has access to their data, 'cause it'll confuse that agent. So I need a walled garden for every single customer with an agent team that collaborates around that data, but they should all have shared access to the parent brand that says: this is what we do, this is everything you need to know. But for every customer I want them to be separate. Same for every agent and employee internally — they all need their own separate world model with a shared world model, so they can't interfere with each other's.

I think this is a data architecture structure that organisations will need to start planning for to support the agentic enterprise.

Start with one agent [23:01] — Host

Host: Just to bring it right back and really simplify things: there will be a huge amount of people that, just as a very simple step, start with one agent that has the right amount of risk profile and access to their information and data. If you're listening and you don't have an agent, that's a really good starting point. There's a bit of setup involved, but once you get it going, the benefits of what it can do for you far outweigh.

Ubiquitous agents [23:37] — Host

Host: So it'll become ubiquitous — they'll be on every surface, that agent. And that's what we're seeing now: it is almost every surface. I can be in Slack and access my Stephanie. I can be in WhatsApp, Google Meet. Pretty much anywhere I want to work, I can bring her into it. I think it'll just become natural to have it, and you'll probably talk to it in your ear in the future. It'll just be normal.

Enterprise is one year out [24:00] — Host

Host: Coming back to Buzz and how it applies to businesses, because that's what we are really focused on as organisations adopting this AI technology: why are we talking about Buzz if it's not really for medium-to-large-size enterprises? The reason I like talking about it is because this is the future. We're just seeing the first glimpse into the future of how organisations will interact. The tools just aren't ready yet, but they will be soon. Platforms like Hermes, like OpenClaw, showed us what an agent was like — wow, this is a multiplayer agent, we can have many of them, they can do different jobs, we can give them different access and skills. That's much different to just using Claude or Codex as an LLM. But Buzz has given us this new vision of the future, which tech companies and organisations need to start looking at and going, "Okay, this is something we could actually bring into our company now. Maybe it won't pass our security test. What if we could give our internal IT team access to collaborate on this — would it make them faster and more efficient? And then what if I onboarded the marketing team to use it? Maybe they have a bit of a safer playground to play with." You can start to think about what if we had this.

Then it's just a waiting game on which platform tech provider — like Salesforce or Microsoft — is going to provide the enterprise tech that's the same as this. Because it's happening. I give it one year and we will have this for enterprise. There will just be data security and governance layers built into something like Buzz, and it will be adopted very quickly. We can already build out harnesses for businesses that get agents running and actioning things within your business that have those safety guardrails in place, and it will become like a blend of that and Buzz.

Shared compute [25:58] — Host

Host: There's one thing we didn't talk about before, and that is Buzz gives you the ability to do shared compute. This is pretty cool. Everyone can pool compute for heavy tasks through Buzz. You can give up your machines — say you're not using the full capacity of your machines, you can put them into a pool and other people can access your machines to create and build. That lowers the barriers for small teams running serious agendas.

Decentralised power [26:33] — Host

Host: It just takes the concept of decentralisation — which I assume Jack Dorsey is all about — and puts it into this business: with compute, with cryptography, with everything.

Host: This is where I get really excited. Again — there's macro ethical questions that need to be managed by humans when it comes to AI, but I get really excited about this when it enables people to work together and have that shared access to power that they wouldn't have had before.

Host: Analogous again: we're rolling out these systems for our houses that run the electricity in our house. It's a solar system, but it can feed back into the grid. When technology can help community, that's pretty cool.

Free and open source [27:14] — Host

Host: And it's free as well. This is all free software. It's open source. You download the code itself and you host the code. There's nothing to buy. There's no SaaS to purchase.

What's Jack's play? [27:25] — Host

Host: What is Jack's play? He loves it. He loves crypto. He loves open source. He sees this is the future. Why do you think he's doing this? How's he going to monetise this? Maybe he's not — and maybe that's awesome. What's the play here?

Host: I reckon this is the foundation of the next Twitter. I think the next type of Twitter will be created from this.

Host: He's probably been noodling this idea for years, maybe a decade or so, 'cause these concepts have been around for a while. But agents — that's fairly new. Having an autonomous agent, that's like… 2022. When was ChatGPT first out? It's so fast, how fast it's moving.

But I think business leaders, when they see this, they all go, "Oh, okay. That's impressive. I can actually structure a whole team." I've been doing org design for my clients around adding agents as members. I've designed org charts with all squares on them.

Host: But you've got to then create that agent, and it has to be an autonomous multiplayer agent — or a swarm of agents where there's one agent per employee. It's just a different concept to giving access to code.

Host: So every human employee — at the very least, my prediction is every human employee will have one agent, and they will probably have multiple. Depending on how many, that is what the force multiplier of yourself will be, and they will all start to collaborate in platforms like this. For businesses, you can see the writing on the wall.

Raising agents [28:53] — Host

Host: Is it perfect? No. We're both troubleshooting these agents, raising them right now. They're challenging to set up. You need to put guardrails in place. You need to train them. But that will only get easier and easier, and then they can train themselves. And that's the thing, right? Put them in a chat with each other in Buzz — they're training themselves and iterating and improving.

Host: It's just a beautiful world. It's going to be a race of who can do this fastest. If I'm a tech startup and I leverage this technology versus doing manual coding or just using "full code" even — which is fast — this is accelerating everything you can already do, 'cause you're having a team help you do that same work.

Sovereign AI [29:45] — Host

Host: I think there is a major trend towards data and AI sovereignty. If I'm renting my AI from Anthropic or OpenAI, I don't have a guarantee that I'm always going to have access to that. They could change their pricing, they could remove access to a country, they could limit the access to intelligence to only their country — which they've already done. China's doing the same. So I actually need a way to have sovereign AI as a path towards business continuity. It's now part of a risk plan. How do I do that? It's really self-hosting. And companies don't do self-hosting because they've been in the SaaS world for so long. They think it's better to put it into the cloud — and it's not yours. It's not sovereign. You're renting the space, you're putting your data in it, and if it cuts off, you don't really own that data unless you've exported it or backed it up. You're not really sovereign. I think this is a shift towards that: sovereign data, sovereign technology, sovereign intelligence. It will shift towards this architecture. So Jack isn't necessarily wrong in how he set it up. It's maybe just a bit hard to set up for the average Joe, but it is probably going the right direction of the future.

The solopreneur billionaire [30:57] — Host

Host: And I can see this is also one step towards — everyone loves talking about the solopreneur billionaire — this is one step towards that. Another piece of the puzzle, another tool that will possibly enable that. I think it's already happened, but this feels like it will make it more possible.

Enterprise learnings [31:18] — Host

Host: You are seeing enterprise technology take learnings from these platforms. Gemini has an OpenClaw competitor coming soon. They're trying to build this stuff. So it'll be a matter of time that "Salesforce will probably buy Buzz" .

Host: If it's ever for sale.

Well, Dorsey founded Twitter. He now attacks Slack, which is owned by Salesforce.

Claude, Hermes and Buzz [31:40] — Host

Host: So we've got Claude, Hermes, and Buzz. Buzz is what ties them together. Do you think you'll ditch one or the other, or keep all of those in play for now?

Host: I think I'll keep all of them in play. I really like using Claude because it's got the latest Anthropic models. When I really want to use the most expensive, that's where I'll put it, 'cause I'm just on a plan where it doesn't cost me extra. With those agents, I'm using "GLM 5.2" , which is like a cheaper model, but it's an all-rounder, similar to Haiku.

Host: Yeah, I think I'll still use Claude for the heavy-lifting work and probably use it as my development push engine. It will do the heavy lifting there and then Buzz will coordinate before it deploys. So Buzz and Claude Code will work together.

Buzzed out [32:27] — Host

Host: There's our Buzz deep dive. I'm trying to avoid clichéd puns. Is there anything that we haven't covered on Buzz, do you think?

Host: I think developers will have a different point of view on it, because they already go through the dev process — we're not hardcore developers. They will get the most value out of it initially because it's kind of in their flow, and it allows them to structure their agent teams in a much better way. The power of Buzz with Claude Code for a developer is unstoppable.

Host: Yeah, I totally agree. "Anthropic's doing 80% of its coding with its own LLM now" . But it annoys me when I hear this rhetoric around coders will just be replaced. We have an incredible technical team, and coders with these tools will just be unstoppable.

Host: I've got a couple of mates — 'cause often coders have ADHD — they're like superhuman. They can do a million different things and it's incredible, and they know what the output is. I think a lot of the challenge coders have that are super technical, on the spectrum, is they don't know what to build. Having a product owner or manager — someone like you and I — that's like, "This is what I want, I don't know how to build it." And an ADHD nerd with Buzz and Claude, I reckon you could build anything.

Host: Unstoppable combo — EQ and IQ.

Host: Yeah. EQ and IQ, powerful combination. Anyway — Buzz. We're here to talk about Buzz. Are we buzzed out? Is this it? Should we leave it at that?

Host: I think we're good.