How Generative AI Is Already Changing The Communications Field — With EZ Newswire
Channel: Alex Kantrowitz
Published at: 2023-07-24
YouTube video id: ZJBi0DyZp8k
Source: https://www.youtube.com/watch?v=ZJBi0DyZp8k
we're joined today by Caitlin Kelly and Neil Shaw they are the co-founders of EZ newswire Caitlin and Neil welcome hi Alex thanks Alex thanks for coming on so for for viewers uh it's we're gonna switch it up a little bit we have a little bit of a different format today on the channel today we are debuting our new partner chat format when I see a company that I think would be a good partner to bring on to you guys a little bit outside of our typical show I'll select them they'll come on we'll talk a little bit about the industry and we'll talk about their product and that's what we're going to do today so let me kind of explain why I thought this was an important discussion for us to have we've heard so much about generative AI Theory where this is going to go how this might be applicable in certain industries you know is it going to upend work as we know it but very few companies are actually applying this in products today except for like the general use chat Bots like chat GPT and then maybe a little bit in Enterprise but what if you baked it into a new product and actually build something completely new with generative AI at the core that's what these two have done so let's just start uh with the question for both of you which is you know what do you think the the General application of of generative AI is going to be like is it are we all just going to be like living in chat Bots or is this going to be something deeper uh that people aren't quite expecting it uh sure I'm happy to take that look I think um I think generative Ai and large language models are going to make a lot of tasks easier and more productive I think it's going to increase the pace at which development will happen I don't think it's a solve and a Panacea for all problems it's really the application of the of the large language model and how it's trained that's going to make the difference yeah and Alex one of the things that we talk about a lot here specifically in Industries like Communications which we'll get into a little bit um that haven't seen a ton of innovation um we think there's a ton of upside um and it behooves the companies that are building these products to make the experience really seamless really easy to use and really beautiful so it feels like a natural extension of your workflow not living inside of a chatbot yeah so Caitlyn you come from the comms function so it is it does seem like this is going to be you know one of the places where we start to see this you know quicker than ever before so just kind of talk a little bit about from your perspective how this might I mean I you know I'll just say as a journalist like we get a lot of pitches so like are people going to be able to like Stand Out pitching this way or is it going to be a completely different use case that people aren't anticipating so my first reaction to that question is that I I don't think that um I don't think that a pitch necessarily is going to be able to stand out amongst a sea of good pitches that's not going to be the outcome but what I'm excited about in terms of outcomes is up leveling skills and capabilities and competencies for the average practitioner um and so what you can do as an agency or as a inside of an organization instead of a comms organization at any company you can really perform so much better so much more efficient uh efficiently um than you would have been able to do before yeah why don't you take us through your journey seeing something like this and going from oh that's pretty cool to like oh that might be something that underlies different companies uh sure I mean I'll give you an example but I know I prior to to launching easy newswire I spent 15 years in private equity and as an investor the hedge fund and I think it probably took me a good 10 years to develop the skill set of um of really being like an Excel junkie I was very proud of how quickly I could build a model how accurate it was all the very complicated functions and formula that I knew now you can plug something into um to open Ai and you can develop a model in minutes so that's taking probably something like 10 years of work and being able to replicate it and maybe not the perfect capacity but you can get pretty darn close in in you know in fraction of the time it means that if you're evaluating a DDO it's going to be far easier for you to evaluate it with less resources it certainly levels a playing field from you know um from uh in a second third year analyst who might not have the same experience if someone's been doing it for 10 years to be able to at least look at um prospective opportunities and uh and I um more constructive way so it's the way I look at it really is it's you know we kind of lived in a world of screwdrivers and now someone shows up with a power drill but it's the application that really does matter right because in some instances if you're hanging a picture on a wall you don't need a power drill you end up putting hole in the wall but if you know how to harness that power drill effectively you're going to get to the right answer faster right and so speaking of which like there's an the press release seems like a proper place to start and there's obviously a whole ecosystem that's built around it which is actually kind of garbage that we're going to talk about um and maybe that's the thing that you guys are trying to fix I won't give it away for you but um you know how does this and uh what role does the press release play for companies and then how could generative AI actually help improve what we're seeing today in the market yeah so the way we think about a press release is that it's not um necessarily some a document that we are meant to flip into an email and in a date journalist inboxes with um what the press release really is is a document of record for any organization to put a stake in the ground about their own momentum it can be used as an internal motivating tool just as much as it can be for an external or for a journalist facing pitch um so we're really excited about um enabling more companies to talk about their momentum in a way that's standardized and formalized in this uh in this sort of format or structure um in terms of what we're doing with this um uh product in terms of producing these press releases is um creating a better experience around it um historically this has been one of of lots of different editors um lots of people different people involved lots of 3 A.M nights sort of cramming out this content and what we see this um could potentially be is something that's much more streamlined much more easy to use um and something that can really elevate your brand yeah the the use of um uh large language models and AI in our product is is only really it's only a it's like a hook you know it's it we're by using it in the way that we are we're making the press release more accessible more companies uh smaller companies especially ones that don't have thousands of dollars to throw a PR to copywriters Etc uh can now create this professional content in in seconds and for many it's just a starting point we give them a really well thought out foundation and they make edits to it and then personalize it as they want to but it's certainly leveling the playing field in an industry where the cost and the barriers to entry have been quite High right so there's there's what I'm hearing is there's two sides of your business like one is the drafting of this document the press release but then there's a whole other side of of where do you publish it so maybe this is a good moment to dive into oh to dive into either the product or to talk a little bit more about how you're going to challenge the existing power structures that exist in this industry yeah I'm happy to speak a little bit to the products because as a practitioner I've been using these products for uh over a decade and what I found is that the experience is clunky uh it's largely um reliant on Legacy distribution um lists that are lacking any semblance of transparency and so our mission from day one was to make the process easy and simple and clear right transparent um and so that's what we're doing with the UI our promise to customers is not only can you create a press release in a matter of seconds um you can also distribute it to the Publications that are most meaningful to your business and that quite frankly is a game changer yeah I want to get back to some of the um the things that you mentioned here but maybe this is a good opportunity for us to actually take a look at the product yeah sure I can um I'll share my screen and quickly um give a little video demo of um an example that we put together earlier today this is an example that's based on Alex's new um product launch for this new this new video series um so it's pretty straightforward you add your your media contact information this is the person who you would want a journalist to contact or any sort of external party to contact if they wanted to write or learn more about your news you add in a couple of sentences here we've drafted this for Alex and and pop it into our UI we make this super easy to use we give you some pointers on on what constitutes a good prompt um because the better obviously the better information you input the better the outcome will be um and then you hit create it takes a couple of seconds Neil do you want to speak to the prompt and yeah look the you know the power of the large language models is really their generality and however for highly technical applications like writing a press release uh you really need to train that model and ask the right questions uh and um and and provide key inputs necessary so you don't have what's known as hallucinations and the output or things that are made up and so forth so what we've done is really codified our operational experience through Decades of operational experiences in this in writing press releases and Communications in and translated that into a process and a engineering exercise to really leverage large language models in the best way so that the content that is created is professional it's accurate and it's free of any you know uh unwanted improvisations so what they you what we just ran through with that demo while he was talking was we created the press release um the the AI piece of it the actual um engineering of the press release took a matter of seconds um we have it in a in a very traditional text editor box so you can go in there and add your headshots and and add quotes add links make it fancy and uh you know presentable to everyone um and then you publish it so we created a um you know our wire is called easy newswire our first media partnership was with the Associated Press um they as you know are doing really interesting things with open Ai and I think that they'll be um uh open and willing to partner with other technology companies um and now we're working with Publishers across the country to help distribute business announcements that are relevant to their communities news so that means that a publisher in Charleston South Carolina for instance is now going to run uh business announcements and press releases from Easy newswire that are relevant to Their audience um and we're really excited about that yeah yeah go ahead and try to go back to the video uh but to show how you know the process is very quick right um there was a point in the process where the user gets to decide where they want it to be published and that is we've also demystified that process for PR the existing providers uh their checkout process or their selection process is really meant for practitioners and is lidgered with it's a Chinese manual littered with the options it's not transparent on pricing it's not transparent on placement and so forth by partnering with Publications and guaranteeing placement we've given that power back to the consumer to let them create their own shopping cart and decide exactly where they want their news to be placed yeah there's so many sides of this there's the creation there's the selection and there's the actual publication and payment um so let's talk a little bit about creation about uh what it would be like to you know build a press release in here so I watched that video as it went on that was pretty accurate in terms of what I was doing I mean at a glance um but it looked like legitimately like big technology it does newsletter and podcasts and starting this new partner program that was like a very good uh press release so Caitlyn just talk about it from like the eye of a practitioner here um you said many sleepless nights and and lots of sweat that went into you know I guess putting together the press release and then sending it out so talk a little bit about how this generative AI can actually change the way that practitioners make this happen it's huge I mean for me as a practitioner I've been doing this for years and years I started off early in my career and I would I would be responsible for that first draft right um and then it would go up sort of the totem pole and it would touch every single person on that hierarchy all the way up to the CEO so if you add that up the accredited you know time spent on that one document is actually massive and that equates to a lot of dollars spent for any given company um and so what we're so excited about is not just the fact that our technology is enabling companies like the ones I've worked for before to have a bigger faster Head Start to get this content created and published but also that we're enabling a whole category of companies that have never ever known that this was possible for them that's so exciting to me to be able to walk talk and and work with the small to medium-sized businesses that we're talking to every single day to demystify as Neil said this process which they've never felt like they had access to yeah so what does it take to be a newswire like the competitors that you're going against are like PR newswire so by the way like as a reporter in my inbox when I get a press release like people be like here's our news you can go to PR newswire to see the press release and like one of the things I always wondered was like well how did they become a newswire like did they just need a website like what does it take to become a wire well the wire actually dates back to 1846. the prehistoric times um it's a Jurassic era practically but it was really the the foundation of the wire was really laying the telegraph cable between New York and uh and London and that was the Genesis of creating uh the transfer of that news content and making it more efficient and it's actually incredibly interesting because the Associated Press in fact was one of those first companies to help distribute that news from Europe and the rest of the world now today what that means is that we have an RSS feed that feeds directly into the cms's of many of these Publishers um and that's how that information is transmitted that's pretty cool okay so you don't need to build that like transcontinental wire to do what you're doing right but the the ones that are incumbent right now it's they seem to have sort of taken advantage of a bit of uh you know a competition a a position that sort of insulates themselves from competition and uh maybe acted as if you need to build a wire legitimately to challenge them right well distribution is an incredibly hard thing to build right um so I'm not trying to undermine the complexity and the work that those companies have dedicated to building that Network um but like with all Industries Ai and technology is changing how things are done and we hope that um we will have a better way of attacking some of those distribution endpoints as they call them in the business um to get our customers in the hands of the media companies that matter to them a little bit faster yeah so what's your pitch when you go to these media companies and say hey why don't you work with us I mean as I'm I'd imagine it's kind of a tougher pitch because I you know having worked inside some newsrooms like in media companies they're like fairly slow and like they tend to really drag their feed on Innovation and you're coming to them saying hey not only are we building you know a new new newswire but we're going to use AI in the in the creation process what are those conversations like I do right well look we the conversation with Publishers actually um uh relatively straightforward we provide them three things Revenue content and advertising leads so they like us we give them three things uh we also do all the heavy lifting we have a great technical team we work closely with the Publishers to make sure that the integration on the technology side is seamless um on the topic of AI we make it pretty clear that look this is and and to be transparent about half of our users uh leverage our AI writing wizard the other half they have right the other half have their own content ready to go and they really just want better uh and more uh transparent distribution uh but the we make it clear to our our customers and the Publishers that this content is really a draft and the customer can modify it personalize it in any way they want and we also have we and and on top of that Alex we have a lot of filters in place to check for the um safe brand safety of a of a um of a piece of content to make sure that it fits the narrative or fits the um the voice of the publication and so forth and just in terms of the appetite of the Publishers um I think that things are changing so rapidly that even two years ago when you worked inside of some of these larger newsrooms um there is a bigger appetite and more um fluency around some of these Technologies than ever before um folks are willing to give things a shot it's just a matter of how much lift does it take from their side um so when we come to the table and offer to do a lot of that heavy lifting it's a it's a much easier way to get to a yes that's cool and so what is like a pricing comparison like let's say I wanted to put a press release on one of your competitors I mean I you you did show in the video at uh you know checkout screen that seemed you know far more affordable than other options that I've heard about yeah so um we are more affordable than many of the players on the market today we're also more transparent about our pricing um one of the things that I encountered as a practitioner was that there's not um a lot of upfront communication around how much things cost and that's a problem um I've I can tell you anecdotally that I've been hit with an invoice in the thousands of dollars for very early stage companies that do not have the funds to support an expense like that um and so after the fact that's coming in after you oh geez so there's an AP process with many of the the Legacy services and and that's worked for them and that's fine but for earlier stage companies for smaller companies that don't have processes or funds bank account sizes like that um it's really an issue and so we've been very committed from day one to make sure that the price point is one that is palatable for the majority of companies this is not a one percent product we are a product for the others right for the 99 um and we've been really committed to that um and we have that conversation with Publishers on day one to make sure that they understand who they're serving they are serving the small businesses that are operating in their market and they understand that and want to work with us because a healthier business ecosystem in their Market benefits them long term totally so what would a price look like for you guys it starts as low as 25 we actually don't charge for writing the content the the writing the AI writing wizard is complimentary uh you only pay really for the endpoints that you want guaranteed placement in so those those endpoints we work together with the Publishers to develop pricing that makes sense but on average Alex we're 90 cheaper than um than the incumbents so when a press release from Easy newswire goes to a publisher does it go to like a certain section is it like a release section and let's say it goes to the AP for instance where like does that get distributed on the AP newswire where do I find them it does so the we have a direct partnership with the AP that's actually the first partnership that we did um and and so does uh some of the Legacy players uh but the AP has a specific section on their site for press releases and it's segmented by their partners that they work with including easy news wire so if you go to their site you'll see it um in addition to that we work with the AP to distribute it now it's not guaranteed but often um you know the average pickup is around five to six links with about three thousand member newsrooms that subscribe to the ap's feed across the country and that's everyone from like the New York Times comes down to um you know a a local publication to a college newspaper that's cool so how long have you been in market for we launched we came out of beta and launched um about two months ago and we've been working yeah we've been we launched our Alpha in um in August of last year we've been working on this now for about three and a half years I'm just curious what the feedback has been from clients who've used the generative stuff what are they telling you so far 80 of our clients when we ask them we have a feedback survey that we we asked the client customers to fill out um 75 to 80 have rated our content um creation a five out of five uh which means that there's minimal or no changes um uh but we're making changes to uh to our uh to our models our parameters every single day so we hope to get that number closer to 90 but you're right it's a pretty good number yeah what models are you building on like are you building on open AI models or we use we use uh we do use gpt4 which is an a open AI product and there's a there's been a couple of new um releases from some of the competitors out there that we are um we're also uh involving into our Tech stack it really depends on the use case and that's a great thing is that we can really leverage off the tens of billions of dollars invested in in uh in AI over the past decade in order to really fine-tune and make our uh our content better but for any single you know you know for any single release we're typically calling our models about 20 to 25 times it's very different than if you just took you know uh that same piece of content that input and went to for example chat GPT which is more of a One-Shot model uh and got you know really one uh one output we're really iterating behind the scenes uh so that the quality of that content is professional and like I mentioned before there are very few hallucinations I want to ask both you you come at this from very interesting backgrounds I mean Caitlyn being a practitioner to decide to go to build a tech company talk a little bit about your journey there if I would have known how hard it was before I started I never would have done it um but no I mean I talked to folks about this a lot it's the most it's the hardest and most gratifying thing I've ever done um but I do think that having this practitioner this like first-hand operational experience has made all the difference um we've lived in the shoes of folks trying to um create a narrative trying to explain and express and um Pitch that narrative to media and journalist um and so it's incredibly powerful in terms of how that manifests in the product experience um and it certainly made a difference to the experience of our of our end customers yeah Neil look when I speak with investors they always say I got a good good life uh you know they get to evaluate different companies and investment opportunities and and uh they get to chill so what made you trade that in for the entrepreneur lifestyle yeah well you know I actually started my first software company when I was pretty young and um I I exited it when I was a sophomore in college and I I was awarded under grad so I've kind of got sucked into that the the um the High Finance um Vortex so you know really that 15-year journey in finance was uh was my sabbatical and then I returned I would say back to being an entrepreneur it's you know like the world of being an entrepreneur in the world of of startups is an incredibly difficult path so you have to some someone be a glutton for punishment and like The Daily Grind that comes with it yeah oh I hear you on that 100 any did I miss anything or is there anything else that viewers should should think about when they are considering this or thinking about this story um the thing and I think this is more of a question for you or just like to turn that like um interview there is sort of a um there is a lot of questions or a lot of conversation around journalism and Ai and how AI is sort of impacting and influencing journalism and one of the things that we didn't talk a lot about during this call was that the content that we're creating and Publishing in media companies at you know the AP for instance or any of the other media companies that we work with is clearly marked as sponsored content it is paid content just as this conversation is paid with you we are paying these Publishers to run this content and it's marked as such um and so we do get a lot of questions around the distinction between editorial and paid and how that line might be blurring and so there's not much there's I don't you know I don't really have a point or a question but it is a it's something that people are talking about a lot um and I just you know just sort of wanted to open up the conversation that none of this technology none of this content is ever going to replace the work of well-researched Journalism yeah no I'm with you yeah all right so no I just think it's an important thing to to mention or just keep mind of as we're especially we think about it every day um but in terms of how we work with publishers it's always on a commercial basis it's always a revenue conversation um and then if there's upside and in an editorial perspective that's great for the brand definitely and yeah some people have said that the best ads are content and if you're that what I like about your targeting system is you're giving the person who's putting the release out some level of control to be able to find Publications that fit and that that is a real service because oftentimes they'll just show up in places that they don't belong and when those can get better targeted it works for everybody the the um the company doing the release and the publisher and the reader yeah yeah so well listen we really enjoyed it this has been really fun as as we've all said this is a Shakedown business and we're trying to be eight anti-shakedown as well said to me before so um thank you for having us in this conversation we're excited to talk to some of anybody who wants to continue the conversation about it and um we'll see how it goes thanks for being here I really appreciate you helping me pilot this new series couldn't think of a better company to do it especially this is pretty newsworthy stuff very cool so just shout out the URL or the way to get in touch with you before we we go I'll put it in the show notes but for those that are watching right now what should they do yeah it's uh it's easy so letters easy newswire.com that's our uh that's our URL uh and you could reach us at you can reach me at NS EZ newswire.com and CK newswire.com or hello.com and they'll get to us amazing well Caitlin Neil thank you so much for coming on for opening my eyes and our viewers eyes to a really interesting new area and a practical finally a practical application of generative AI we're here and we're just gonna started all right thank you so much