AI Content Is Now Undetectable Without AI — Max Eisendrath, Red Flag AI

Oh, I mean there's no way people are going to on their own be able to tell. Probably even even now. I think it's I think reaching a point without like advanced types of AI. You need a a comparable detection to tell cuz some of it is it's just so good. >> Hi, my name is Demetri Bonichi and I'm a content creator, agency owner, and AI enthusiast. You're listening to the AI agents podcast brought to you by Jot Form and featuring our very own CEO and founder Idakin Tank. This is the show where artificial intelligence meets innovation, productivity, and the tools shaping the future of work. Enjoy the show. Hello and welcome back to another episode of the AI agents podcast. I'm here with Max Eisendrath, the founder and CEO of Red Flag AI. How you doing today, Max? >> Hey, Demetri. Good to be with

you. Good to be chatting. Um, so first of all, tell us a little bit about how you got into the world of AI. Um, and tell us also what got you interested in making Red Flag. >> Sure. Well, it's a long story that I won't bore you with all the details, but suffice it to say, I've been working in software development and AI related use cases for uh over eight years now. Originally it was focused on sentiment analysis and kind of largecale social listening projects. So crawling the web at scale and for specific pieces of content. And that morphed somewhat naturally into content protection which are focus at at Red Flag these days. And the AI originally was to handle efficiency in crawling and content detection for the business cases we were dealing with. But more and more we're dealing with AI generated content, separating

that out from uh original content, nonAI generated content uh and and validating the uh antipiriracy results that our our system is finding. So it's it's taken on a bunch of different meanings which is pretty interesting. So, ranging from core efficiencies to uh what we're actually looking to detect at at Red Flag today. >> What would you say is kind of the uh the thing that sparked your interest in that area specifically? >> Well, I mean, it's it's always interesting to try and do stuff at scale in an efficient and and cheap way. I mean, that's kind of a classic problem in software, as I'm sure you know. So that was originally kind of the the main use case is, you know, how can we accomplish something that's been sought after for a long time but really just didn't scale in a practical way. So people

weren't able to to really do it through typical kind of SAS applications and and software. There's a lot of manual intervention still required for some of these use cases until very recently. Uh so that was kind of the initial interest. Uh and then now at this point the problems of authenticity and proving out the origination and ownership of content is incredibly important and it's only going to become more of a problem. So it's an exciting area to work in right now. >> Um did you have a background in in content at all? Uh creating it or reviewing it or >> I just watching a lot of it. >> Okay. All right. >> No, not creating it. No, no, definitely came at it more from the the technical side. So, it's kind of a an interesting area to learn about, too, talking to a lot of

creators and publishers and uh yeah, content owners these days. It's good education. >> Yeah. No, for sure. What do you think is the uh main uh issue right now with the with the world of content? Seems like you talked a little bit about how there's uh some some decently uh you know, you kind of alluded to the AI content. tell give people a little bit of background who maybe don't know about that situation and and how it's >> uh causing these copyright issues. >> Sure. So we started working on content protection almost four years ago now and the original focus and even at that time it was more kind of classic antipiriracy is everyone generally thinks about it you know movies software music not [clears throat] being leaked and not being shared in places where it shouldn't be and this kind of endless whack-a-ole game

of trying to remove it and having it pop back up and securing the content. And so automating that process and and making it as efficient and scalable as possible, which was interesting to work on, but honestly the world has changed a lot since even that time. And now it's you know a lot more about live content first of all. So that's uh far more of a pressing issue is the the leaking of of live streams and live events, sports primarily handling how to trace uh premium live leaks for content owners and broadcasters. So that's become a big focus of ours in the last year, two years. Uh and then content creators themselves. I mean that whole industry and the creator economy has really blown up um for quite a while now but especially in the last 1 2 3 years and the the share of

people who are making majority of their livelihood from their content uh has really exploded. And so the concern about protecting it and making sure that content creators aren't having their content re-uploaded without their authorization and therefore losing ad revenue uh when people and their audience are watching their content elsewhere. It's a huge concern. So that's been an interesting addition uh and a a lens to see this through. we've been working with YouTube and some of the other platforms directly to support the creators uh who are concerned about that and and then lastly on the AI specific angle here I mean there's there's different parts of it that are that are AI related but AI generated content and you know deep fakes different depending on how you look at it whether it's uh generated for malicious purposes or not just differentiating between um what is AI

generated and what's not is more and more of a concern and of course like name image likeness and copyright attribution for that is more and more of a concern. So that's something that we're also working a lot these days. >> Yeah. No, I could I feel that. What do you think uh is the the main areas or is it like all areas that you're seeing people kind of stealing uh these pieces of content? Ah, you mean you it's crazy the it's it's everything that you'd think of and and more. I mean, at this point, especially on the creator side, it's not necessarily even the top premium shows or events. I mean, of course, that's a huge issue and kind of the classic piracy uh landscape of leaks and people watching stuff uh getting around pay walls, etc. But I mean because it's a lot of

this is based on just audience capture and shifting revenue from the original creator to anybody who can re-upload it. I mean anything that has any audience of any significance is is being ripped off uh and uploaded on all platforms by other parties. And in some cases, there are very well-developed uh businesses that are doing this. And it's being done at scale um by farms of accounts that are immediately stealing content from uh large audience channels and re-uploading it uh and cycling through different geographies. So, you know, as is the case with piracy and cyber security stuff in general, there are some very competent bad actors. So, it's uh >> competent bad actors. >> But yeah, that's how it is. >> Competent bad bad actors is a great phrase. Um what would you think is kind of the main uh ways that how are you stepping

into to help and what are some of the practical uh things that you think are being changed on on platforms to help with this in the in the future too? Or are you seeing anything from the platforms or is it just kind of like a you know you're kind of stuck where you're at? >> Yeah. So I mean I think in a big picture way one of the important things to do and one of our main priorities now and certainly for this year is just to make the access uh very readily available for anyone who wants to use it and essentially take what has been previously enterprise level security and solutions that required a lot of onboarding and handholding wet glove type of implementation, make that available to the general public and anyone who wants to use these solutions. So, we're making it as self-service

as possible. Uh, and that's a big trend, I think, for this, especially as it's a solution that's going to be requested by by more and more people um in the full range of the spectrum from small creators to large enterprises. And as far as the platforms, it is definitely being taken more seriously. I would say YouTube is the best uh at handling this and prioritizing this at the moment, but but others um Meta uh X and and Tik Tok are starting to take it more seriously too. We're working, as I said, directly with YouTube and we uh handle tons of coverage on the other platforms. Um so I'm familiar with some of the tools that are coming out. There's going to be an interesting NIL solution on YouTube. It's coming out in the spring or the summer that will be able to be used by

individuals for well, as the name says, their their name or their image and likeness um being used without their authorization by other channels. So, that's going to open up a whole can of worms, but also a good set of features to protect um people's people's content. >> Yeah. Um what are you uh kind of looking to improve in your product this year uh versus how how it's operated in the past or like in this past year? >> Yeah. I mean besides the the onboarding kind of self-service aspect of it which is you know definitely a big focus I would say it's the ability to handle all different types of content for all different devices. So I mean this might get a little into the weeds on on how we do watermarking and and fingerprinting and the detection uh of content but obviously with something like

this the main thing you want to avoid are any false positives. So we we definitely pride ourselves on the very high standard we hold the solutions to and you know the rigorous testing that all the detection methods go through to make sure that what we're linking and matching up is what we think it is. Um there are many different ways in which content is shared uh around the world. So different formats and protocols um codecs and devices that the content and therefore the watermarking and security has to be compatible with and we have you know 99% of that covered but there's always more. So there are a few things on the horizon uh especially um some codecs and formats that are used in other countries that we'll be adding this year uh to expand the coverage doing a lot of work in India and Southeast

Asia right now. Um so yeah that's going to be a big focus for us in the next few months. >> Interesting. So do you think that there's like anything that's going to happen in the in the realm of I guess the content that I mean, you know, let me let me take it back a step. So, you're working directly with with YouTube. >> That's what you said, right? What's that kind of experience been like? >> Oh, it's been great. Uh, we started really doing it in earnest just a few months ago, but they're very um collaborative and you appreciate the help in providing the support to to content creators on the platform who have these concerns and they have a lot to manage. So, I think having third parties do this for them is is a priority from what I can tell. Um, talking to

their team. So, yeah, only good things to say on that for sure. And and I do see that happening more at at other platforms, too. I mean, we've talked a lot with the other major platforms, too. And I think it's going to be more and more of a priority um this year and and certainly into next year uh on that front, too. But they're very supportive and and they know that both the classic you know, piracy and re-uploading issues are more important to more people now. It's not just the big studios. And also, the AI generated content is, you know, crazy and nobody really knows what to do about it and how to handle it, whether to be concerned about it, how to monetize it, etc. So, that's a that's a big concern and I think an open problem. And I guess I should add

on in terms of things we're working on. This is going to be something we're releasing towards the middle of the year. Uh there's a a big demand for what we're calling a red flag shield product that will essentially make video content created by anyone very difficult to train models on and essentially much more expensive to train on in terms of the GPU time required and provided in the form of essentially encryption where the content owner can provide a key to decrypt or remove the shielding on the content. And that would remove the additional expense to incentivize AI companies and model companies to actually pay the original owners for their content before they train on it. So that's going to be a big focus as well. >> How does that work? >> The getting them to pay for it. >> Well, no, no, no. How does

it like how did how did you me Yeah. How did one how did that come to your mind and how does it work? >> So gener Well, I mean it's a huge problem. We work with publishers for examp um we work with a lot of book publishers in terms of >> anti classic antipiracy you know finding files and textbooks and books of all types audio books that are appearing where they shouldn't and there was that major you know anthropic lawsuit last year where a lot of the major publishers uh won a settlement on the training on a lot of their content and so that conversation we were part of uh quite a bit and raised some interesting questions and in researching kind of the the state-of-the-art of what's been done to protect content from uh malicious training there really hasn't been much beyond some work

on static images that's been done in in academic content but not really for other type of of content. So the idea here is essentially introduce a level of noise that can't be removed that will confuse the training process uh and prevent the models from being able to effectively train um on any given piece of content in a predictable way as quickly as they normally would. Mhm. >> And that is something that you can throw a lot of money at and potentially get around, but it becomes very economically cumbersome and they don't want to spend more money. So, that's the thinking there. >> H Okay. Interesting. What is your uh personal kind of favorite uh thing that you've you've managed to do to help someone out? maybe it was a content creator um or otherwise that kind of has validated uh why you're doing what you're

doing, you know, personally. >> Yeah. I mean, I I think honestly a lot of this content creator work has been super gratifying because we've seen the amount of money that's going back to these creators and it can be quite significant. I mean, in one case, we return more money to a channel than they had made in a month legitimately, which >> Oh, wow. That's crazy. So, you know, that happened recently and uh it's it's a significant amount of money that that people are losing to these re-uploads and it's basically free money for them. So, it's uh a service they're all very happy about. It's nice to be on that side of the of the issue and and actually getting people >> Yeah. I doubt any anyone you're working with like mad for any reason really like, "Oh, you're getting me money back. Thanks." >> No,

they're they're definitely happy about that. Yeah. So that's that's been great and and hopefully it leads to you know I think a wider adoption of this across platforms and across the internet where you know shutting stuff down and removing content there it definitely needs to be done in some cases. Uh you know piracy contentious issue for sure but people put a lot of time and work into creating these things and they should be compensated for them. So that's that. But just removing stuff and this whack-a-mole approach, which has been how this has been handled for the history of the internet, it's it's not the best solution. And having a system where you can actually just monetize the results and get people compensated uh is much more important and I think in the long term will be more impactful. >> Do you think that there's going

to be I'm I'm curious as well. I know individual content creators is one thing and I think they should be probably prioritized, but there's a lot of commercial content out there that just gets repurposed. So like I I go on YouTube a lot, right? I create YouTube content. Fair fair enough. I want to be on there. The >> uh my shorts feed is literally just filled with clips of different TV shows sometimes that have that same uh slowed edit uh music that's like it's called No Bida. I I finally found what it does and it actually kind of slaps. Uh the song's actually pretty good. But what's funny is like some random clip from a show like Shameless or some some random uh like Cop Show or like uh Chicago Met or whatever and someone says something and then like it does a a like

a a flashing screen with the music uh hitting the bass drop and I'm just kind of curious what what is YouTube doing about that? Do you know anything you have any insight on that? >> Sure. Oh yeah. And on on Tik Tok there's a lot of this like star sparkle filter stuff that people do to get around. Yeah, I I know what you mean. And >> it's like that's like oh it's unique. I'm like bro this is not unique. You're literally just >> Hey, I mean it's tricky. That's the thing is it's inherently you know kind of going back to the competent actors and sometimes people don't even mean much by it but it's an arms race in terms of one tool to get around whatever the detection method is and then you have to combat that and come up with the solution. So there

there's always going to be uh you know somebody coming up with some new technique but it is getting very very easy to uh accurately detect. >> Oh it is >> at a at a super granular level. I mean that's what we spent a lot of our time doing with with the watermarking side of things. And um you know the the filters and and the clever bass drops the days are a little bit numbered with that. But yes, thank God, bro. Thank god. I'm so annoyed by it. It's not even funny. [snorts] >> I know you value that. Yeah, >> it's it's my entire shorts feed. It went from not being that whatsoever and then I was genuinely interested in a couple of the clips from the shows and then what ends up happening is it's just my whole feed and I'm like, >> "Yeah, yeah,

no, you're done. You gotta give it a We can come back. >> I'm cooked. I'm absolut I'm just absolutely cooked. And I was like I wasn't I wasn't planning on that. Thanks though. >> Yeah. Yeah. So that's an interesting one. But I mean soon it's all going to be AI generated versions of the same stuff. And that's why >> really again it's more going to be about attribution than trying to shut everything down because they're going to be a million different shades of the same content and uh it's going to be more important to like recognize the whole and then attribute the money and have economic incentives for handling that. So >> what do you think is the uh best part about um doing what you do? Yeah, we do. I mean, working with a lot of different interesting creative people is definitely the most

interesting part of it. Uh, and seeing kind of the the inner workings of the production and the distribution and how people think about what to make. Again, ranging from very niche small creators around the world. Uh, different cultural communities, too. That's been pretty fun. Uh, see how different countries think about which platform and types of content they want to create. and then ranging all the way up to to the biggest studios and broadcasters. And I mean, it's pretty interesting, pretty complicated stuff that they figured out, workflows and distributing this stuff live. Uh, it's it's pretty cool. >> Yeah. No, absolutely. It seems like it I I I find what you're doing really cool. Um, what are you uh trying to I guess I would say do actively right now to make it accessible? I know you said you're doing a self-s serve product that's probably

coming out. So, how how are you helping to make this more accessible? What's that going to look like? Yeah. So for the creators themselves uh any individual channel on different platforms they'll be able to sign up and complete the entire onboarding process in a matter of minutes so that they can start testing out how much money they would earn back and recover uh through the monetization and how many takedowns they can get across the internet on places that maybe they can't monetize yet but they don't want their content leaked to. So that's, you know, the next immediate step that'll be available later this month. And for larger enterprises, I mean, we've really reduced the onboarding time for even large live sports premium content owners who want to watermark content and track its distribution and even shut down uh leaking accounts or fake accounts to a

matter of weeks, which previously it has taken sometimes over a year uh to do this. And that's going to be really nice to to roll out to more and more broadcasters and uh streaming services in the next few months. So those those are things that we're kind of focused on right now. >> Okay. Nice. What is your um kind of uh opinion overall on how you can continue to grow with these different platforms as I'm sure some are there internal systems improving on this at the same time? I'm sure. >> Yeah. Yeah. So, we work closely with the internal systems and kind of layer on top of that and in some cases inform them at different platforms. Um, I mean, I I think there's almost going to be as much like interplatform interfacing as work we're doing within each individual platform. So, I mean, as

every content owner and distributor knows, you want to put your stuff out everywhere. So really the most common request from content owners is can you handle multiple platforms? I mean people have their favorite and one they care about the most. But being able to be like interoperable between all of them and have kind of one central view where you can see the monetization the protection of your content across everything. That's what we're providing to everyone and focus on on improving this year in terms of that kind of central command station. They can see everything uh across all platforms and take whatever actions they want without having to check a dozen different things and you know have a bunch of headaches. >> Nice. >> So trying to reduce the the number of of points of friction there. >> Yeah, totally. Um, I guess kind of curious

your your general opinion on some some more AI stuff in general. What do you think has been the general um positive and negative impacts of AI content generation after spending so much time seeing this, right? Because I'm sure you know some pe some individuals themselves are using AI to generate some content um inside of their own videos and stuff like that. And then obviously there's this other side that you deal with. What do you think is kind of the two sides of the coin as uh as you've experienced it probably the most in depth out of of the majority of people? >> Yeah, I mean it's a good question. It's, you know, it's amazing. Obviously, it's pretty crazy what is possible now and what people are able to make. And I think it's just obviously the beginning of a whole new kind of art form

where people are going to be creating a lot of their own content directly from these models and they're going to want to protect that, too. So, uh, that's all cool. I'm I'm a fan of of all that stuff. I think there's also like you said some kind of intermixing of the AI generated stuff with you know traditionally generated content. Um we're seeing a lot of that with with advertising which of course always leads the way with some of this. So customizing certain um advertising ad placement stuff within content using AI generated tools. I've met a number of companies that are doing that recently. Um, and yeah, I mean, putting different characters in in different universes, I mean, mixing different creative content together, that's that's obviously uh a huge thing people are doing now. And, uh, has a lot of attribution questions related to it as

well. But yeah, I mean, I don't need to tell you, this stuff's happening just at a crazy pace. Um, in fact, I think I was just seeing the uh most popular streamer on Twitch is AI generated now as of like last week. >> What do they make? >> How much do they make? >> No, no. What? Uh, that was a poorly friend question. What uh type of content do they make? >> I wonder what Yeah. How much they make? >> Yeah, I actually want to know the first question I asked. But >> you know what? I don't know. I just saw like a headline uh about this, but >> it's a ridiculously stupid headline that actually kind of triggers me. Um but hey, you know, I if if they make interesting stuff and it's it's like >> Yeah. >> Yeah. I'm not sure. But I

think there's going to be a lot of a lot of I mean, I don't have anything to say on that that anyone else doesn't already know, but it's it's really just moving at a crazy pace and it's it's exciting. So >> that's pretty cool. >> I mean, fair enough. What do you think is the the stuff that you're you're most looking forward to in the um well well actually I asked uh on the positive note, >> right? So with negatives, what what are other uh what have been in your opinion some of the cool things that you've seen on the AI generated side that didn't go across the line, didn't just turn into a no bott um AI edit? what what does what does kind of uh get allowed because you go, "Oh, that actually is unique and interesting." >> I mean, you know, I

think a lot of this stuff is is super creative and people are going to want to own the attribution and the rights for just as they do with any other type of art. I I mean, I definitely believe that's happening. It's there's obviously a lot of the slop out there, but I think the proportion of legitimate art created this way is only going to increase. Uh, so, you know, there are a bunch of different types of it. I am definitely not an expert in in all of it and not following all of it super carefully, but I've seen, you know, videos, uh, images, music that is all very high quality. So, I mean, it's a it's an interesting thing to come to terms with, but I think people are definitely going to want uh to attach their their name to the stuff they created with

those tools uh more and more this year. Uh and I think there'll be platforms more and more dedicated to that type of content or sections of platforms. Um and filtering within platforms. I think that's going to be an interesting thing. We have had some discussions with a few of the big social platforms about that just in the near future in the next year or so how can filters be applied broadly to you know exclude anything AI generated from a feed which is a tough thing to do accurately um you know watermarking and other detection will help with that but uh that's going to be an interesting an interesting part of this too so when people do or don't want to see that type of thing. Um, like filtering is going to have to get very nuanced to handle that. >> I was going to say,

do you think it's going to become the AI generated content is going to become so unique at some point that like kind of we're we're going to be in a position to to not be able to do much about it? Obviously, I uh, you know, have full optimism on the the prospects of what you're doing as a company, don't get me wrong. Um, but it's >> Yeah. uh do much about it as far as like tell whether it's AI generated. >> Tell whether it's AI generated and there might be a level of where they can claim like true uniqueness. >> Oh, I mean there's no way people are going to on their own be able to tell uh probably even now. >> I think it's reaching a point. Yeah. W without >> I mean not to totally sell our stuff, but I it just is

what it is. I think without like advanced types of in AI, you you're Yeah, you need a a comparable detection to tell cuz some of it is it's just so good. >> Ju just to kind of speak to this at kind of a meta level, what I what I have noticed is interesting to me is you are working in in such a way where you're essentially a um you're you're one of many companies that I've noticed in this year. So you are kind of the counter to a lot of what's going on, >> right? >> So in AI, something I've noticed is there is a bunch of opportunity out there >> for creating a response piece of uh AI to an output. >> Right. >> Right. and you're kind of sitting there in an interesting position because you know you're working with uh the response

to AI generated content by analyzing AI generated content. Do you think at at kind of a fundamental level then that level of discernment is only possible due to it knowing how like like I'm trying to ask functionally how does the AI itself since it is a visual >> able to determine that it was a AI visual versus how a human can determine whether it's an AI visual or not? cuz I I'll be honest. I I saw like a behind the scenes thing that was popping up on my screen from the new Avatar movie and it almost looked like they were all dressed up with um makeup and stuff like that. And then I went, "Wait a second, don't they mocap this?" >> Yeah. Yeah. >> Right. And then I looked at the comments and people were like, "Oh my gosh, I just realized like after

watching this for 60 seconds it was AI." So, don't get me wrong, I get how the human eye could be, you know, confused, but how is it that you as a company I I'm I'm genuinely curious. How does the A You don't have to give away the secret sauce, but is it like mechanistically, how does it figure out that it is AI? >> That's a a very good question and kind of gets to the point of of all this and kind of the fundamental thing that we and and everyone else doing this has to grapple with because there are really two sides of the coin. You know, deep fate was kind of the term everyone used to talk about this until the last year or two, but you know, just to use that as the example, you can try and see whether something is AI

generated or or deep faked and see if it has six fingers or whatever, you know, when those were the tells. Um, so just kind of approaching it from the top down and analyzing the pixels and any, you know, aberrations and that's just inherently not going to be perfect and it's really always going to be uh behind the cutting edge generation. doesn't mean you shouldn't do it and we have done it and uh there are requests always to have detection methods like that but the the surefire way to do it is own it from kind of the bottom up and have like initial filtering at the point of creation and this gets into like a whole thing about all the different ways people create content but stamping it watermarking it however you want to describe it at the the point of origination to say this is

how this content was created you you know, it was created with this model. Uh, it was created with this piece of hardware. Um, it was submitted through this account tracking the whole chain of provenence and and ownership. Um, and that's going to be the most reliable uh sense of of trust people have. I mean, we work with some of the large uh news organizations and I mean, they're having a really hard time right now. And there's we're part of this organization called C2PA uh content providence and authenticity and they uh on the news side you know somebody could send uh any type of video to the Associated Press and say hey this I just took the video about they don't know you know like so >> that's really bad. Yeah, they want to know, but they want to be able to have methods of submission

where it's like when was this created and how was it created, which piece of hardware, which channel. So, long way of saying I think there's going to be a series of checks both like the trail and the provenence of how something was created, what the original source of creation was, whether it's a model or a camera. Um, and then layer on top of that the top down like are there six fingers and all that together is going to be how people know what's going on. Um, but it's a really interesting problem. I mean, and it's not solved yet. >> Yeah. Um, when do you think it could be solved? >> I don't know. I mean, it's one of those things where in a sense it never will be 100%. But I think there's going to be more reorganization around uh you know I'd say kind

of less closed garden ecosystem which has been web 2.0 you know, internet of the last 15 years type stuff. Bless you. >> Um, and more there there's going to have to be more adoption of shared standards around this stuff, which is already happening. I I think within the next two to three years, there will be more uh clearly adopted standards around authenticating the provenence of stuff. Some of it's at C2PA group I mentioned. Some of it's going to be, you know, other types of protocols that everyone's going to have to work with just as they did with, you know, HTTPS and and other things. So, >> Okay. Well, cool. What do you think some of the cool new trends that are going to happen in general with um AI um are going to happen this year? Well, I mean, I guess getting away from the

content stuff, uh, I mean, there's a lot of cool stuff within companies and software development, which I'm sure you talk to a lot of people about. >> Yes, absolutely. I talk about that a lot. Yep. >> Yeah. I mean, a lot of those tools are are super exciting and interesting and save everyone a lot of time with busy work and um, boilerplate stuff. So, I I'm a fan of all of that for sure. Uh I think and I think a lot of it is gonna this is going to be a big year for that for sure in my opinion. I mean it there were interesting tools but it was very wild west and kind of every week it seems like there's been a different thing for whether it's coding or you know operational stuff or >> Yeah, I was going to say what's kind of

the uh I mean on your company internally what do you use a lot of to kind of save time do more >> uh I mean we use a lot of different uh automated software development tools. I mean we've used >> um you know a few of the different kind of sampled different ones but classic and codecs and other things just to especially automate testing um code rabbit uh for like PR reviews and other things like that. I've tried a few um and you know the quality of the core team I think matters more than it maybe has in the past because if you have a really good core team then you can do a lot with these tools um and maintain a similar sized team which is pretty amazing. >> Yeah. Do you think we're going to see some like leaner teams this year? >>

Oh for sure. Yeah. Yeah. >> Okay. Are you pretty like are you white pill on like the what that means though? Like do you feel like it? Well, I keep using that term. Do you know what white pill means? >> Okay. Okay. I gota get it. Sorry. It's like a I am like a at 6040 people actually aren't getting what that means recently. >> Also chronically online to me. So I I >> Okay. Well, listen, as long as we're on the same page, brother, then then we're good. >> Yeah. Uh I'm a Yeah. I mean, we're a we've always been a a small company in terms of headcount, and that's been a focus since day one about we want to stay as lean and small as possible because for for a bunch of reasons, but I am a big fan of that. Um, and yeah,

I am very optimistic about all of that. Um, and I think when used right, it it makes everything better. uh and makes people's lives easier because it's less of annoying tasks and and busy work. So >> yeah, totally. I don't Yeah, I I agree with you. I think um everybody in a practical level probably wants that. Uh >> right. >> But people don't want to negatively affect them and take away from things they enjoy doing. But I haven't really seen that happen yet in in our in our world specifically. >> What are you um most excit What are you actually using internally um that you're a big fan of right now at the company? Uh like the main uh like if you had to name one favorite AI tool that you use uh yourself. >> Oh, I mean it's nothing special. I I don't think

anything we're doing is something that that other places are using. I mean it's it's mostly the big ones. I mean honestly having codeex for uh coding partner I think has been the most impactful and and these other two tools for >> uh like PR review and testing uh on the software side has been >> nice I mean it saves people 20 to 30% of time um I mean the rest of it's just clawed and JPT and you know I think that just saves time with market research search, but uh like a little sales side. Um I don't really have a specific one to recommend, but all good. No, I just kind of that's my been my question is just asking people who are AI founders kind of what is their favorite u tool they're using right now because you guys are kind of on the

forefront and maybe you're you've seen some stuff. I mean the thing is I realize we're in a bubble but like you and I individually and others but even in our little bubble we all seem to know some unique stuff. Oh, Universal Cool School is uh Granola, this note takingaking app >> uh which >> Granola. >> Yeah, I'm I'm like a horribly unorganized person in many cases. So, >> okay. >> Nice and recommended to me by organized people who I trust. So, they uh yeah, it just, you know, kind of a in browser tool. It will transcribe calls but in a very intelligent way and organize them, provide follow-up. And I've seen a few of those and this one seems like it's the best. >> Nice. Well, I really like um I use Grain. Um that's a that's my meeting recorder of choice. To be fair,

I have kind of learned that there is uh some opportunity out there to uh kind of make your own um if you know how to do some basic coding with um >> like just lovable or whatever. Like it's it's actually so so easy to to kind of uh hack something together nowadays. And I'm actually about to switch over to like my own thing that's going to manage my email or whatever. But um >> Oh, cool. >> Yeah. Yeah. It's a it's a fun time out here to to just kind of >> Yeah, it's always a new thing. A lot of new tools. >> Always a new thing. And you can even make your own new thing for like 50 bucks or less. So, >> um well, I think that's uh kind of all the questions I have for you. I just want to kind of

close things out by asking you to plug what you're you're doing so we can uh uh get everyone over >> Sure. Well, yeah, thanks. Those are great questions and good discussion. I guess I would just encourage anyone who's interested to go over to our website redfogi.co. You can request a demo uh reach out on email, set up a meeting, and soon have that self-service for a >> and uh if you see any no botto edits of uh shameless on your YouTube uh uh downvote it and uh actually actually wait a second, don't downvote it. market as spam or whatever the heck the that YouTube feature is because guys, if you if you dislike something, it helps the algorithm. Just fun fact for your people. It's uh it's something that has been bothering me as a YouTuber for years. It's like >> when it's like people

would dislike a video I made, I'd be like, "Guys, guys, guys, this is not what you do. You're not you're actually helping me. So, thanks, bro." But >> yeah, you're telling the secrets now. Yeah. >> Yeah. Exactly. All right, man. Thank you so much. Uh, everyone make sure to go and check out uh, Red Light AI. And with that being said, also make sure to leave a like, comment, and uh, let us know what your thoughts were on the product and what your thoughts are on in general with uh, all this cool AI stuff, whether you're um, bullish on this, whether you've been seeing the no bat edits or whether I'm the only one who's been uh, scarred for life by this being on my feed. Thank you so much for watching, and we'll see you in the next one. Peace.