UpSCALE

This conference session features a series of lightning talks that explore various topics related to technology and open source. The speakers discuss unique aspects of software, community platforms, and ways to democratize technology. Adriana Frick presents a contemplative take on process management using Linux principles, while Jonno dives into penetration testing and the OWASP Top 10 vulnerabilities; he emphasizes the importance of understanding security in web applications. Evans introduces the Fediverse, highlighting decentralized social media platforms and their benefits over traditional big tech options. Anelise Stenny traces the history and revival of ASCII art, linking it to cultural and technological nostalgia. Finally, Christian Alexander shares insights into the marketing strategies of major tech companies versus open source projects, and Eric addresses the implications of AI integration within open source software. Each talk aims to inspire thought and discussion about the future of technology.

is this mic. Oh, amazing. It's on. I have the scary mic that doesn't have an on and off button, so I'll try not to say things uh during the event, you know. Hi everyone. Welcome to Upscale. >> This is great. We're going to practice a lot more clapping, so you're already doing a great job. So, many of you have been to an upscale, but if you haven't

before, this is our fun, entertaining lightning talks. It is low pressure, low stakes is what I keep telling the speakers. So, let's make that happen for them. Um it is five minutes and the catch is their slides auto advance every 15 seconds. They have no control over their slide deck. One just to see how perfect their timing is and to give us a little bit of entertainment

because you know talks although are very great. Sometimes you just need to see if they mess up a bit and that's half the fun. But that being said, it causes stress for speakers. So if they get confuled by this arbitrary constraint we've put on them, we do something that I like to say is clap them through it. So let's practice. Someone's slides are a little bit off

or they need some time to kill. [applause] Great. You're doing a great job. Um I tell them this is a very friendly audience and it's a great for firsttime speakers. So don't make me a liar. Okay. Perfect. So without further ado, I'd love to welcome our first speaker up, Adriana. Hi everybody. Uh my name is Adriana Frick and I'm here presenting BTOP 2, a contemplative meditation on

process management. If you attended last year, you might recall that I did a a speech on BTOP and using that internally as a person. So this might seem a little odd for me. These ideas have always been connected. I was a system administrator before I became a yoga instructor. So that's how I understood these things. And if you missed out on my talk last year, since this

is a sequel, that's a link to the original video. So here processes are not equal in value. We can agree on that. The same is true of your thoughts. Not every thought is worth prioritizing just like not all of your thought your processes are worth prioritizing. So how can we apply Linux ideas to our thinking? Now we know that we have priority cues. I'm supposed to stay

behind the podium. We know that we have different ways of handling processes in the Linux world. Some ideas have to be or some processes have to be dealt with immediately. The same is true with your thoughts. So in the yoga tradition, we have true or correct. Those are things that you logically deduced, that you personally experienced, credible um sources, and then you have things that are maybe

misconceptions or just biases. So they're called false thoughts, but they could still have value. The same is true of memories. They aren't always true, but they're still of value. So is imagination. And then you have zoning out. This kind of sleeplike state where you're blank of thoughts. So you can kind of go through your thoughts and label them. You find a point of focus. This is something

we practiced last time. And you go, okay, maybe this is a true thought and I can continue having it. Maybe I'm just zoning out. Maybe this is imagination. Maybe I can cue that for later. Now, we can find any point of focus. A penguin perhaps. Last time we did body, we did breath. But think about as we're going through this, I may have triggered a thought in

you. Maybe you had a thought from a past dro. Maybe you had a thought from our last presentation and you can say well that's a memory or I might have bias from the last time I attended or you might be daydreaming right now. So just observe your thoughts and notice whatever comes up with non-attachment without judgment. Now sometimes we have sticky thoughts that don't let go just

like in BTOP. Now in BTOP if you have something that's using excess resources it will actually be colorcoded and this idea is present in yoga too. So, we're going to call these discolored or alarming processes. And in Linux, you send an interrupt of some kind. And we have different kinds of interrupts because sometimes you need to kill that process immediately, even when it's not listening. Sometimes you

can be a little more graceful about it. Now, if you've seen sneakers, this actually is very close to when you have discolored real thoughts. If you recall, he goes from point to point, and there's one where just a ton of data starts coming at him. Sometimes you have that in your own thoughts. Okay, I'm thinking about this thing and suddenly something hits you and you have too

many resources dedicated to this thought. So if we had a BTOP for our brain, it might look like this. Now you're sending an interrupt and it's not listening. That is a sticky thought. It is not going away. It is persisting. And you're trying to label it and you're alarming the whole time you're labeling. Okay, I have this past memory. I'm remembering a past role I had and

it's maybe making me feel bad or I'm imagining something about who I am. I'm imagining something about what I'm capable of. So, we have tools for this too. Then you may want to label it ignorance. I didn't know any better. Or wrong. I amness. I love this one. You have classified yourself as a thought. You have classified your selfhood as your role. And you're much more than

that. You may be attached to something because in the past it gave you pleasure even if it won't do it again. You may push it away in aversion because in the past it gave you harm. Even if that's not true, maybe you are sticking to this thought because it's making you fearful. So you can use these labels and it doesn't send an interrupt. Unfortunately, with the sticky

ones, you have to just look them in the eye. You have to say, you know what, I'm not that person anymore. I'm overly attached to this historical role, and that's not who I am now. I didn't know any better back then, but now I do. And now I'm a a bigger person than I used to be. And over time, that color fades and those thoughts become less

sticky. So when we're managing our internal processes, we need to remember we're the system administrators. We are not the process. We aren't the thought. We are the thinker of the thought. And by reducing the power and the the coloration of these thoughts, reducing the resources, we can regain that power. So if this was helpful, you can find me at the Ubuntu booth this weekend. Last year I

said if five people signed up, I would do a newsletter. Four people signed up. So maybe one of you makes it happen. Thank you very much. You have to keep clapping till I get to the podium. Yeah, there we go. Great job, guys. I really wish that I could kill my thoughts with like a command line prompt. That would be incredible. So, that's what I'm going to

be daydreaming about now. Thank you. All right, please welcome to the stage, Jonno. Hello. Hello. All right. Hey everybody. All right, we're going to talk about penetration testing game day. Um, who here has heard of the OASP top 10? Okay, cool. Oh, it's a popular thing. Okay. Um, yeah, I uh what can I say about me? I am a software engineer. I've been doing it for a

long time. I also do a bunch of open source stuff and security stuff is very interesting to me. um working on development teams, you'd be surprised how many people have not heard of OASP top 10. Why should you care? If you're writing code, you need to understand why it breaks. And it's a lot of fun to break. Also, okay, cool. Okay, so penetration testing, if you happen

to not know what it is, it's like an authorized way to attack something, attack a a surface. We're talking mostly about web stuff here because that's mostly what I'm working in. Um, and it's legal. Some people get paid to do it. You know, why do we practice? Uh, because you get better at understanding what the exploits are. Um, and doing it in a safe environment matters, not

just going and hacking some big company. That's okay, too. But that's what people get paid for and in a legitimate way. The OASP, if you don't know what it stands for, it's an organization called the Open Web Application Security Project. It's a community. Um, they do a lot of work to collect information and educate people about security practices. The OASP top 10, that's a mouthful, is the

top 10 uh most vulnerable things that the the group has decided are important at the You've probably heard of of a lot of these individually, but they also change over time. Um, and here's like some sample ones. Broken access control, crypto failures, uh, insecure design, and it's not just, uh, web necessarily. So, I'm going to just talk about one of them to give an example. Uh, injection,

probably the most widely known one is when there's user input and it's not sanitized. So, it can do something malicious on your system. you will probably know this slide here. SQL injection Bobby tables. Everyone here knows this, but uh not everybody knows this. Um, OMS do help. We use them. They can only help you so much. Uh, broken access control. This is number one. Uh, users do

things that they shouldn't. Okay. Uh, the UI doesn't show them a button. So, this is sort of the security through obscurity sort of thing. Um, this is very important because you're trying to make a secure site. Cross- site scripting. Uh, this is where the user can inject something on the browser side and have it uh execute something that the user didn't want. It's like another type of

u injection. [snorts] Um, and OASP is always evolving. So, I you know when I it doesn't change that often, but when it does, it keeps up with things. A couple years ago they added some GraphQL injection stuff which I think is very relevant or it was a couple years ago. GraphQL seems to have already come and gone. Um your browser is a tool. That's the main tool

we use in this workshop. Um you just need your web console a regular browser maybe some extensions usually not much else. Uh that's one of the reasons why I offer give often give this to people who are not even developers. Okay. The OAS Zap is uh an application that helps you do a bunch of attacks. It will guide you through a bunch of things. It has builtin

the top 10 attacks. It you can do spraying attacks, dictionary attacks, all this stuff. And it has scripting built in. The OAS juice shop itself is a vulnerable application that is made for attacking. That's what we use in the workshop. It is a gamified system that actually gets you through some of the challenges and keeps track of it. Here's what the the juice shop looks like. It's

a very conventional web app. You can order things, you can have conversations, there's support, there's um ratings and uh privacy policies and all this stuff you'd see on a regular site. Um so on the site you're actually seeing your scores. So there's a a separate scoreboard area. And actually the first challenge is to find the um because it's sort of hidden. Uh but you'll see how far

you are and you could spend a long time doing it. Uh this is a CTFD server. This is a way if you want to plug it together so everyone is running their own instances. You can see everyone's score over time and it's kind of fun for people to watch. Um where I go next. So if you want to learn more about security, there's a ton of other

stuff. We're we're the Linux people, so we know about Cali and Metas-ploit and that stuff. There's a CTF event here uh tomorrow, I think, over in the other building. Capture the flag if you don't know what that is. Um, and that's it. Awesome. Thank you, Jonno. Uh, I think maybe the juice shop should come to game night next year. Let us know what you think. All right.

Please welcome up our next speaker, Evans. All right, my name is Evans and I like social media and enjoying a reasonable dopamine boost by scrolling on Limmy and Mastadon every day. That was faster than I thought. So, this is Limmy. Um, it's a community-based news aggregator like Reddit. It has memes, news, comics, uh, posts. Uh, the most upvoted post of all time is a broken link to

an image of a can of beans because, you know, they're just weirdos. Um, this is Masttodon. It's like Twitter. Instead of communities, these posts come from individuals, brands, or self-identified bots. George Takai, uh, aka Mr. Sulu from Star Trek, is the most popular account with Eugene Roko, the founder of Masttodon, coming in second. So, what is the Fedverse? It's a collection of social media services similar to

Twitter, Reddit, uh Instagram, Tik Tok, YouTube, etc. There are already or there's about 32 different types of services right now by my count. Uh each of those services consists of many sometimes hundreds of individual servers. Fedverse apps are ethically better. Um they're decentralized, federated via the activity pub protocol, uh free open source software, they don't have ads or addictive algorithms, and they are self-hostable, so you can

run your own servers. Um many services we use are centralized uh a single service controlled by a single entity. They start off okay but as they capture market share the initificication begins. More ads worse privacy and manipulative algorithms keep you online so they get to you know more ad revenue. Decentralized services are smaller uh run by people who just want to run them and are usually paid

for by the admins with a little help from donations. They're more diverse, more resilient, and usually self-hostable. So, uh, if you can't find a community you like, you can build one. Um, distributed or peer-to-peer services are, in my opinion, the most interesting, but it's technically difficult to orchestrate a reliable, responsive P2P service. Uh, check out Secure Scuttlebutt if you want to see what the future could be

like. It's actually kind of already dead, but that's all right. Um, federation is how email works. The simple mail transfer protocol is how Gmail users can talk to AOL users or Yahoo users. Email's not a single platform. It's a website that you log into or it's not a single platform website. Ah, all right. Anyway, I don't need to explain why open source is important, right? So, you

know, we're all here for open source. yeah, never use closed source, never go back. >> So, um, I don't think there's any advertisers asking Fetty admins to put ads on their servers. And with hundreds of servers to choose from, people would just move to a different one if they tried. No ads means no monetization. No money means no need to keep you on the site. No rage

baiting algorithm. Um, look at this map. Be the first Montana or Wyomingite to run your own server in a server in your state. This is a real superpower of the Fedverse. Anyone can run a server. Don't like big tech? Run your server. Are you in Iran or Syria or Cuba or some other embargoed place? Run your own server. No one can stop you. Um, but no one

has to listen to you either. Other admins can block your instance. If you want to spout a bunch of hateful, intolerant crap, um, you'll probably get blocked. if you like radically progressive politics and memes and comics and world news and other stuff, pictures of corn and beans and cool Linux tips, then welcome home. If that sounds lame, run your own server and, you know, push that the

other direction if you like. Um, there are a few projects that claim to be decentralized that I find suspicious. Um, this is just my personal opinion. Uh, I'm not an expert, but I don't see a lot of blue sky servers out there. Um, Threads uses the activity pro protocol, but a lot of people in the Fediverse have have deferated them for reasons. Um, there's a bunch of

other services that are part of the Fediverse. So, here's just I ran out of time to put in pictures here, but you know, Twitter, Masttodon, Reddit, Limmy, Instagram, there's something called PixelFed, etc. Um, there's a lot of micro blogging. Miss Key is one that's mostly in I think Japanese, so I don't really understand the content very well. Um, Friendica is okay, like a Facebookish thing. There's a

bunch of different services, tons of them. Like I said, there's about 32 in total. Why should you join? Well, um, if you care about free open source software, if you care about taking the power back from centralized, uh, big tech providers, if you care about being more than just a consumer, uh, run your own server, join a server you like, join a community that talks about knitting

and kittens or or one that talks about activism, whatever. How to join? Uh, first, pick your kind of type of service, whether you want a Reddit like service or Twitter like service or whatever. U, pick an instance. There's a bunch of different instances. Uh, yeah, they're crazy. And that's it. Uh, enjoy I have to ask, anyone from Montana or Wyoming in the crowd? >> All right, looks

like we're going to get one state on your map. We've got one. Awesome. I look forward to the update next year. All right, thank you so much, Evans. Let's welcome up Anelise. Hello, my name is Annaise Stenny and I'm giving a brief talk on ASKI. Um, so before Figma, before GPUs, uh, they really had computers making art with just the punctuation. So going to be talking about

that. So it has 95 characters and that's kind of the canvas and every single character takes up a identical plot. So it's kind of similar to tiles being placed and that can be you know your building block to building art. So this kind of dates back all the way back to ancient Greece and we have poets there arrange text um to represent the symbols and the things

that they're talking about. Um, and in 1918, a famous poet arranged words into shapes of birds, fountains, and the iconic uh, Eiffel Tower. The poem didn't just describe the thing, it became the thing as well. And typewriter art kind of popularized in the late 1800s, and artists were feeding paper through typewriters multiple times, rotating it, overprinting characters just to build up shading. And there's no computers involved,

but this was kind of setting the precedence for that. And then in 1966, two Bell Lab researchers converted a photography or photograph in a grid telephone symbol. Management pulls it off the wall because it features a nude woman, but the New York Times puts it on their front page and it goes back on the wall. So printers uh had no graphics at the time and basically, you

know, they had zero graphic capabilities. Terminals were texton only. Emails couldn't carry images. Characters became pixels because there was literally nothing else. And you see in the 80s and 90s bulletin board systems, uh, Usenet emails, no images anywhere being carried. So artists built entire underground culture around text releasing monthly art packs similar to how albums are released. Now we see kind of the precursor to emojis. And

we see this actually pop up from a university bulletin board. And in 17 in September 19, 1982, it proposes as a joke. And that lays a foundation of emojis as we see today. And here we see early video games taking also that concept and also inspiring future generations of genre in that regards to RPGs. And then we also see here Joan G. Stark aka Splunk um posted

hundreds of ASKI work in the Usenets between 19 1996 and 20 2003 and over 250,000 visitors to her geo site page. Um she's cited as one of the you know woman who kind of helped shaped early internet and then we see here that guey interfaces come. Um the internet went graphical and the terminal felt a little bit something like your dad would use. So ask key seemed

to gone. But with anything, nostalgia isn't just sentimental. It's a signal when people start reaching backwards. It usually means something the present isn't quite working for them. The early web felt like visiting someone's home, and now it feels like a commercial shopping mall. Um, so 10 years of the same website, you know, it's colorless, bland, and arguably a little bit boring and soulless. Minimalism without personality is

just emptiness. So you see this kind of counterculture pop up in MySpace or even Aztec or Aski Design Studios and it you can even see it with the BBC BBC Sports rebrand where they actually take a lot of you know these inspirations to go build it out and then as of recently you know it doesn't just go into uh it goes into the underground culture of DJs

uh visuals and even Charlie XEX's new Brat album takes a color from the 1985 Unix prompt as well. And then we see this being put onto bodies where Aski tattoos, pixel art on clothes, and a whole Y2K, you know, revival is built around visual language and kind of early computing. And then you have people, you know, a generation of tattooing literally on themselves. Um, in 2004, sorry,

2024, ASI broke AI and they were able to kind of sneak in harmful prompts and then it passed and it was basically outwitted by a six-year-old technique. But we keep the aesthetics of era and everything whether the tool that we use, we usually tend to want to go back and this is because the tool changes but that feeling of trying something new doesn't. So on my last

closing thought is that as he didn't survive because it was efficient, it survived because people loved it. Proof that culture shapes technology, not always the other way around. Um, and that's worth something. Thanks, Anelise. I've changed my mind. I've decided Slack's out. IRC's back in based on that. So I can make that decision for everyone. And there you go. Announcement. Um let's welcome to the stage Joss.

Hello. I'm looking forward to getting a lot of clapping because I haven't practiced this. This is going to be great. Um, I'm here to talk about um making it fun to demonopy or as people usually call deouling, but including all of the big tech platforms. Um, there's a QR code here and also at the end if you missed this that goes to the website for uh Seattle

Resist Tech Monopolies. Um, and we're going to have featured a lot of cats in this that are in a calendar that we have that you can buy. Also, um, Resist Tech Monopolies is a group of us that wanted to use our skills to self-host and learn, but also to help other people get off of proprietary tech. Um, one of the things that we have been experimenting with

is the concept of a disco tech, which is disco and tech. We uh got this from um we got this from a group in Detroit um and we modified it a little bit. We do it science fair style with tables and each of us sits at a table and advises and learns about a different topic. Um we have like one group one table for social media and

one table for email and things um different topics that people can come in and learn about. We do it um at libraries. We um do it at um larger groups um that are having other events. Um, and we've been experimenting with just what works and which kind of stations work. Here's some examples that are in one of our zenes. Um, and we also interview people during the

event to ask them like what kind of barriers they've run into as they've tried to use open source tools. Uh, one of the things that we came up with to engage people that kind of don't know what to do when they show up is um the bingo game that you might have played before where you collect signatures from people for like is there somebody who um and

so for example, is there somebody who uses Linux, who has a cat, who's chatted on Signal, who has a library card, who knows what Project Nimbus is, who knows the song that's playing right now because we play disco music while we're doing it. um who got here without Apple or Google Maps, etc. Um and then we learned that we needed to build some zenes uh because we

need to have something to give people during the disco tech and when we're helping them get set up with this software that they can have something something to take home with them. Um and one of them we actually got a partnership with um an artist who was kind enough to make one of these for free for us about signal and we're going to be making more about

the other technology that we help people with. Um you can find all these on our website and download them and feel free to use them. Um we also came up with a calendar. Um each month features different technology, open source technology and it features one of our pets and quotes about from the pets about uh the technology and information about how to implement that technology. We also

have a list of all of them on our website and you can see the rankings of how difficult each of them is to implement and we offer prizes for people who do all 12. So contact us if you're interested in the calendar or if you're interested in trying to do all 12. Um this is a honorary dog because we're we use cats mostly in all of our

stuff but we have a couple of dog people in our group too. Um, there's a QR code on each of these that gives the information about each of the the non- big tech tools. Um, finally, okay. Um, and we actually have this for sale at one of our local bookshops. Um, and we're going to be coming out with stickers that you can put on when you accomplish

the different tasks of deleging different software. Um, another thing that we do is we invite people to come in from the public to do book clubs with us. Um, we're really interested in trying to create a kind of culture of um, welcoming when we do deooding. Um, it's not about like telling people what to do. And so the book club helps with that. Um, tomorrow I'm going

to be leading a talk um to talk not about how to deleg but about how can we help other people de Google that actually works. It's called tech glamping as a service. How do we help busy non- techies doole? So please come join me to discuss that and give give us ideas that I'll bring back to my group. Um, and here is going to be another QR

code and our mastadon and blue sky. Did that talk want to make you all hear investments and Goldman Sachs and other things like that? No. Okay, good. Because Christian's not talking about that. So, welcome to him to the stage. [laughter] >> Yeah. So, Christian Alexander, I was at uh I was on Wall Street for about 30 years. I worked at Goldman Sachs for 15 of those. Been

in open source for about six months. Uh I work at a nonprofit foundation, Appario, which promotes open source in education. My job is to raise money. Um so these are just my first impressions having limited experience in this field or in the you know in this universe. Um Megate Tech spends way more money um on advertising than it does on R&D. The numbers there are outrageous and

that's because it works. There used to be a saying that my boss had which is perception is reality. If I think you're an idiot and I'm your boss, you just became an idiot because I think you are one. And the reason that uh that these big corporations use so much advertising is because it works. And the reason that open source struggles is because it doesn't. Um, so

this is just a fun slide. You know, I don't know if you remember this thing from the 80s, but we've got a headache this big, the eedran headache, all that. Um, you know, it it is a uh feature of our culture that we all can remember. Got milk or you know all of the popular ad things, advertising campaigns that you saw as a kid. Um, now this

is to the point of where we live in open source and higher ed. Microsoft spends as much money on marketing and advertising as the IT budget of the entire University of Massachusetts system times 490. Um, Bayer spends 750 million bucks in 2025 on advertising. Aspirin spent zero, which makes our product the equivalent of a generic drug. And then look at the pricing differential. It's going to be

three times the price for half the amount of aspirin if you buy it from bear, which is really the same hurdle that this audience confronts here in going up against um you know well-branded products. Um now down to the space where you know the space where we inhabit um instructure which is one of the big learning management providers spends two and a half times just under two

and a half times more on marketing than on R&D. Um now meanwhile the LMS space is still in the process of consolidating. So every three years almost on the button you either get a merger, acquisition or bankruptcy which then is going to put your un your average university in the space of forced migrations. This is the um the Sunni system. They have gone from um what was

it? Lotus to angel to blackboard to whatever they're on now. They've done four they've had four different systems. They've done three force migrations. They've spent 75 million bucks that they flushed down the toilet and probably spent double that on the migrations. Now, let's look at the companies they're buying stuff from. Microsoft is 10 times the size of Cisco, which to me is just a mind-blowing fact. Like

Cisco, I think of an enormous company. Um Cisco has done 220 acquisitions in the last 30 years. You effectively what they've done is like a bailing wheel just go and gobble up all of the stuff that's out there that could that could potentially compete with them. And you'll see it in their valuation trend next. So look at there's the cumulative acquisitions and there's the valuation of the

company. Um you know so in time what happens is each sector gets hollowed out um and becomes one dominant provider with monopolistic margins. Um Microsoft here has done the similar number. It's about 250 that they've done in the last 30 years. The only time that they've taken any time off was during the dotcom bubble burst, the financial crisis and now. And the reason for now is there's

nothing big enough out there that they could buy that I think could really move the needle. There's the valuation and there's the cumulative acquisitions. And again, this is something that's 10 times the size of of Cisco. Um the next one is going to be Apple, which is really just mind-blowing. Um look at you watch their margin trend. So you can see the transition to the monopoly phase

there and then you can see the monopoly phase and look at how the margins have gone up even after that. this it's it's uh the amount of margin packed into a commodity product is just mind-blowing. Now there's Apple versus Dell. Margin trend for Apple way better. And this margin spread is like a full, you know, 20% over. So you can charge 20% more for an iPhone than

for the other thing. [snorts] Um now where we have an opportunity I think to reframe how we're seeing is 80% of the software out there in operation and in function in universities and elsewhere um is at its core open source stuff and for us I think the opportunity is kind of you know hey mag you're soaking in it like is to reintroduce or to to promote this

idea at least for us in higher red that um that what you're using is already thoroughly loaded in open source and to try to frame the conversation um are we a generic medication or are we something more exciting like a DeFi you know decentralized kind of revolutionary thing and at least for us in education the time I think is right for that because there's enough duress in

the space to hear new ideas okay I fear that I've committed a cardinal sin It is 2026. We are six talks in and we haven't talked about AI yet. So, rounding out the show, let's welcome Eric. >> Hello. Okay. Uh, I can maybe talk about Cisco too. Uh, I was part of one of those how many 200 something. Okay. Well, I got to focus on the slides.

So, I'm going to talk about AI slop. Uh who here uh is knows the term AI slop? Yeah. Uh who here is like uh not concerned about AI slop in open source? Not concerned. Yeah. Okay. Uh who here thinks that we should ban all uh AI coding agents and AI code in open source? Okay. You four. I'm going to try to convince you in the next however

many minutes I have left uh that there is a way that we can actually achieve a world of abundance with AI tools to help open source maintainers. Uh so uh you're probably mostly familiar with all these. I'm not going to go over each one, but there are a ton of problems with people using AI uh in open source. Open source maintainers are already burdened. Uh they're already

doing like amazing work. uh that's kind of thankless and the entire world runs on it. Uh but you probably some of these call out like curl uh the CVES. I think that was the first uh time for me that that I saw kind of this danger. Um and uh a bunch of people just opening up CVEes because they could get money if it was accepted in in

a in a pentesting kind of hackathon. Um and uh that that was pretty bad. There's a bunch of others here uh as well. And I'm realizing that um you know I thought I was being kind of smart uh using the uh 15-second auto thing and I was like I'll just duplicate a bunch of slides and now I'm just sitting here and [laughter] so okay so you you

could just say no you could just say no and I'm not going to use these tools they're not helpful at all or you can adapt right there. So there's a lot of this going on right now. Who has heard of vouch from Mitchell Hashimoto? Yeah. Who uses ghosty? Ghosty. Liby. There's a lot of you using lib ghosty. You don't know it. Okay. Uh so the there are

ways that you can kind of look and identify uh something that it's that's AI slab. We've put out uh we try to be as open as we possibly can around evals of all these models to cut through hype uh at the company I work at. And we also uh put out things that can help you identify. We can see in different models kind of keywords that are

used or key patterns. Uh it's nothing that's deterministic, but you can easily identify uh you know you're absolutely right if you've heard that one before. Uh but there are a lot of other ones as well. Um a lot of grammatical patterns as well that are just baked into these models. Really easy to identify. Then there's a lot of stuff where you can uh gate it. So there's

the kind of hard gate of just like banning anyone that's not a contributor, anyone that you don't personally know. Uh but that is closing off not only your ability to maybe use these tools, but it's closing off the next generation, right? Are we going to be the last generation of people maintaining open source? Are we going to be open to to bringing more people in? Because if

you just say you've never contributed to open source before or you've uh not you're not part of this this this family of of contributors uh every single person that has ever contributed to open source was at one point a person who had never contributed to open source. So we should not ban people. Um, and then where we can get to is what I'm calling abund the the

world of abundance. Uh, which is that you can actually use these tools uh to help people become contributors to your project. So for instance, if you have a contributing guide, who who here maintains an open source repo? Yeah. Yes. You probably have a contributing guide and probably people that are pring are not following the contributing guide. So one of the first things is like, hey, can you

follow? Yeah, I see the snaps. Uh, you can just auto handle those things. That's a great thing for an AI tool to help with. Uh, that's something like issue triage. Great thing to help with because there's so many things that are just, oh my god, we're telling them to use the contributing guide. Oh my god, they're not I don't know what they're trying to do or they're

just pring something without there being an issue. All the process, everything. Anyway, uh, grab your phones and scan this. It just takes you to our website. Um, it's free for open source. It's code rabbit. It's AI code review. It's free for open source. Plus, you don't even have to use this. Uh if you want to be to get money, we're trying to give as much money. We

have a goal of a million dollars in cash each year to give to open source. Yeah. Please uh take our money. And if you don't take the money or if you do take the money, it doesn't matter. There's free cupcakes in the back. Carrot cake cupcakes. So, thank you. Thank you all. All right, that concludes our upscale for the year. Thank you so much for joining us.

All the speakers are up front if you want to chat further about any of their topics and see you all tomorrow at the keynote.