This Week in Enterprise Tech Episode 557 Transcript

Please be advised this transcript is AI-generated and may not be word for word. Time codes refer to the approximate times in the ad-supported version of the show.

Curt Franklin (00:00:00):
Today we have the New York Times arguing that chat, G p t will have to start over because of copyright issues. Verizon is saying their five G expansions will certainly help rural America. And we'll be talking to Dan Lawyer of Lucid Software about how AI is changing the way our teams can collaborate. Quiet all on the set

TWiT Intro (00:00:25):
Podcasts you love from people you trust.

Curt Franklin (00:00:30):
[00:00:30] This week in Enterprise Technology, episode 557 for August 18th, 2023. How lucid are your collaboration tools?

Lou Maresca (00:00:51):
This episode of this week in Enterprise tech is brought to you by a c i Learning. A c L Learning has partnered with the best in the industry by adding insights. [00:01:00] The new skill gap analysis tools is also there to ensure you that the training you're providing is actually working. Visit go dot acl earning.com/twit twit listeners will receive at least 20% off or as much as 65% off an IT Pro enterprise solution plan. The discount is based on the size of your team, and when you fill out their form, you'll receive a proper quote tailored to your needs and by an optica, reduce the complexities of protecting your workloads [00:01:30] and applications in a multi-cloud environment. And Optica provides comprehensive cloud workload protection integrated with a p i security to protect the entire application lifecycle. Learn more about panoptic aopa.app.

Curt Franklin (00:01:46):
Welcome to twt this week in enterprise tech. I'm your host, Kurt Franklin. Happy to be with you this week and I'm happy to be here. Along with my co-host, [00:02:00] wanna welcome both of them. First, Mr. Oliver wrist back with us for one of his specials, star co-host engagements. Oliver, how are things with you?

Oliver Rist (00:02:14):
Things are good. Things are good. The weather up here is incredibly stormy and, and rain ridden, but other than that, I'm, I'm doing fine.

Curt Franklin (00:02:23):
Well, we're glad that you've made it through the raindrops to be with us. Also, glad to have with us my co-host [00:02:30] and our producer. We'll talk about that later. Mr. Brian Chi. Brian, how are things on your end of the neighborhood? I'm doing fine. I'm actually finally getting out and about. COVID was not my friend. It really, really laid me out for almost four weeks. But I'm recovering. I still have a cough. I'm still sucking down cough drops like

TWiT Intro (00:02:56):
Crazy.

Curt Franklin (00:02:58):
But yes, I have [00:03:00] been getting your emails and I've been trying to reply as I can. And I, I should say to some of our viewers, every once in a while, we, even though we try to stick towards the technology and the policies and you know, the decisions behind the technology, some, sometimes it starts coming out like a commercial. So we apologize for that. We attempt to stay more on the behind the scenes view but it doesn't [00:03:30] always happen. So apologies for that. We appreciate that. And today we're definitely gonna, not going to have anything that could be confused with a commercial. We've got a great guest. Before we get to that great guest though, let's talk about some blips. Hey, do you use Zimbra? Then you may be in trouble because of phishing campaign targeting customers of the zimbra Collaboration software Suite has spread [00:04:00] to hundreds of organizations in over a dozen countries.

(00:04:04):
According to an article on dark reading, Zimbra a collaboration suite that is a niche offering with far fewer users than the major office productivity suites has been hit by a series of security issues this year, and the new phishing campaign is simply the latest. According to researchers at eep since April, 2023, an unidentified threat actor has been using scattershot phishing emails [00:04:30] to call credentials for privileged zimbra accounts. The primary targets have been small to midsize businesses, which happens to coincide with the open core software primary customer base. There have been some government organizations that were hit as well. Mostly you would think at the city, county, and state level. Now the phishing email messages all follow the same formula. They say they're from the zimbra Security Service. Talk [00:05:00] about a change in the login webpage and close with a salutation from the very convincing ZIMBRA boss administration.

(00:05:10):
The message has an H T M L file attached, which directs the user to a generic ZIMBRA login page, but has some identifying elements customized for the target organization. And the page opens in the user's browser, though it is on a local file path, and it prefilled the username field in order to give the impression that it is a legitimate zebra [00:05:30] login page. And at this point, any user who types in their password will be sending the sensitive information straight to the attackers to avoid compromise. Researchers recommend standard security hygiene using strong passwords, multifactor authentication, and updating to the most recent version of zimbra. I'll add one more. Hey, don't go clicking on obviously fake email phishing messages. Zimbra boss. Wow.

Oliver Rist (00:06:00):
[00:06:00] So despite all the tinfoil T 1000 warnings and the fact that chat G T P is already carrying not one but two, kill it. Now lawsuits as I think I mentioned last time that I was here everyone is still making a mad dash to deploy ai. So that means new infrastructure and according to the register that's already having a big effect on how data centers get built. So in a data center, the conventional heat dissipation strategies seem to work great for six to 10 kilowatt [00:06:30] racks. But those apparently face plant, when you put 'em up against all the GPUs that you need to run an AI training model that's forcing data center operators to make some very serious changes. So the example that the register used was Tesla. Though it doesn't seem to be a hundred percent on whether that's accurate or not, but they think so because apparently the Musk grad put up a job description last week for a data center program manager would be responsible for building and I quote first of their kind data centers, unquote.

(00:06:58):
And yes, that's plural. [00:07:00] There are a bunch of reasons Tesla might be building us on data centers from evolving self-driving capabilities to supposedly they're developing a humanoid robot that needs an offsite brain whatever, though all that digital thinking is gonna be done on GPUs and those going to eat power like it's going outta style, which is gonna keep those data centers pretty damn balmy. So Tesla wants a place where it can support, you know, their homegrown supercomputer that it wants to build. Apparently for the same reasons that I just mentioned they're calling [00:07:30] the thing Dojo and in July, Musk was bragging that it'll pump out on the order of a hundred ex exo flops, which is a lot of floating points that you gotta keep going. Any data center that's gonna handle that kinda load for completely customized system is gonna have some really nasty power and cooling issues. So it seems to make sense that Tesla's gonna run at that from scratch. Now, that may be a pretty or semi extreme example of this problem. But they're not gonna be alone with this [00:08:00] AI issue. So if your boss is suddenly having really, you know, bright eyes when somebody mentions ai, it's probably really good time to figure out how all your new GP, new G p U nodes are gonna do when it comes to dumping all that heat.

Curt Franklin (00:08:16):
Actually, you know what's really funny? Oliver, Kurt and I all worked on a InfoWorld article called Pimp my Data Center, and we were actually handling heat loads that big [00:08:30] almost 15 years ago, I think. Anyway, so my blip is from our friends at the Verge. Verizon is saying bragging that it's weeks away from a major upgrade to Rule five G. Well, what it's saying is it's up to 200 megahertz. C-Band, five G Spectrum has been cleared for use four months ahead of schedule. An upgrade it says, will bring more bandwidth [00:09:00] and faster throughput to its customers, particularly those in rural areas. The company will use the new spectrum to expand its five G based home and business internet bi business in addition to improving its wireless phone service. The spectrum is available after the F C C validated the moves by satellite communications companies, s e s and Intelsat to switch to higher C-band frequencies, clearing the lower ones for commercial cell phone use [00:09:30] on X.

(00:09:31):
Formerly Twitter people have recently reported seeing a hundred megahertz of spectrum for its five G ultra wideband service in rural areas like Intercourse, Pennsylvania and Bristol Virginia showing ulu speed test, screenshots of wireless throughput, normally only seen in small sections of large cities. Cities Verizon says has made its new C-Band Spectrum available to 359 markets so far. The Verge has asked for a full list [00:10:00] and will update if the company ever responds. Now, I'd like to remind folks that one thing that is contributing to this is the shutdown of three G in a large part of the continental US and reallocation of those bands for the five G shuffle. 'cause This just doesn't happen with one service. Think of it like the so-called trickle down effect where bands shuffle around to make more space for services like the five G Ultra wideband. Where more [00:10:30] channels are available for bonding. That's the key to the bigger bidder. Better speeds. This is especially beneficial to those areas with lower congestion like rural America, where low user densities allows users to bond more channels together for better speeds, which ought to make our co-host Heather Mo Williams very happy, often a little cabin by the woods. [00:11:00]

(00:11:00):
Well, this is starting to sound like a broken record. A White House is yet again telling the executive branch to take cybersecurity seriously. In this case, it's ordered federal agencies to get their cybersecurity safeguards up to date by the end of the year since they lag in their ability to implement President Biden's executive order, which was issued in 2021 in a dark reading article, they report that Jake Sullivan National Security Advisor is complaining [00:11:30] that due to a lag in essential cybersecurity implementations, the US government has been exposed to malicious cyber intrusions and agencies that have fallen behind are failing to set a good example for other organizations. We can all rest easy though because senior officials have insured the National Security Advisor that they're going to achieve full compliance with the security requirements included in the executive order. By the end of this year, prepared to see a flurry [00:12:00] of statements designed to reassure the White House and the public that they're just on the verge of putting in place the best darn cybersecurity this world has ever seen. Unfortunately, security remains a well tier three priority at best at most of the departments until there's a major breach. It just doesn't get real leadership attention. Now, I hope I'm wrong about this one. I'd love to see some of our federal agencies have a fighting chance against at least the weakest [00:12:30] of the threat actors lurking out there. Well, that is it for our Blips Bites are coming up, but before we get there, we need to hear about one of our sponsors from our leader. Well, Lou, take it away.

Lou Maresca (00:12:49):
Well, thank you guys. I get your back to your enterprise and IT news in just a moment. But before we do, we do have to think a really great sponsor of this week at Enterprise Tech. And that's a c i Learning. 94% of [00:13:00] CIOs and CISOs agree that attracting and retaining talent is increasingly critical to their roles. Now with today's IT talent shortage, it's actually more important than ever that your team skills be current. 87% of companies say that they have skill gaps in their employees. Although the challenge of assessing your IT staff skills is really an overwhelming thing, but it doesn't have to be. In fact, a C L Learning has partnered with the best in the industry by adding insights. The new skill gap analysis tools is also there to ensure you that the [00:13:30] training you're providing is actually working. Now in a quick one hour assessment by Insights a c I learning will provide your whole team with key diagnostics.

(00:13:40):
So this is solution IT managers have been waiting for. In fact, with Insights you can identify specific skill gaps in your employees and see where your team's weaknesses lies. Empower your team with personalized training 'cause blanket training really wastes money and time insights. Offer detailed solutions [00:14:00] and support and strategy by issuing recommendations and training plans for individuals and your whole team. In fact, compare results against other organizations so you know where you stand and test those skills and close the gaps with Practical Labs that allow trainees to focus on the skills they need most. A c I Learning helps you retain your team and entrust them to thrive while investing in the security of your business. More than 7,200 hours of content are available with new episodes added [00:14:30] daily. In fact, a C L only stomps its competitors with 50% higher completion rates. These are the training solutions your business has been waiting for. Future proof, your team and company with insights from ACI Learning visit go dot ACL learning.com/twit. Twit listeners will receive at least 20% off or as much as 65% off an IT Pro enterprise solution plan. The discount is based on the size of your team, and when you fill out their form, you'll receive a proper [00:15:00] quote tailored to your needs. And we think a c i learning there's support of this week, enterprise Tech. Now back to you guys.

Curt Franklin (00:15:11):
Thanks Lou, we appreciate it. We'll hear from Lou again later on, but before then, we need to talk about something that, well, I don't know that anyone else in the tech industry is talking about. It's this radical new thing called generative ai. Might have heard of it. Chat, G P [00:15:30] T, all of that out at Black Cat and Defcon last week. Pretty much everyone was talking about it and there are just a whole host of issues that these technologies bring up. We'll talk about some of them later on, but right now we've got a bite where the New York Times is gonna help us talk about the legal aspect. Now, several weeks back, New York [00:16:00] Times updated its terms of service. Its t o Ss to prohibit AI companies like Open ai. That's the company behind chat, G P T from scanning its articles and images using them to train AI models.

(00:16:17):
Now, after doing that, it seems like the times may be preparing to sue open ai. And according to some experts, [00:16:30] the results could be devastating to open ai, including the de making them destroy chat GPTs dataset and fines that go up to $150,000 per infringing piece of content. Now, N P R, which was one of the companies that broke the story, spoke to a couple of people with direct knowledge who said that the Times lawyers are mulling whether [00:17:00] a lawsuit might be necessary to protect intellectual property rights. Now this is frankly a long, deep loud echo of some of the lawsuits, some of the issues that have been around since companies like Google started crawling the internet and scraping content to, to put into their search engines. This is especially true when they go [00:17:30] beyond merely putting a link and a title and actually put some of the contents on the page on the search page.

(00:17:38):
There's a very real question about just how much of the material that is on the internet can be used by an automated process that goes out and gets this. Now there are book authors who have been looking at this. There are [00:18:00] all kinds of companies and they say that the Times is primarily worried that Chat g p t could use the New York Times material to become a competitor to the New York Times for, you know, why would you subscribe to the Times? If you could just ask chat g p t to summarize the articles from the Times, which is [00:18:30] something that you could do now chat, G P T is going to probably claim that it's doing using the fair use exemption. This one's gonna get complicated. It really is. Now some companies associated Press is one, have struck deals with open ai. So [00:19:00] they are licensing their content to open ai.

(00:19:03):
They are, we can assume getting some sort of royalty every time some of their content is is crawled. This is an issue for the large language model or generative ai because the larger the sample of data that feeds the model, the more accurate and useful the model becomes. So there is not only an economic, [00:19:30] but a technological incentive to go out and hoover up all the data that's available. There is likewise some real incentive to make sure if you are an original content producer, that that content is not effectively used against you and words aren't. The only thing, one of the major issues in the SAG after [00:20:00] strike is how actors likenesses can be used after they are recorded. Now th this is not a such a huge deal for the Brad Pitts of the world. They have high paid lawyers and agents to negotiate some such a thing, but I've seen several people give the example of an extra in a film who signs a thing saying that as an [00:20:30] extra they can be scanned and their lightness can be used in case they need to recreate them a few dozen times in a crowd scene.

(00:20:40):
But once that image is scanned, how often can it be used? And if that person should become famous later, can the studio just go back to their extras reel, pull the image and use AI to animate it without paying the actor any additional money? Big, [00:21:00] big issues. So I wanna go, Oliver, I just have the strange feeling that this is the sort of thing about which you might have an opinion. So do you think that companies like the Times have a real issue here? Is this something that they should be able to control how their articles are used? Or, you know, is this simply fish wrapper that they shouldn't be worried about? Someone else [00:21:30] using the day after it's published?

Oliver Rist (00:21:34):
Definitely be be worried about it. So even before ai if I'm writing for PC Mag, we would have this issue where these Podunk sites usually in Asia would grab an entire article word for word, stick it on their site, sometimes with my name on it, sometimes with their own. And then we'd have to send 'em a cease and desist. But if they just paraphrase the article and link back, we wouldn't, we wouldn't bother. In fact, we'd like [00:22:00] it. So the question is how is chat G P t surfacing this stuff? If it's giving credit and it's giving links back then fair use really doesn't have that much of a leg to stand on. If they're barfing up the whole thing or you know, a substantial part of it and they're doing it word for word, then you know, no, but it, it gets even more complicated.

(00:22:23):
Like, can they change? I, I mean, I don't know, but like, can they change how the stuff comes up? If it's just used as data for [00:22:30] a query? And it summarizes that that data if they can turn that on and leave the word for word stuff off that a judge is gonna look at, like sure. I mean, it's, it's, that's fair use. That's anything up there. But the N Y t, they have a paywall. And if you've ever been on on on N Y T that paywall comes up awful fast. You have to do, even after you do a a free account, you get maybe 1, 2, 3 articles a month and then pow you gotta pay. So a how did OpenAI [00:23:00] get that? Like, did they get like do a pay account and then just abuse the hell out of it?

(00:23:05):
Or did cheap or did chat G p t like bully its way through which a previous lawsuit I think I talked about the last time that I was here. That was what they were alleging. They were alleging that not only were they grabbing you know, writing content, they were going after email stores after medical records. I don't know what happened to that lawsuit, but if that's all it was doing, basically finding anything that wasn't wasn't nailed down and sucking it up and then using it yeah, that's gonna be, [00:23:30] that's gonna be a big, big no-no. And honestly, I think you're right. I mean, it depends. If they get a nasty judge, they could be dead.

Curt Franklin (00:23:38):
You know, this, this is going to be interesting and I think it will be fascinating to watch this because if a lawsuit against OpenAI is successful, does that mean that they're going to go after Microsoft and, and binging? Does? we know that Google is getting ready to come out with their own instance. That's only are they gonna go after Google? [00:24:00] Yeah. you know, this is, this is one of those cases and we, we hit them fairly often where the technology has run way ahead of our regulatory and legal frameworks. I mean, we we're, we're way off in uncharted territory. Brian, I I'm interested in your take because you and I are both book authors and it's funny that Oliver mentioned the idea of somebody scraping. I remember [00:24:30] when you and I sold the the simplified Mandarin translation of our most recent book.

(00:24:37):
We both agreed that we would either sell one copy or we would retire on the royalties. I think we came a lot closer to selling one copy. You know, what, what do you think as a creator of intellectual property about what should be happening and what you suspect is going to happen [00:25:00] around this use? Well, I, I think it comes down to our friends in Washington DC and actually I've been hearing rumblings that there's been unofficial conversations at the UN about this very topic. It's AI is on a lot of people's minds. Now, what I am gonna do is, I'm gonna go back a little bit back [00:25:30] when Netscape was one of the big powers in the emerging internet. One of my ex-girlfriends was one of their original programmers, and we actually had a conversation that where the heck is all this web browsers information going to come from?

(00:25:55):
And then I had another off the record conversation [00:26:00] with a gentleman that was actually one of the original programmers with Alta Vista, and we had a similar conversation. So this is not a new topic. Where the heck is this information going to come from and how are people going to keep credit? And I'll tell you right now, the federal gov, US federal government doesn't know all kinds [00:26:30] of agencies have had information floating around in the government back channels. It's labeled for government use only that, how the heck do we regulate this and should we regulate this? So I'm going to answer this as I would really, really, really like to have more than a single digit royalty check for [00:27:00] our books. So that's one I'd, I'd like my, our books to not become the public domain because of something like chat G P T, but I also think it's going to be something that's not going to have an answer anytime soon.

(00:27:18):
My suggestion to our viewers, like always is write your congress critters. If you have an opinion, email it, write, [00:27:30] call your congress. Critters are looking for directions. They're trying to figure out how to keep people's intellectual property their own. I'm sure the New York Times case is adding fuel to this fire. I'm reasonably sure the Associated Press licensing deal has muddied the waters. So there's good, there's definitely two sides to this coin [00:28:00] and I am of the opinion that yeah, intellectual property needs to stay intellectual property, but I'm quite fond of web bro web crawlers and getting information that way, being able to find things. And I do, I have played around with chat G P T and the generative art programs like Dolly and so forth, and I think they're a really cool tool, but how do we keep [00:28:30] it so that people can actually make a living?

(00:28:33):
That's my big question and I don't have an answer. Yeah, it's one of these things where some very smart people are working on this. As I said, out at Black Hat and Defcon, people were, were talking about these issues and a whole lot more. This is one where we have the possibility for not only intellectual property issues, [00:29:00] but privacy issues, security issues, sensitive information issues, just all over the place. And what I think a lot of people are both looking forward to and fearing is if there will be regulations promulgated sooner rather than later that tell pe tell the companies how they can use it. Unfortunately, I think everyone is worried that if the regulators, and especially [00:29:30] the lawmakers get their hands on this, any regulations they're going to come up with are going to be wrong. They're going to be less useful than they could be and more harmful than they might be. And oh, by the way, how about those artificial intelligence engines that aren't within the United States? What do we do with those? This one is a big deal and I will be really surprised if [00:30:00] this is the last time we talk about it here on twt. Well, that's gonna do it for this particular bite. We've got a great guest coming up. Before we talk to the great guest, though, we've got to hear from our great regular host Eska telling us about another sponsor of this weekend in Enterprise Tech. Lou, take it away.

Lou Maresca (00:30:24):
Well, thank you guys. I'll get you back to Enterprise in 19 News in just a moment. But before we do, we do have to thank another great sponsor of this week in enterprise [00:30:30] tech, and that's panoptix in the rapidly evolving landscape of cloud security. Cisco panoptix is at the forefront, revolutionizing the way you manage your microservices and workloads with a unified and simplified approach to managing the security of cloud native applications over the entire lifecycle. PANOPTIX simplifies cloud native security by reducing tools, vendors, and complexity by meticulously evaluating them for security threats and vulnerabilities. Panoptix [00:31:00] ensures your applications remain secure and resilient. Panoptic detects security vulnerabilities on the go in development, testing and production environments, including any exploits and open source software. It also protects against known vulnerabilities and container images and configuration drift all while providing runtime policy remediation. As Cisco's comprehensive cloud application security seclusion Pan Optica ensures seamless scalability [00:31:30] across clusters and multi-cloud environments. It offers a unified view through a simplified dashboard experience, reducing operational complexity and fostering collaboration among developers, SREs and SEC OP teams. Take charge of your cloud security and address security issues across your application stack faster. And with precision, embrace panoptix as your trusted partner in securing APIs, serverless functions, containers and [00:32:00] Kubernetes environments, allowing you to transform the way you protect your valuable assets. Learn more about panoptix at panoptix app and we thank PANT for their support of this weekend Enterprise Tech. Back to you guys.

Curt Franklin (00:32:18):
Thanks Lou, we appreciate it. And now it's time to get when we get to what most people consider to be the very best part of this weekend, enterprise Tech. And that's our guest this week. We're gonna be talking [00:32:30] with Dan Lawyer. He is of Lucid Software. Dan, the chief product officer. Dan, welcome to this weekend Enterprise Tech.

Dan Lawyer (00:32:46):
Thank you Kurt. I'm so glad to be here. Appreciate the opportunity.

Curt Franklin (00:32:49):
Well, we appreciate you taking the time to be with us. Now, before we start talking about Lucid, wanna take just a minute to talk about you because we have a wide [00:33:00] range of ages, experience levels, and experiences in the viewership and listenership of this weekend. Enterprise tech, most of them want to know what it was that led you to the chair where you are sitting. Can you tell us just a little bit about the career path that brought you to Lucid?

Dan Lawyer (00:33:24):
Yeah, happy to do that and it's, it's a lot of fun to reflect on that. You know, my, my journey [00:33:30] has been long. I've been in technology for about 30 years now. Started out at Word Perfect, which is a great divider, you know, people who do or do not know what word perfect is right. And then moved on to Novell. I was connecting with Brian actually earlier before this, this show talking about some past history that overlapped with us at Novell and it was at Novell that I first got into product management and was over like fairly highly technical things. You know, I was doing product management for network operating system kernels and, and file [00:34:00] systems and, you know things like that. And then progressing into more consumer facing things over time. You know, my more recent companies prior to Lucid software are places like Vin Smart Home or ancestry.com or Adobe.

(00:34:14):
And so those have been a lot of fun, but it's been, you know, the last 23 years, all product management and increasingly product management plus some other function. You know, so product management plus engineering, product management, product marketing, product management, UX strategy, things like that. But it's been, it is been a really fun ride for me [00:34:30] to, you know, start way back in shrink wrap software and progress up to now where we're, you know, predominantly a SaaS world and at least the world that I live in and to see how technology's evolved. And and some things have remained consistent and some things have really changed.

Curt Franklin (00:34:45):
You know, a lot has changed and it is been interesting. You mentioned Word perfect, I remember Word perfect and, and its predecessors. So it is fascinating to to talk to people who have no idea what [00:35:00] the world was like before these original big word processors came on the scene. Well, you, you know, one of the things that these word processors did was allow us to collaborate, but even collaboration has changed a great deal, and I know that that is what Lucid does. Can can you tell me a little bit about how Lucid [00:35:30] approaches the idea of collaboration? What, what are you trying to help people do?

Dan Lawyer (00:35:36):
Yeah, I'm happy to that, you know, lucid plays in the space of visual collaboration. And I think the fundamental problem that we're trying to help people with really comes down to this, is that when people work together, they quickly discover that written and spoken language is imprecise. And so, you know, for thousands of years, people have looked for ways to bridge that gap. And historically maybe that was, you know, riding on [00:36:00] a cave wall, or more recently is riding on a whiteboard or a napkin or getting into the same room. So we can see each other's body language, some way to augment the conversation visually. Visual collaboration. And Lucy specifically really focuses on how we help teams to see and build the future. And we think about how, how does a team start at an idea and progress along a journey from that idea all the way to reality. And, and how can they do that in a very collaborative, visual way? And the visual aspect of what comes in precision, [00:36:30] it helps teams to get to clarity faster. It helps 'em to get to alignment faster and ultimately it helps them to build a future faster.

Curt Franklin (00:36:37):
Well, building a future faster is, is really interesting because I think it's what all companies are trying to do, but how does Lucid approach collaboration in a way that is fundamentally different than any of the other collaboration programs? You know, collaboration as a category [00:37:00] had been growing for a while, and then during the pandemic it seems to have just exploded. So, so what is Lucid doing that is different than what some of the, the other companies are doing?

Dan Lawyer (00:37:13):
Yeah, I think there's a couple of things there. And, and really during the pandemic, there was immense pressure put on the patterns of how people collaborate. And, and, and there was immediate need for new ways to do it because people were suddenly in different working patterns. Fortunately for Lucid and, and, and, you know, solutions similar to Lucid [00:37:30] we were well positioned and already had solutions that, that knew how to play in this new world. And so I think the thing that make Lucid quite different from other competing offerings in the space it really comes down to two things. One, you know, I mentioned earlier that Lucid really focuses on this complete journey that people take from idea to reality. And Lucid spans and covers more of that journey has more breadth across that journey and our functionality and capability than any other player out there.

(00:37:57):
And so the idea that we can support the collaboration [00:38:00] all the way from the earliest ideas through to where it's like out there, it's real and people are doing it. That's the one thing. The second thing that Lucid is really known for and, and that people love about our platform and our offerings comes down to a, a, you know, an underlying platform that has a set of data and automation capabilities. And so you're not just doing everything on your own on a canvas, but Lucid is doing a lot to help assist and automate that process. And even though you're playing what, you know, feels very freeform on a canvas [00:38:30] that information can be data backed. And so as you are doing things visually on a canvas, it's actually updating the underlying systems of record or updating the other data sets that you need to integrate with other systems or to carry the work process forwarding farther. So it's a combination of breadth of workflow and then the depth and the data and automation capabilities that really set Lucid apart.

Curt Franklin (00:38:52):
Now, lucid is, I'm assuming a browser-based set of software, or do you [00:39:00] have your own clients that companies load?

Dan Lawyer (00:39:03):
Yeah, lucid is predominantly used in a web browser. We also have native applications for iOS and Android devices. And, and really the origins of Lucid started out when one of our co-founders got frustrated with, you know, another similar product on the market, Visio at the time. And his basic thought was this simple, which is like, why can't this be web-based and collaborative? And that was the origin of how, you know, our first product Lucid Chart came to market [00:39:30] was. And so it's always been a SaaS-based, web-based collaborative product from the very beginning. We've then expanded that portfolio over time to, you know, broaden it and introduce beyond Lucidchart things like Lucid Spark and Lucid Scale to cover, you know, a breadth of different use cases that knowledge workers go through.

Curt Franklin (00:39:49):
Yeah. I've got to ask a couple of things. Number one, you've talked about the mobile clients. Do you have a good sense of how frequently people use this on a mobile [00:40:00] device versus using it on the larger screen of, say, a laptop or a desktop or I suppose even a large monitor in a conference room? I, I think we still have conference rooms.

Dan Lawyer (00:40:13):
Yeah, we, we, I, I've heard they still exist. And, and based on you know, requests from some of our largest customers, I think that they're coming back into fashion. We have a lot of customers that are making sure we have strong support for large form factor touch devices, right? So like some huge [00:40:30] monitor, which is touch enabled, which becomes basically like a digital whiteboard inside of a conference room, which we do have and support and increasing support for those devices. But as far as you know, deployment, it's interesting, I think the, the majority of usage is still in a web browser for our products. And, and we see patterns of usage. And, and I think what it comes down to is when people are in a heavy content creation mode, they're often looking for, you know, more screen real estate, more capability you know, with [00:41:00] a larger keyboard, their mouse, a stylists, things like that.

(00:41:04):
We also don't see a lot of that increasingly moving toward tablet size, form factors where people are interacting with the stylist and things like that. And then also on a mobile phone the nature of a lot of the content you know, you're doing a lot of pinching and zooming on a mobile phone if you, if you're looking at, you know, complex diagrams are, are robust you know brainstorming sessions or things like that. And so I think that's part of what drives it, is that the use case, the use, what people are creating [00:41:30] tends to work better on a larger screen, whether that's a tablet or, or a desktop or laptop.

Curt Franklin (00:41:36):
Oh, I know with my eyes, I appreciate every square inch of, of screen real estate I can get. So I understand that we've got a lot more to talk about with our guests, but before we do, we're gonna pause here for another word from a great TTW sponsor, and welcome back, it's time to bring my co-hosts [00:42:00] into the conversation. Oliver Wrist would like to hear the sorts of questions that you have for our guest.

Oliver Rist (00:42:11):
Yeah, so since we were talking about it AI to me, AI and collaboration, I mean, thinking back on all the C R A P problems that I had, collaborating with people digitally to me AI would be a huge boon. I mean, I can look at it in several scenarios and be like, [00:42:30] damn, I wish I had had that. Then. is that on, on, on your roadmap? And if it is like, are you gonna do your own or if you do third party access, like the, you know, the chat GPTs of the world, how would you structure that kind of relationship?

Dan Lawyer (00:42:47):
Yeah thanks Oliver. So we actually have a beta of our first set of collaborative AI features going right now and anticipate later this year that it will be out of beta and fully available to customers. [00:43:00] And that that feature set is built using you know, on the backend chat, G p T would go through Azure and into the Azure services to provide that service. And what the current service is, is you know, an initial flow of how a team can start out with an initial kind of brainstorming ideation session, progress to that session, assisted by the AI synthesize, meaning out of that session assisted by the ai, create a synopsis of it [00:43:30] and turn that synopsis into action. And so it's kind of this like end-to-end workflow of how you might use AI to assist you through a brainstorming session.

(00:43:37):
And that's our first step into it. And, and again, in beta and will be broadly available and not too long from now. The, you know, I think how it plays out over time for us and, and as we've engaged with our customers, we see a lot of interesting patterns, but I, I'll say like the most commonly requested things and things that are very much on our minds, and we actually have early prototypes of [00:44:00] the people could go look at some of these is the ability to either through text or speech, just type or say something and automatically have a diagram or some piece of content created. So using generative AI to do that and we actually created our first chat and G p t plugin that's available out there available today where you can do that.

(00:44:21):
You can, you know, feed it some entry and it will create, create a diagram for you based on the entry. It's been a lot of fun to play with that. And, and you can imagine that, [00:44:30] you know, working its way back into our products mainstream, where that's just part of how you use our products. The second thing that people have really wanted in our products is the ability to have AI almost start to become the interface and the assistant similar to what Copilot is trying to do, I think, but to where you can say something or write something, and I no longer have to be the expert at finding features in the product or whatever. I just have to have an idea of this capability probably exists, and so I'm gonna ask Lucid to do it for me and expect that it will be able to do that.

(00:45:00):
[00:45:00] And, and I think that type of interface is really compelling to people and is probably what the future looks like for a product like Lucid is, you know, the ability to just talk or type to it, and it's the expert, it has the skill, and it does things for you that you want, and then you engage in a conversation and it, and updates and fix it and refines it for you. The, the structure of that gets even better, though. If you can take, you know, these public LLMs and combine them with private information from that company, it'll enrich the context. And, and that's [00:45:30] really one of the core conversations we're having with a lot of customers is, and, and it really goes back to the conversation earlier on the show around the legality of these things and the lawsuits and things like that, which is, you know, every customer I talk to, they want the promise of ai, they want the value of it. Nobody wants their, you know, their IP or the value that they've created to be given away without their permission. And, and so it's find that balance of like, like, how can you, you know, put the two worlds together in a way that protects the IP of a customer, [00:46:00] keeps everything safe and secure for them, but still provides this amazing value of, you know, higher automation and capability for our, our customers. And, and I think that's, you know, that's what we're working on right now is figuring out that balance point.

Curt Franklin (00:46:15):
Hey, Dan, that actually gets me a nice segue. I've got a question about APIs and your private LLMs. You know, a lot of corporations, a lot of organizations create their own knowledge [00:46:30] base. How does one do that? You know, 'cause every once in a while a lot of conversations and different types of organizations are eventually feed into a help desk app or pull from a help desk app and then eventually populate knowledge bases. But one of my biggest, biggest complaints, I I used to work for Xerox, and one of my biggest complaints is [00:47:00] not having the help desk tied into institutional knowledge. Do you see this as something lucid could start solving eventually?

Dan Lawyer (00:47:10):
I I think Lucid certainly will be part of that solution. And, and the reason I think that is, is today as our customers use Lucid and they create content and they collaborate together, the end result of that, often, like, if you think of it from a certain perspective, the the result of that is often becomes a blueprint of their business, a blueprint of how their business operates, right? They're [00:47:30] documenting and diagramming the business processes or the technical workflows, or the stack or the, like, any number of things about how their business runs is likely documented inside of Lucid someplace, including their original ideas of how they even came up with what they should do and how they should modify it and do it over time. So a lot of that content exists in Lucid and, and so I think that can become available, you know, through an ai to the customers.

(00:47:54):
And, and you can imagine like the security model that need to be overlayed on top of that, so that depending on who the user is, they have access [00:48:00] to the appropriate information. That gets more robust though, if you can take that and stitch it together with other content about the company, right? Like the, the Lucid doesn't have all of the content about the company. We have a lot of information about how the company works and, and that could be used, but there's a bunch of other information. There's across all types of other formats. And, and so I think ultimately you'll probably see the world start to converge toward where there's, you know, maybe graphs of these types of documents and, [00:48:30] and files that get stitched together that become available to an ai. And, and then there's a strong security model at the top to make sure that nobody ever gets access through the AI to things that they shouldn't get. And that gives you a really compelling view into a business to where I don't, I, it is, it's less of me like hunting and pecking through a knowledge base to me, like asking normal questions. The AI understands a lot about my context, the teams I'm on, the projects I'm working on, and can pull back really, really relevant answers to those questions.

Curt Franklin (00:49:00):
[00:49:00] All right. So if I wanna try and bring it in, so this doesn't sound like a system that's going to be driven from someone in the trenches, this is probably something that's gonna be driven from at least a director level. What kinds of thoughts, what kinds of key note points should we be documenting? Should we be getting ready in order to justify making this investment? Because it sounds [00:49:30] like an investment. This, you know, this is obviously not a copy of Visio, right? This is something quite a bit more, and it probably doesn't get licensed in the same category as Vizio. So what are the decisions?

Dan Lawyer (00:49:44):
Well, it's, it's clear. I mean, there's a lot of companies forming policy right now to start with one, when they're, they're excited about what's coming. They're also very worried about making a misstep. They're worried about doing something. Or especially when you deal with companies with lots and lots [00:50:00] of employees, how do you make sure that every single employee is well informed of, of what they should or shouldn't do in terms of ai? So there's a lot of policy. I think you'll see then you know, software solutions to try and enforce that policy. A lot of the technology needed to do that exists today, right? Like whether that, you know, at multiple layers of like a technology stack to try and do that enforcement. But there will be, you know, new things on the radar. I I, I think it starts with that, which is like, what's our policy?

(00:50:26):
Then? It's, then it's starting to think about. And if you don't have like really strong [00:50:30] role-based understandings of, of you know, what people in your company are doing and, and what based on their role they should have access to, well, that's a place to start cleaning up it. You're gonna want to be able to think about based on role, what people should be allowed to do and what they should have access to if that's in place, I think that that transition to where there's some underlying graph with all this information in it becomes pretty smooth, right? You plug it in that understanding of rules, you plug it into policy and, and then, you know, select the graph system that you want to use. Which [00:51:00] by the way, I don't think you'll have to build yourself, right? There'll be many people will offer a solution that way. And then, you know, you that comes together and, and you're probably up and going now, you know, the time horizon on that. That's, that's not a six months from now thing, but at the rate things are going May, is that a 12 month? Is that an 18 months from now thing? Yeah, I I think it'll come faster than we think. Yeah,

Oliver Rist (00:51:21):
I did have one more. So I'm a contractor most of the time work for many different clients. So teams Slack, [00:51:30] you know, 'cause I have gray hair maybe someday a long time ago. Lotus Notes I have not used lucid. However, that's what I kind of wanted to walk through. So like the teams guys or the Microsoft guys have a really compelling, you know, Microsoft 365 play where their AI will put stuff from your meeting everywhere. How does that work in a lucid environment? Like, like if I have a meeting, like, walk me through, like what happens when I'm using other apps to, you know turn my, I my, my ideation [00:52:00] in a, into something real? How does that like work?

Dan Lawyer (00:52:03):
Yeah, so the, the way it works is we take an approach of integrating with a lot of these main systems. So with the exception of Lotus Notes everything you named we are integrated with. And so they played very nicely together, and you can go back and forth smoothly between those worlds. In fact, specific to Microsoft Teams at, at the recent Microsoft Build Conference we announced with them our initial integration into copilot with our AI and the ability to, you know, when you're in Microsoft Teams to use copilot [00:52:30] to you know, point to some lucid document or lucid board and ask for a summary in the teams environment, it'll create a summary and give that that back to you. That's the first of many things I think we'll do together to integrate the AI and our products generally, the teams environment.

(00:52:44):
We we're integrated with all of the Microsoft Office products that, that most people are using today as well as, you know, a broad range of other products. And so if, if there's a place where you're collaborating today or where you are communicating something like Slack, [00:53:00] things like that we're likely already integrated there to where you can use Lucid in the way that that gives you the most value. And then that content can show up in the work streams other places, or you can start in those other places and come into Lucid. So it all plays very nicely together, back and forth.

Oliver Rist (00:53:17):
Cool.

Curt Franklin (00:53:19):
Jump in and ask one final question. Human resource onboarding is the giant bugaboo for human resource people. And [00:53:30] in my moving around the onboarding process in every single collaboration tool I've ever used has been clumsy. Is this something that Lucid been working on?

Dan Lawyer (00:53:46):
Yeah. thank you Brian, for asking that. It's, it's a, I think a point of pride actually for Lucid. When we look at our N P s surveys that we do, the most frequent verbatim response from our users of [00:54:00] why they love Lucid is because of its ease of use. And we have constant investment into keeping it easy to use and making it easy to onboard and, and making it easy to go faster. But that, that's, that's one of the reason, reasons that we've been picking up momentum and winning in the market is because it is easy. And, and I think our, our business model is dependent on it. Our, our business model has a very heavy you know B two C acquisition where we're getting millions of users coming into the platform independently on their own. [00:54:30] And to do it at that scale, you have to be able to have somebody that you've, you know, pull in through SS e o or s e m off the internet, be able to come in and on their own get started.

(00:54:38):
And so we, we, from the beginning, we paid really strong attention to that ease of use and how people onboard. And then people, you know once you get many users inside of account, then it tends to consolidate and they want a higher level features, they want more IT control, and so it goes up and then it becomes more of a B two P transaction. But it's, it's those strong roots and, you know, product-led growth, those strong roots, and knowing that we have to be [00:55:00] able to survive, you know, millions of people coming in independently and, and onboarding themselves into the product, that, that have created a constant focus for us on how easy it's to do that.

Curt Franklin (00:55:12):
Well, we're getting down close to the end. I just had a couple of quick questions, and one of them is that so many collaboration systems are great, but they require a massive team for their deployment care and feeding [00:55:30] for yours. Is your service something that's applicable only to massive enterprises, or is it accessible to some smaller companies?

Dan Lawyer (00:55:41):
Yeah, our product scales from one to tens of thousands of users. We have deployments across that whole range today. And, and the, you know, models for how you acquire Deploy it, administer it, scale across that full range. And so we're in, you [00:56:00] know, some of the largest companies you can think of today at full scale with tens of thousands of users. And, you know, we're on one person's laptop you know, using it as a SaaS service independently. And so we have a full range of ability that way.

Curt Franklin (00:56:14):
Well, given that wide range, how would someone take the first step if they were interested in exploring the potential of what Lucid has to offer? How would they get started?

Dan Lawyer (00:56:29):
The easiest way to get started [00:56:30] is to just go to a web browser and go to lucid.com. Or if you're looking for a really specific app, you could go, you know, you can also type in Lucid Chart or Lucid Spark or Lucid Scale and get to our individual apps that way. But any of those paths, you know, lucid.co will get you there as a starting point, and then you can go and discover more.

Curt Franklin (00:56:50):
That is super. Well, and we thank you for being with us today. We [00:57:00] appreciate your time your expertise. Any final words for the TWI Riot, all of our listeners?

Dan Lawyer (00:57:10):
I, I just really appreciate Kurt Oliver O'Brien letting me be part of the conversation today. It's been fun to participate and to, to listen in. And, you know, I'm just excited about where this world is going. I'm excited about the amazing things that are happening with ai, whether it's on Luc's platforms or other places. And, and also like [00:57:30] really curious to see how this plays out. There's big questions ahead for us all as this plays out. I, I'm, I'm an optimist about it, but I also, I'm a realist about, there are some, there's some hard questions to, to solve

Curt Franklin (00:57:42):
Lots of questions. Glad that we can play a role in helping folks find answers. Good on you. Well, we've been talking with Mr. Dan lawyer, chief Product Officer, lucid Software, and we've been doing that [00:58:00] because we've been bringing this week in enterprise tech to you, our listener. You're, the reason we do this and the crew that has done it for you this week includes me and my co-host. First, want to go to Oliver Wrist, Oliver, what's coming up with you for the coming week?

Oliver Rist (00:58:24):
So right now, I just finished talking to Maserati about their, their strategy for ev, which [00:58:30] I wasn't expecting. I was expecting to talk about a vehicle, but no, they're gonna redo their entire line. So that story will be in worth if it isn't already. And after that, I'm about to publish my first book in simplified Mandarin. It's called how did Brian Get a Girlfriend Before Kathy?

Curt Franklin (00:58:49):
Gripping Thrilling Reading. We're all looking forward to that. Speaking of which, we go to the man himself, our co-host, and our producer, the [00:59:00] man in charge of bringing us guests like our this week's guest Mr. Brian Chi. Brian, what's going on with you in your part of the world this week? Well, I'm gonna answer Oliver's question first. That girlfriend happened to be one, the very, very early programmers. I went to school with her at the University of Hawaii, and she was actually one of the programmers early [00:59:30] enough that she had an Easter egg in Netscape. So that was kind of cool. Anyway yeah, I've been getting comments from different people. Some of our, some of our episodes do get heavy on the marketing. I'm hoping the people that didn't like one of our episodes that went in that direction are happier with today.

(00:59:54):
I think Dan did a wonderful job talking about some of the policies and decision making and [01:00:00] why you might want to go to a more sophisticated collaboration environment. But, you know, we'd really like to hear from you. Questions, comments? I've got pretty thick skin, so you're welcome. I do a lot on Twitter I still use. It's even now it's called X. And I'm a D V N E T L A B advanced net Lab. You are also welcome to throw email. I'm seabert spelled C [01:00:30] H e e B e R t at twit tv. You're also welcome to use twit at twit tv, which will hit all the hosts. Would love to hear from you and, you know, show suggestions are always welcome. Take care everybody. Thanks to Oliver and O'Brien. And over in the control room. We also have Mr. Victor this week doing our online [01:01:00] TD and production stuff. Victor, what's, what's going on in your part of the world? We, we know you're back there. We can hear you breathing,

Victor Bognot (01:01:09):
<Laugh>. Yeah, thanks Kurt. I, I just wanna say like one of the for Club Twit members one of the other shows I work on is called Home Theater Geeks with Scott Wilkinson. And this week and last week he did an interview with two [01:01:30] t h X guys, one of them the c e o and another one a director of product development about spatial audio. So really good interview. The first one is Free to the Public. So even if you're not a Club Twit member, you can check that out. But it's a really good interview. So if you haven't listened to Home Theater Geeks in a while here's a good reminder how, how geeky Scott can get with Home Audio

Curt Franklin (01:01:58):
<Laugh>. [01:02:00] Oh, we are glad you are there doing that. We'll look forward to checking that out. And real quickly, if people want to find out what I saw at Black Hat and Defcon, I've got some articles coming out over at Dark Reading also gonna have at least one that's up on omnia.com and I'm taking the paywall off for that, so I'll be pointing that in social media. Anybody will be able to read that. Let [01:02:30] me know if you were there. I saw some members of the TWIT Riot and would love to catch up again. Well, that's it for us this week. Thanks for spending another hour with us. And remember, if you want to keep up with what's hip hot and important in the world of enterprise technology, just keep TWIET