AI Without Compromise: Why Data Sovereignty Is the Next Enterprise Battleground

00:00:10:25 - 00:00:31:27 Michael Welcome to this ambiguous. I'm your host, Michael Fauscette. Each week, we interview experts in artificial intelligence, generative AI, and business automation to help business leaders understand how to use these tools for the biggest business impact. 00:00:32:00 - 00:00:43:18 Michael Our show today is AI Without Compromise why?

Data sovereignty is the next enterprise battleground. I'm joined by Clayton Bryan, head of enterprise at Quill meetings. Clayton, welcome. 00:00:43:21 - 00:00:46:20 Clayton Thank you, Michael, for having me.

Looking forward to a great conversation. 00:00:46:23 - 00:01:04:08 Michael Yeah, I think this will be really interesting. And certainly data sovereignty and and all the other topics around that are definitely hot these days. I, you know, just to get a started, could you give us an introduction, tell us a little about yourself and your journey and how you got to Quill meetings and maybe a little bit about Quill meetings as well?

00:01:04:10 - 00:01:30:08 Clayton Sure. Yeah. So I, have, spent the last ten years as an early stage investor at a shop called 500 global, formerly known as 500 startups. And I started as a, as it relates to my journey with Quill, as I was doing my work as an investor.

And, you know, I'd be meeting with, hundreds of, founders on a monthly basis. 00:01:30:10 - 00:01:50:23 Clayton I found myself, you know, constantly needing something that could, capture, my, my meeting notes. Help me process those meeting notes. Help me stay on top of of of my job, essentially because I was just talking to so many different founders, and, Quill was introduced to me.

I happened to know, the founder, Mike O. 00:01:50:23 - 00:02:19:16 Clayton And, he told me he's working on, Quill. And I said, hey, let me take a look at this, because I can absolutely, you know, use this. I the analogy I like to use a lot is I felt like a very busy doctor in a very busy hospital.

I'd be, you know, going from meeting to meeting to meeting, meeting with different founders, talking about everything from portfolio support to fundraising to, in some cases, even playing the role of, the founder marriage counselor, trying to make sure that my teams did not, have fisticuffs with each other. 00:02:19:16 - 00:02:35:09 Clayton So so it was it was just a lot of information for me to always have to process. And Michael said, hey, check out Quill. And so I started using Quill, and I quickly became a super user of Quill.

And I love Quill so much that I said, hey, you know, I, I think that you have you really have something special here. 00:02:35:12 - 00:03:01:10 Clayton And I want to I want to join and I want to help you out. And, and here I am leading, leading enterprise and, and some of the things that make Quill special is really around the architecture. And the architecture is built in a way in which, the transcription is local.

The, summarization does get processed in the cloud, but if you're, enterprise, you can bring over your entire your own lens stack. 00:03:01:10 - 00:03:40:00 Clayton So we have a Bill lamb, feature. And so all the data stays under the control of those enterprises. And so once I was able to tell my compliance officer that at at 500 and I, you know, broke down the architecture, why is this different?

I, I was I received the blessing to use it. And, once I saw that story up front, I said, hey, if I had to go through this and I'm at a $2.5 million AUM, operation, I'm sure that there's going to be other users in regulated industries that are also going to really, want to have the same AI magic at their fingertips, but know that they're 00:03:40:00 - 00:03:48:03 Clayton going to get blocked by their compliance policy. And so I wanted to come on board to, to help lead that charge because I think it's such a special problem.

00:03:48:06 - 00:04:06:19 Michael Yeah, I mean, it. Quills tagline is AI without compromising. I mean, that's in the world we live in today. That's a pretty bold claim.

I especially where you know, a lot of the AI tools that companies use, you know, they're presented to them in a way that it's a cloud service. You have to put your data in the cloud. 00:04:06:19 - 00:04:21:13 Michael Data sovereignty is kind of questionable. Data residency is questionable.

I mean, what what makes Quill different? And you hit on it a little bit, but let's go a little deeper there. And and what a founding team decide to build it this way. You know, from the ground up.

00:04:21:16 - 00:04:46:14 Clayton Yeah. So the founding team is has a really strong philosophy around local AI and around data sovereignty. Your data should stay with you. It should stay controlled by you.

And that that's a part of their first principles. Thinking and, and, and and so that, you know, as, as relates to just everything, any, any of their data. 00:04:46:14 - 00:05:10:02 Clayton Right. And so because of that DNA that, was present with the founders that you see that also, within the product itself.

So, so again, the audio never leaves your device for Quill. There's, we leverage whisper for local, transcription processing. And then by default, the summarization in the minutes will, go to the cloud. 00:05:10:05 - 00:05:34:10 Clayton However, our, enterprise clients, we, we have a, bill lln and all the enterprises, that's what they choose to do.

So they they leverage they have their own GPU cluster. They have their own private cloud. You know, they're able to put Quill, have the processing go through, you know, their, their stack so that they understand and they have full control over where that data goes. 00:05:34:10 - 00:05:36:17 Clayton And that's the way that we think it should be.

00:05:36:19 - 00:05:58:10 Michael Yeah. I mean, that is definitely unique. And I, I've used several, you know, probably no surprise, but I've used several different types of services and most of them have been cloud based. And, you know, honestly, in the end, you start to wonder.

There's a lot of, and in my business, a lot of NDA type information that I, you know, conversations I have with my clients and all that. 00:05:58:10 - 00:06:17:16 Michael I just got really uncomfortable about so I could see where that makes a lot of sense. And particularly, you know, tied to it's on my local device. And then it sounds like there are plenty of enterprise options to, to to set that up and, and have it have the architecture run in a much better controlled way.

00:06:17:19 - 00:06:45:08 Michael And it sounds like you you also offer this kind of an air gapped option, right. So you can you can bring your own model, disconnect from the internet. And, you know, I'm curious, as you talk to companies and have been working with companies, what kind of organizations are choosing that path? And what does, you know, what's the does the demand for that level of isolation tell you about kind of where enterprise AI adoption is, is headed?

00:06:45:10 - 00:07:21:20 Clayton Yeah, it's the companies that have the most to lose. And so it's companies that are dealing with, confidential, classified information. So, you know, I think, enterprises that touch, the power grid, that touch, the fence, that are deep tech, you know, that they can't have, vital information, seep out or, or in other cases, you know, they just, you know, they need to be absolutely sure and have absolute control, of the servers where that information, resides, and is processed.

00:07:21:20 - 00:07:37:24 Clayton And so those are the types that are saying, hey, you know, we've been kind of left out in the cold as it relates to all this AI magic. We we we obviously we, you know, they talk to to peers and, and friends and they, they understand the power of of what some of these, no takers can do. 00:07:37:24 - 00:07:57:10 Clayton But they absolutely also know that, the, most sensitive information, that they have, will be discussed in these meetings. And, they want to make sure that it is fully, closed off to, the internet and, in any type of, breach capital.

00:07:57:12 - 00:08:17:18 Michael I would think that works well in regulated, highly regulated industries, too, or even like medical, you know, health care where you've got a requirement and there's a lot of a lot of regulatory issues or even like in in Europe where, you know, privacy protection is much greater. And, and some of the states in the US too. So that that makes a lot of sense. 00:08:17:21 - 00:08:58:11 Clayton And if I can just add you, we've had a lot of very conservative, staunch GDPR type of shops.

Reach out to us because they said, hey, you know, your architecture is, is, is really exactly what we need. And you know, so I have over the last four weeks, I mean, I've kind of gone on a little bit of a not in person, but a virtual tour of different EU, enterprises that are telling me, you know, just, how, critical their, their GDPR stance, is and, and, and just how exhaustive of research that they've done to try to find, products, that fit, their policies. 00:08:58:14 - 00:09:22:18 Michael Yeah.

I mean, it does sort of make sense, too, because, you know, let's be honest, it's a lot easier to to set up a cloud application with a single delivery method than it is to build something that gives you end user choice or business choice and is flexible enough to, to use a model, bring your model, deploy locally to pull in the cloud, deploy in a hybrid way. 00:09:22:18 - 00:09:28:09 Michael I mean that that's that's more complexity, which certainly adds to costs, right? 00:09:28:12 - 00:09:50:19 Clayton Yeah it does.

So convenience is a big key a key part. Right. But you know but like you said, if you are regulated by, you know, an alphabet soup regulator, if you, you know, CMS and CMC, Iter, GDPR, you know, all the acronyms, right. You know, then you convenience, is is not as significant to you.

00:09:50:19 - 00:10:02:13 Clayton You really, absolutely need to have, your information, control over where it lives, where it's processed. You know, that's just table stakes for you. 00:10:02:15 - 00:10:38:15 Michael Yeah. I mean, we were talking earlier.

You talked about an interesting adoption patterns where, you know, you have an individual who discovers Quill starts to use it, is happy with it, and then, you know, later their IT team, they're CSO investigated and they end up, you know, really surprised that it works the way you say it works. I mean, what does that tell you about the gap between kind of enterprise expectations around AI vendors and what they're mostly seeing delivered, particularly on the privacy front, where we know there's a lot of a lot of risk and and concern.

00:10:38:17 - 00:11:09:11 Clayton Well, I'll say, you know, CISOs, compliance officers, are really starting to wake up to the fact that, you can't trust these third party, servers for the highly regulated, industries. And, you know, it is, there's a lot of fear, you know, there's a lot of fear about getting, subpoenaed and, breaches and, and so, you know, naturally, the gut reaction is, hey, you can't use this, right? 00:11:09:11 - 00:11:28:22 Clayton Because they're just thinking of, okay, well, you're probably just like everything else.

You're probably just storing it at third party cloud. You might have good data protection agreements, but but we still don't trust that. But as soon as they start to scrutinize the architecture and they see that, hey, you know, our data sovereignty, is that bolted on? 00:11:28:24 - 00:11:39:20 Clayton This is a a a, foundational principle, that is at the core of our architecture.

And once they understand that, then they give the green light. 00:11:39:22 - 00:12:17:03 Michael Yeah. I mean, that is a that is a very good story when we think about a lot of the different, conversations that I've been having lately and, and on the show, too, like, we've talked a lot about data sovereignty and governance get built in governance, you know, as companies are bringing more agents, you know, from that posse, you know, proof of concept kind of environment out into and into an actual workflow and, and start to scale up, and, you know, most organizations are kind of bolting compliance on their existing tools, maybe.

00:12:17:03 - 00:12:34:04 Michael And it seems like you guys have taken a different approach when you're designing privacy into the architecture from the start. I mean, why do you think bolt on workers broken? And then what should companies, you know, really look at when they're evaluating AI tools? 00:12:34:06 - 00:13:08:25 Clayton Well, just thinking about like the retrofit of of doing that I mean, governance can't be a layer.

It has to be the foundation. And, and and if it's not, you're going to leave yourself vulnerable, to things that are going to, be much more expensive, timely, just big headaches down the line. And so it's important that at the, at the start you're thinking about this and at every step, every iteration of the product is something that is, standard that you are, building by, because it's just not going to work. 00:13:08:25 - 00:13:31:13 Clayton We've seen you know, some of the other as a, as a, parallel to this, just in terms of just even the adoption within AI.

Like, I do think that there's this notion of a productivity J curve. You know, if you look at innovations throughout the course of history, you can't just bolt on an electric motor to a steam, workflow. 00:13:31:15 - 00:13:48:05 Clayton You know, you can't you can't bolt on, governance and data sovereignty to an existing, SAS product that has never had to think about that before. There's just too much that you're going to have to go back and and fix too many holes that are possible.

00:13:48:08 - 00:14:14:24 Michael Yeah, that makes sense. I mean, I've been writing and talking a lot about in the last, you know, 5 or 6 months about governance by design, the idea that we have to shift the way we think about governance in the genetic world from a process that sits in the top at the end to something that's inherent inside of the platform that has to be built from the ground up because of the fact that you're you're talking about governing something that operates at machine speed.

00:14:14:24 - 00:14:28:13 Michael It's it has access to, you know, all sorts of tools and data and, and it's just such a risk if you don't think about it in this new way of, from the ground up versus from the outside in. 00:14:28:15 - 00:14:31:24 Clayton Exactly, exactly. That's been our findings as well. 00:14:31:26 - 00:14:56:21 Michael Yeah.

So so I know your your co-founder, Michael built, angel lists syndicates, led Angel, his ventures, you know, past $1 billion in assets and, and and you also have an investor, background. How does that background shape the way you think about building Quill, especially, you know, around what enterprise buyers actually need versus, you know, what gets funded? 00:14:56:23 - 00:15:23:17 Clayton Yeah. I mean, you know, when you have a founding team that's managed billions in assets and sat on boards of, of public companies, you know, they they don't need someone to explain why hedge fund won't send meeting recordings to a third party cloud.

You know, they've been that buyer, right. And so it's just deep domain expertise, deep domain experience expertise as it relates to how these industries think. 00:15:23:20 - 00:15:44:21 Clayton I myself, you know, I have a lot more gray hairs than I'd like to admit for my age, because I've gone through a lot of battles with my compliance officer. So I, I understand, you know, what it takes to have tooling that is compliant.

And it's absolutely magical. And it delivers the cutting edge AI output, and you're doing it in a way that's compliant. 00:15:44:21 - 00:16:05:19 Clayton You're doing in a way that all of the gatekeepers inside can live with, can be happy with. In fact, you know, one of our enterprise clients is a, publicly listed deep tech company.

And, you know, the CSO sent us over, a, security survey. We we answered it. And as soon as he understood the architecture, he became an advocate. 00:16:05:19 - 00:16:34:19 Clayton He was like, I, you know, this is a green light, right?

Like I you know, it's kind of like we've been looking for something like this, and we've been so frustrated because we haven't found, the tooling that understands, you know, the different hurdles that we're going to have to go through because of the sensitivity of the, of the data and the and the, in some cases, the standards many get to the standards of the regulatory bodies, that, are governing them. 00:16:34:22 - 00:17:00:12 Michael You know, most of the people, and this is a common thing, right?

Most of the people are using ChatGPT or cloud or perplexity, you know, all the, Gemini, the other kind of major tools, they're really sort of only using it personally outside of work, maybe mostly kind of as a search function. So they're not they're not really getting real leverage, from in their professional lives. 00:17:00:12 - 00:17:11:10 Michael Right. Why do you think that exist?

And then and then how can quil help close that gap and make it, you know, more relevant from a business perspective? 00:17:11:12 - 00:17:40:18 Clayton Yeah. You know, so ChatGPT is a great tool with a massive professional trust deficit. Quill solves that problem for most information rich moments and at any professional day.

And so, you know, I think it really again, it goes down to, you know, we're not asking you to, to change your workflow. You know, we sit at the OS level, we capture the meeting and we deliver the intelligence without requiring you to trust a third party with your most sensitive conversations. 00:17:40:20 - 00:17:59:24 Clayton And, and, you know, a model company can say, like, you know, we're not going to, you know, train on your, on your data, but but really where, where is it?

Where is it going to rest? Where is it going to store, where it's going to be stored. You know, what happens if that company, gets acquired, you know, like what happens to the data that's in that system, right. 00:17:59:26 - 00:18:09:13 Clayton Well, well, for our enterprises that are leveraging bring your own LM, all the data resides within your own infrastructure.

So you don't have to ask those kinds of questions. 00:18:09:16 - 00:18:29:06 Michael Yeah. I mean, that certainly builds a lot more, confidence and and, you know, I think the other thing, a lot of people kind of don't really understand the risk of using a public model for things that should be controlled, like, like privacy information or, or even like in my business a lot. I get a lot of NDA.

00:18:29:08 - 00:18:51:17 Michael Information from clients and, you know, in the conversation and like I, I run a local model that's gapped for that because I'm certainly not going to upload this. The slide deck into cloud or ChatGPT or anything like that because there is certainly risk. So I think, you know, a lot of people don't maybe realize that. Or if they they do, they don't really know how to work around that.

00:18:51:17 - 00:19:18:21 Michael So that that really resonates. Yeah. So, you know, beyond business meetings, which, you know, obviously that's a very good use case for you. But you've talked about other use cases.

So like, oh, I have a doctor's appointment and I'm going to, you know, for a family member, for example, you know, I want to record that or, or maybe I'm having a mortgage conversation and I want to document that at the bank situations where privacy is really critical from a personal perspective. 00:19:18:21 - 00:19:31:14 Michael Right. And, and, and the stakes are high. How do you think about the line between enterprise and personal AI?

And is that line maybe not even useful anymore? 00:19:31:16 - 00:19:51:03 Clayton Yeah. Well, I think that the line between enterprise and personal AI is dissolving. The question is always the same who controls the data?

The answer should always be the person whose life it represents. Yeah. You know, I think that that that's that's very important. You know, we didn't build separate products for enterprise and personal use.

00:19:51:03 - 00:20:01:20 Clayton The architecture that protects a defense contractors classified briefing is the same architecture that protects your family's medical information. That's what local first needs. Yeah. 00:20:01:23 - 00:20:22:02 Michael Yeah, that that makes sense to that.

Really, the problem is the same, right? We're trying to protect whatever that data or information is and knowing where it is, how I've, you know, how I've managed to store it and control it and use it, those kinds of things that actually gives you that confidence that that you can use there. 00:20:22:02 - 00:20:23:27 Clayton So exactly. 00:20:24:00 - 00:20:48:21 Michael So, you know, I've been writing lately quite a lot about, about what I've called the automation trap.

And it's, you know, where companies, enterprise, AI, they bring it in, they, they bring it into their existing workflow, but they don't redesign the process. They don't rethink the work and how that's going to, you know, work in this new environment with the new agent and tools. 00:20:48:24 - 00:21:14:15 Michael Quill seems to go further than just transcription into automating, like the post meeting workflows, like updating CRM or, you know, creating a task in my to do list, task management software, whatever that might be.

How do you think about the, the, that transition from capturing information to actually acting on it in some way and how you could enable that? 00:21:14:18 - 00:21:53:06 Clayton Yeah, I think we said that a great place to do that. You know, we have enterprise customers who are manually copying and pasting meeting notes into their CRM, multiple steps, multiple systems, losing contact context, and any handoff, Quill can push structured data directly into Salesforce, HubSpot, affinity, Pipedrive.

The meetings become the source of truth. And and so, you know, we we see this as a way to help to improve this workflow, and, and show that there can be an evolution of this workflow where within my quill, we have a cool agent colloquium. 00:21:53:09 - 00:22:27:03 Clayton I can take a, have a call, you know, this week I've had several calls with, enterprise, clients of, of Quill. And as soon as I'm done with that meeting, I just tell the agent, go into, our CRM, update, that, that contact update that objects, with additional, contacts from this meeting, you know, so, so that that's that's really kind of where, where we see, it going, you know, I saw that you wrote about the shift from tools to coworkers.

00:22:27:05 - 00:22:43:27 Clayton Meaning intelligence is the foundational input for agents and workflows. Before an agent can act on what happened in the meeting, something has to capture what happens, accurately, securely, and in a structured form. That's Quill's role in this agent. Exactly.

00:22:44:00 - 00:23:11:15 Michael Know that that does that fits really well into this kind of thought process of both of how do I, you know, how how do we integrate this into the work? But then how do I make the work better? By changing the workflow, changing the process. And, and I mean, you know, for years in the, in the industry, we've talked about how bad a lot of company CRM data is because it just didn't get updated or it got updated in multiple ways in the different formats with the wrong emails, with the, you know, whatever.

00:23:11:15 - 00:23:23:22 Michael Right? So having a way to automate that and make sure that it's consistently done, I can see, you know, a huge benefit from a data management perspective as well. 00:23:23:24 - 00:24:06:23 Clayton Data management. But also just think about, we're removing aspects of, of of just tedious work, a lot of manual work.

Right. And and so now that you have time back, that time that you formerly were spending, you know, just reporting stuff over to your CRM, right. Maybe auditing it right. I can just outsource all of that to, my, my agent at this point, and it goes beyond just CRM, you know, we plug also into project management, platforms, like linear and Trello and, and so, so I can, you know, if I find a bug, in, in Quill, I can, quickly, basically have 00:24:06:23 - 00:24:27:06 Clayton the Quilliam, the quill agent, update our linear ticket, or a feature request.

If I'm talking to, you know, a client that wants to do something, that we don't currently offer. Right. And so it's just making me much faster. It's making other organizations much faster and giving time back so that you can now, you know, do other things, be more productive.

00:24:27:09 - 00:25:04:06 Michael Yeah. I mean, I think that conversation is interesting in, in business today because a lot of the, the hype has been, oh, you know, this is going to take away jobs. It's going to, you know, mean that a lot of people won't have this. What's that.

Because things do change. And to me it's it's been more about I can see how skills shift and uplevel and even your ability to do higher level tasks and take care of things that have much more value, because a lot of those kind of routine things that I had to do can be automated, and I don't have to waste my time on them when I could easily, you know, 00:25:04:06 - 00:25:11:19 Michael be out delivering another project or, you know, working with strategy, that sort of thing that is higher value to my company.

00:25:11:21 - 00:25:35:03 Clayton Exactly. And I think that, effective, tasteful applied AI will allow you to spend time on the things that you want to spend more time on. Not not the tedium, not the, the the mundane aspects of work. Right.

You know, if you're, deal person Quill should allow you to spend more time talking to other humans and doing more deals. 00:25:35:03 - 00:26:01:00 Clayton Right? Not like the admin side of it. Right.

And so that that's really, important. And you touched on something around, you know, jobs and job loss and, and and I think that again, properly structured, applied AI will allow you to do a lot more. There will there will be, you know, more jobs, and then less, you know, I think it's a little bit scary because like, oh, it has the ability to do all this automation. 00:26:01:02 - 00:26:26:19 Clayton But I think it's going to also open up new opportunities.

And so we're, we're looking forward to those new opportunities, those new workflows, those new jobs. You know, I tell people all the time, 26 years ago, in 2000, you know, the term data science, you know, wasn't really, you know, an industry, today, you know, you can get a, a degree from any of the world's top institutions in data science. 00:26:26:21 - 00:26:44:02 Clayton Right. And so I think we're we're seeing something similar.

Another analogy I like to, allegory I'd like to go back to is I remember when I was growing up, I used to see an acronym on certain job descriptions, wpm word words for a minute. They wanted to see, you know, how good of a type of story, right? 00:26:44:04 - 00:27:02:24 Clayton And, and typing was a skill that not everyone in the 90s and 80s, not everybody was, was proficient in typing. You know, I think we're in a similar, something that rhymes because now it's about how AI native are you how how effective are you with workflows with automation.

Right. And so I think it's just it can be scary from certain vantage points. 00:27:02:27 - 00:27:18:06 Clayton But I think, you know, looking at our history pattern matching with our history around innovation, I think that we're going to be in a much more, value, creative and equitable future because of the technology that is now starting to emerge. 00:27:18:09 - 00:27:46:23 Michael Yeah.

I mean, I think it I think it can certainly be seen that it is going to open up new avenues. It's going to mean that certain jobs are going to need different skills, new skills, whatever. But but at the same time there's opportunity. Like, I, I know I was talking to a, CTO of a company not long ago, talking about development and, and vibe coding and how, you know, a lot of developers have been really fearful that by doing this vibe, coding and bringing in coding tools automates a lot of the process.

00:27:46:23 - 00:28:14:15 Michael It means developers aren't, you know, needed. And that's absolutely not the truth. In his case. You know, in his case, he's like, no, no.

And in fact, you know, that's the opposite is true. It means that the developers in the process are even more valuable. But they have to learn new skills, right? They provide the conceptual model, they provide the output, they help make sure that things are built correctly and work correctly, because you really you're thinking about it like you've got a roomful of very junior devs.

00:28:14:15 - 00:28:29:01 Michael I don't want them doing everything. I want them to do what they do well, and then I want my developers, you know, that have been there for years and have the depth and capability to do what they do well. And then we we have a much better, more productive system. 00:28:29:03 - 00:28:44:29 Clayton Exactly.

You know, I, I think that, with Cogen specifically, you know, you're, you're able the developers I talk to you, I work with that quill and and and others that I work with, through my career as an investor. You know, they're telling me that, you know, their job, the nature of their job is changing every six months. 00:28:44:29 - 00:29:16:02 Clayton You know, they're not, you know, manually coding as much anymore. Instead, they're managing 5 to 10 instances of cloud code, or leveraging Curser or Codex, to go much, much faster, you know, so we're going to we're going in a direction where, the, the, the nature of work is changing.

The tooling is changing. Now, you probably don't want to vibe code something that ends up in production, but you can absolutely and should absolutely be vibe coding. 00:29:16:05 - 00:29:34:24 Clayton Prototypes. Right.

So maybe instead of more PRD, you see more prototypes that are vibed coded so that folks can, more quickly, easily understand the direction that is going. You know, I, I vibe code features all the time. And I sit with enterprise clients who say, hey, is this something that you would like to see in the product? 00:29:35:01 - 00:29:54:09 Clayton And then a week later, I can push that live in the product right after it's been coded up the right way.

So so I think it's allowing us to go much, much faster in areas of development and business building where previously there just be much longer feedback loops, much longer, development loops. The timelines are compressing. 00:29:54:11 - 00:30:13:08 Michael Yeah. So so, you know, let's sit back and if we get out of a crystal ball and believe me, in AI world, that's a tough one, right?

Because everything moved very quickly. And, you know, I, I used to ask this question, I would go like, you know, 24 to 36 months out. I, you know, some days I feel like 24 to 36 hours out is a good question. 00:30:13:08 - 00:30:35:15 Michael But but anyway, let's think about looking ahead a little bit.

You know, as AI agents become more capable. And then of course, privacy concerns are going to continue to be top of mind and maybe even get, you know, more. So where do you think the the market's heading. You know, will local first privacy by design approaches.

Is that going to become the standard. 00:30:35:18 - 00:30:47:21 Michael And do you think most companies you know that continue trading data sovereignty over convenience? Or are there just going to be better solutions available so that they can actually address those issues? 00:30:47:23 - 00:31:13:14 Clayton Yeah.

You know, I think that the, the honest answer, both paths will coexist. Convenience will always win for consumer use cases. But for the regulated industry, financial services, health care, legal defense, government, local first will become the minimum standard, not the premium option. You know, regulation is a one way ratchet.

The SEC enforcement isn't getting lighter. 00:31:13:20 - 00:31:25:11 Clayton GDPR fines aren't getting smaller. Cmmc requirements are getting less strict. Every year the compliance bar goes up, and cloud dependent AI tools on the wrong side of that trend.

00:31:25:14 - 00:31:56:28 Michael Yeah, yeah. Well, so Clayton, really interesting conversation. I really appreciate you joining today. And, you know, I, I think this is such a hot topic that we really do need to bring to the front, for, for everybody, from a business perspective especially a lot of concern there.

But before I let you go, I always like to ask at the end of the show, you know, can you recommend somebody thought leader, an author, podcaster, you know, somebody that you think the audience would get real value out of if they they checked them out and followed? 00:31:57:00 - 00:32:17:29 Clayton Well, before I do that, one other thing I wanted to do with the last question, I wanted to draw parallel to encryption. You know, 20 years ago, Https was optional. Now it's table stakes.

Data sovereignty for AI processing is on the same trajectory. The question isn't whether local first becomes the standard for sensitive context. The question is how quickly. 00:32:18:01 - 00:32:40:18 Clayton So I think that that's important.

And, you know, the companies that build for sovereignty now won't have to retrofit later. That's a structural advantage. And that's why we built Quill this way from day one. So I wanted to just, add a little bit more to that last, question, and, then I can I'll answer the final question.

00:32:40:21 - 00:32:41:12 Clayton Yeah. 00:32:41:14 - 00:32:49:28 Michael No, I do think you're right, though. I think data sovereignty by design that certainly makes a lot more sense to me than than the way we've approached this in the past. So.

00:32:50:01 - 00:33:11:26 Clayton Yeah. Yeah. And, and 100%, 100%, I think more and more people are going to wake up to that. You know, I think it's been happening.

We've been seeing it. But the momentum of that realization is going to start to compound faster and faster. And so, the last question, who to follow? I'm a big fan of, Matthew Berman.

00:33:11:29 - 00:33:35:15 Clayton He, has a, YouTube, show. He does live shows. But he he's always he's very in the weeds in terms of the latest models who's releasing what? But I love the way that he breaks things down, with his video podcast.

Because it's it's it's easily digestible for the non-technical user as well. 00:33:35:22 - 00:34:05:18 Clayton So if you're curious about cutting edge, you know, he just, did something about the latest, ChatGPT model. The 5.4, it's he's a great go to, and he's also getting some really fantastic, interviews with some heavy hitters.

From, the world of AI. So, the CEO of CEOs of large companies, but also some of his videos, specifically in terms of how to leverage, current tools. 00:34:05:21 - 00:34:17:11 Clayton I mean, I just went down a rabbit hole with everything that he's done with, open cloud, for example. It really, really fascinating stuff.

For, both the technical and the non-technical users. 00:34:17:13 - 00:34:29:13 Michael So. Yeah. No.

That's great. Thanks. Thanks. It's a very good recommendation.

Certainly. Audience to check them out. So again, Clayton, thank you so much. Really interesting conversation.

I really appreciate you joining. 00:34:29:15 - 00:34:34:19 Clayton Thank you. It's been a pleasure, Michael. Thanks.

00:34:34:22 - 00:34:55:10 Michael And that's the show for this week. Thank you all for joining us. Remember to like, share and subscribe to the show. If you enjoy the show, please leave us a review to help others find us.

For more research on AI and other software, check out arionesearch.com. If you're an expert in AI, generative AI, or business automation, either as a provider or an end user, email your information to disambiguation@arionresearch.com.

00:34:55:10 - 00:35:06:22 Michael Get disambiguation at arionresearch.com. Don't forget to join us next week. Disambiguation is an Arion research production, I'm Michael Fauscette and this is the disambiguation podcast.