The Stack Overflow Podcast

Blocking the haters as a service

Episode Summary

We chat with Tracy Chou, founder, CEO, and lead engineer at Block Party, a software platform working to make social media safe and enjoyable by helping users avoid, or at least manage, online harassment.

Episode Notes

Chou, a  Stanford educated computer scientist and electrical engineer, cut her teeth in Silicon Valley with stints at Facebook, Quora, and Pinterest, where she advocated for a stronger focus on diversity. 

Block Party describes its mission as building "anti-harassment tools against online abuse, but more fundamentally we are building solutions for user control, protection, and safety."

As CEO and lead engineer, Chou gets to choose the company's tools. Block Party is built with technologies like Render, Flask, and Jinja. Paul is very jealous of this stack.

Our lifeboat badge winner of the week is Bryan Oakley, who answered the question: How to redirect print statements to Tkinter text widget?



Episode Transcription

Tracy Chou I think that's another area where online society is kind of failing, like, not only are there not really any laws, besides in theory, these terms of service rules, which are not really that enforced, but then they kind of get ignored anyways. But also like the social norms are totally broken, right? The norm is that you can be a complete [censored], and it's okay. So that's a hard problem as well, as human behavior, how do you fix social norms? But I do think we can try to shift that faster than laws or enforcement of laws.


[intro music]


Ben Popper Hello, everybody, welcome to the Stack Overflow Podcast.


Paul Ford Ben, it's so empty in here. It's like tumbleweeds without Sara, it's terrible!


BP I know. I know. It's gonna be double weird because we have all these episodes, like sort of banked with her. So she'll be like, there, just not there. It's gonna be rough.


PF I wish there was some way that I could socially connect with her and her new job. Some way to reach out to her and—


BP Reach out and connect.


PF Communicate, maybe be part of her professional network. But alas, there's no way. There is no—


BP TikTok. TikTok is where Gen Z is going to find jobs. 


PF It's not for me. It's not for me. TikTok would absolutely like, they'd see me dancing for about five seconds and be like, not that one. That algorithm just snick, right across the throat. 


BP Yeah, Paul drinking cranberry juice while riding a skateboard and singing along to Fleetwood Mac doesn't bring in the young engineers?


PF They have a photo of me up in the moderation room, like shoplifters at a supermarket. They're like, don't let that happen.


BP Alright, but I just wanted to shout out we did a live podcast, Paul and I, on a platform called Fishbowl which is just kind of like Clubhouse or Twitter spaces. And it was actually really fun. And I felt like people were kind of there for therapy. We had two people from the audience asked questions that was like, "why does my engineering team hate me or not talk to me, how can I change that?"


PF First of all, poor Fishbowl, which is like, "Hey, guys, come on" and we're like, "it's sort of like Clubhouse" which it is. And then like, actually, it wasn't awful.


BP No, it was good.


PF Now, I gotta give it to club—to Fishbowl. I gotta give it to them. It was a good, it was a well run experience. We get in there. There's like 100 people, an engaged audience. If you want a live conversational experience for your enterprise software conversations, Fishbowl app, it turns out—another thing on the internet, you know, good job Fishbowl marketing for getting us involved. Here you are, here's yet another plug.


BP The audience was polite, honest, and vulnerable.


PF It also wasn't just—this was important. It wasn't just one demographic of enterprise software people that was actually exciting to see. So whatever they're doing, good job. Good job Fishbowl. Alright, so Ben. I'll say, well, I mean, I think it's really simple. Twitter is is often a horrible trash fire that tortures our society. [Ben laughs] And when they try to do something about it, they basically say, "Hey, everybody, we made a horrible trash fire, but you seem to love it. Why don't you take care of it for us?" And so that's not good. That's where we are. And people over time have decided to do things about that. But why don't why don't you take it from there? So I'm just, as we're watching, I'm watching Twitter scroll by in front of me and just hating myself more and more.


BP Oh, goodness, turn it off Paul. But yeah, we're here today to talk a little bit about online interaction, moderation, and maybe how to build better systems that make for a safer, happier internet where everyone can engage in a safe way. So we have a great guest with us today. Tracy Chou, who is the Founder, CEO and Lead Engineer at Block Party. Did I get that right, Tracy?


TC Yep, that's right.


PF Hooray! Okay, so wait, I want to know who Tracy is. But not until, let's find out what Block Party is first. Tracy, what is Block Party?


TC Block Party is building consumer tools for online safety and anti harassment. So our first product right now is something you can set up with Twitter to automatically filter out stuff you may not want to see, which is great if you are somebody who sometimes attracts harassment or trolls or other unwanted stuff. So what Block Party will do is run in the background to filter it out. You can use Twitter more peacefully and happily. And then if you do need to go see what trash fire got collected to the side, you can go through it.


PF Oh, so you hold on to the bad stuff, just in case?


TC Yeah. So there's actually a few reasons for this. Some of it is actually that some of that stuff is not trash fire, it's actually good things that you want to see like that's why we are on these platforms because we want to interact with more than just the people that we already know. So there may be good stuff that you've potentially filtered out you want to go see. And then there's also the bad stuff you need to be aware of, sometimes just to know what's happening. Other times there may be actually dangerous stuff you need to be aware of like I've dealt with physical stalking and—


PF Sure. Somebody's posting your address, yeah of course. But tell us before we get too much further into Block Party. Can you give us a minute on you? Like, how did you find yourself in a position in which you became the CEO of a company that's trying to make the platforms actually usable? And isn't that their job? But how did you get to this place?


TC Sure, there's a few bits of my background, which are pretty relevant here. One is, I'm a software engineer, I worked at a number of the different platform companies. So I worked at Facebook very early 2008, then I was at Quora as the fifth person on the team. And at Pinterest, when it was roughly 10 people up until then it was about 1000. So got to see that early stage of some of these social platform companies, and was pretty involved in like seeing how they design their policies and developed. That's the first part. Second is I've been working on diversity and inclusion advocacy in the tech industry for a similar amount of time, 10 years, and in that period of time, have garnered enough of a platform talking about what I think are reasonable things, but apparently are not reasonable to some people. So I have had to deal with quite a bit of harassment over the last 10 years. And so now I am trying to bring all these things together and build solutions for myself and for other people who are dealing with similar problems.


PF I love the engineering aspect of this, which is like I've had this series of experiences, and now I will build a platform. Which, which I'm sure this must occur to you, right? Like you're building another platform to solve the platform problems. First of all, why the hell don't they do it? Why do you have to start a new company to hold them to account? Do you have a hypothesis on why no one can get their act together?


TC I think there's a few aspects of it within the platform companies as they exist now that make it difficult for them to solve it. One is that their incentive structures and the initial design of the platform is kind of all wrong. So it's hard to retrofit solutions in. Oftentimes, the people who are there don't have the same kind of understanding of harassment and unpleasant stuff to try to build it in instinctively. And I think now what we're starting to see is like the platforms are so large, it's actually very difficult for them to build platform wide solutions. And so I think, actually, the way it makes sense to me, for us to move as ecosystem is that platforms are responsible for some amount of moderation, some standards of safety, and then the more they can open up like Twitter has to enable Block Party to build on top of it. I think, like if if other platforms open up more so, then it just creates a much bigger space of consumer solutions, right, as people have more control over what they want to see.


BP Yeah, you know, it's interesting, you said you were at all these places early on, on the engineering side, and you saw how they were building what the incentive structure was, in 2008, Facebook or slightly after Quora, or Pinterest, was the idea of building in moderation, safety, anti harassment, part of that initial stage? Or that only comes later once the platform is scaled, and they realize it's a problem? Like, if I were to build a social network now, I would hope you know that I, seeing what's happened, you would start there. But did any of those, you know, I initially begin with that as a piece of the puzzle?


TC It's pretty rare that they would think about it early. But one of the very formative experiences for me in my career. So Quora was the first job I had, when I graduated, my first full time job. And the first thing I wanted to build when I got there was a block button. Because even though the site I think only had a few 1000 users, there was already somebody who was bothering me. And I just really wanted to be able to stop him from doing things I would send me notifications. And so the first thing I did was build this block button and make this person, the first person blocked on Quora, which is very gratifying. And also, it was just a learning to me that the people who are on those initial teams have so much influence over how they evolve. And because I felt very strongly about this, I was able to push this through. But it also indicated to like most places don't have someone like that early who are pushing for anti harassment protections because they're not experiencing the problem and so they don't feel it so strongly viscerally that they're going to champion it as something they work on.


BP That makes sense. And I just got it. Block Party. I got it now.


PF Oh really?!


BP I didn't really know but now I got it, Block Party. I was actually gonna ask later like, are you trying to recreate a block party? 


PF No, no. Look, I mean, the fellow dudes who are out there listening to this, if you have never gone into the mentions—so like, I have a nice social media following people. Here's what people say to me. "Happy birthday!" "I disagree." "Good to hear from you." You go into the mentions of a prominent woman in technology, go just spend like some tim, I've done this, like go spend some time, 'cause you're like, no, I can't be that bad. Holy banana cakes! It is just poisoned treachery. Women of color in particular. It is a disaster zone in there. And what you start to realize is like wow, this is a part time job for someone to deal with people just hating them all the time. And this isn't me being like, oh, look at me being an empathetic ally. It's like, Oh my God! That's a different product that has nothing to do with my experience of social media. This is badly broken. Someone needs to fix this from product or possibly start a company. When did Block Party start?


TC About two and a half years ago? Late 2018.


PF So how much, so CEO of a startup, how much are you programming?


TC Right now, I'm doing a lot of programming, more than would be ideal probably.


PF Alright, well talk to Stack, Tracy. You're on the Stack Overflow—what's the Stack?


TC Yes. So we're building on top of Render, which is a platform similar to Heroku. Our back end Postgres database. Python flask back end.


PF Good stack. It's a friendly stack that people can learn.


TC Yeah, lots of asynchronous—yeah.


PF Every engineer I work with just just stood up and threw something in my head. But like, it doesn't matter. God, it's good. Flask is so good. Okay, sorry.


TC I really like Flask. Yeah. So our back end, we do a lot of asynchronous task processing. So we're using celery for that RabbitMQ Redis. The main application logic, like I'd mentioned, flask, we use SQL alchemy OM to connect to the database, graph qL API. And then the front end is a little bit of a mishmash, we have both an older school Jinja templating, for some easy front end stuff. And then we have react for the more interactive things.


PF Never apologize for Jinja. That's like the greatest stack in the world. I love it. And I'm going to send a resume. That is, no one's allowed to use that stack anymore. Unless they're the CEO and they say, this is how we're gonna do it. Because everything has to be all JavaScript all the time. Great. Okay, so I can go in and I can, I can sort of say like, Hey, here's my Twitter account, I would like people to behave a little better, or actually, I can't make them behave a little better. So at least filter that stuff out for me.


BP Yeah, I was gonna say like, I was curious, because, you know, I have blocked words and stuff from Twitter. And I know that they have settings, you know, what do you want to see? So there's a layer that that Twitter has created, which obviously, you know, I know, from, from what I hear in public is often insufficient, but like, do you build on top of that? Do you sit alongside it? Why do you think like the ability to block words and filter comments that Twitter provides is not enough? And what does Block Party do that's different?


TC Yeah. So back to your earlier question I wanted to answer it's not just that you want to filter things out, sometimes you do actually want to go see it later. Being able to control when you see is still very powerful. So not having to see abuse happened real time is great, you can isolate it to when you're ready to go look at it. So it is important to be able to see that which is why just the notification settings that Twitter natively provides are often not enough. What's nice also about having a separate folder that is on Block Party that you can review later, if you want as also you can delegate access to somebody else review for you. So you can start to tap into the community aspects of reviewing content on your behalf. So in the worst cases of harassment, we've heard from folks, and I've had this as well, where it's just too traumatizing for you personally, to go look at it. So you'll have somebody else review for you to make sure there's nothing you need to be concerned about.


BP Right. Right. That makes sense. I guess, are you using any kind of semantic or sentiment analysis when you're doing this? Like you're deciding to block certain key phrases or words? What is it you know, that allows Block Party to decide this is something I'm going to let through now. And this is I'm going to put in the look later folder?


TC Yeah, we're not doing any semantic sentiment analysis right now. We actually get asked this question a lot, if we're using AI. I almost want to introduce AI into our stack just so we can say that we use it.


BP What's your machine learning strategy?


TC Talk about machine learning as if we use it. That's our strategy. So we're actually just filtering on pretty straightforward rules right now things that are characteristics of the users. So you can have a lighter, more permissive filtering, which says just filter out people who have new accounts, don't have a profile photo of fewer than 100 followers, those kinds of rules are pretty similar to what Twitter has. You can also have more restrictive filtering that says, by default view, anybody who at mentions me, unless they are somebody, I follow somebody followed by someone I follow. So kind of like in that social network, verified user or somebody I've interacted with recently. So these rules actually work pretty well. And I think this is one of the key insights. Which is like, you don't have to do really fancy machine learning to get filtering that works pretty well. And also, because of the construct where you can still go see everything later, the cost of getting the wrong decision is not that high. So we don't have to get it perfectly right. To get like whatever 99% accuracy, because if we get it wrong, you just see it later. And that's usually fine.


BP No, yeah, you'd have to do machine learning do well, that's so you can raise the VC and become a unicorn. The machine learning is just you know—


TC What's funny is like I actually had machine learning based tools for moderation before. So when I was at core, I built a bunch of moderation tools. Some of them were based with machine learning. And I worked on machine learning things at Pinterest, like on the home feed and recommendations teams. And I studied, I have my Master's in AI. So like it's not as if I don't understand how to do these things. But it's sort of like cost benefit analysis of like an early stage startup where we want to invest our engineering time, machine learning is not it.


PF Well, I just like this, this tactus gets better and better as far as I'm concerned. So like easily understandable heuristics and Python just sounds like the greatest day, like no wonder you're still programming. That actually sounds like, yeah, my God, something got through, let's figure out why rather than like, you know, cross our fingers that the the model trained on evil will come up with the right probability.


BP Yeah what's inside the black box of evil catching? We don't know, change the lights.


PF This is a funny thing, right i think this is a great moment to point out that like not having that black box is really valuable in this context, which is if we understand and can really negotiate and talk about the heuristics that we're using to block people, we have a more transparent model that's going to be better for your users longer term right for them to kind of be able to understand, yeah.


TC The understandability is super key. Eventually, it might be nice to plug in some machine learning, but still be able to understand a bit of what it's doing. I think that is kind of an active area of research right now within machine learning, especially with like deep learning and these neural nets where you've got to just like construct your models, and then let them train without really knowing what's happening. And it can be pretty alarming to see what the outputs are, when you don't know what the training data was, what biases they're picking up what exactly they're doing. So it was kind of nice if we've just sidestep that for now, we don't have to worry about understandability of some blackbox model, our rules are super simple, they should be easy for people to know which ones are working or not working, and they want to toggle them on and off. It's all very easy.


PF I've hit a problem that you have figured out a way to overcome. I did a weird sort of rd project years and years ago, it's called Anxiety Box. And the idea was that you would tell the computer, your anxieties, and then it would actually send you really mean emails all day. And the joke was that anxiety is like spam, it just kind of is like "Hey, man, looks like you didn't finish your book. I guess that's because you're a terrible person looks like he didn't get your work done. It looks like you—" And he would send you like 12 or 13 emails a day. And then you would start to like, kind of, for me, I had to do this, it was really therapeutic. Because I was like, oh, my God, my anxiety is really stupid and terrible. And then it got picked up on This American Life. And people were like, how do I use this? So I set up a Google Form. And I was like, go ahead, I'm going to build this thing. Tell me your anxieties. And I ended up with like 8000 people putting their anxieties into a forum. And I'm like, I can't build this, I own, I now own 8000 people's greatest fears, I have to send them emails about them. I'm not a medical professional. Like, I'm like, how can I mess with all of this trauma safely. So I really kind of had to stop the project, because I'm like, I don't have the tools, money or time to deal with this kind of human vulnerability in a way where I can guarantee people's safety, security, and so on, especially not for a weird art project. So let me throw that back to you. Right. So like, I hit a wall, and over in the bowels of a Google spreadsheet, lurk that data set, which I really need to just blow away at some point. It shouldn't exist in the world. Here you are with a lot of rules and a lot of knowledge about people's, the worst stuff that's happening in the world and you are taking on a responsibility for their safety. They're coming to you and saying, "I can't deal with this anymore. This is just too much. Can you please protect me from the larger world?" So talk a little bit about that, you've lived it like how do you build a product that takes care of people, instead of you know, the typical product strategy is we're going to aggregate as many human beings as possible? What is the, what sort of product strategy, like helped you along this path?


TC It's pretty simple. We just try to think from the perspective of a person who has to deal with this stuff, and what would be helpful to them, we don't have to promise to solve the entirety of the problem. If we just make it a little bit better, it's still good because the bar is so low right now. It is also relatively easy for me to work on this product, because I am in the core demographic. So this is one of those things that Silicon Valley startup lore likes to say, like build something that you use yourself. So I have some intuition from just dealing with this problem myself, where I've noticed that traumatic impact on me of seeing trash on Twitter directed at me in my mentions. I can feel the mental health impact of it. And so just realizing that if I can just filter it to a later time to see it. I already feel so much better. And I don't get that mental health like doing that already helps a whole bunch, and it doesn't have to solve the whole problem. Like there's still a bunch more that I need to deal with. I still have to go to the police when I have more severe stalking and harassment, like Block Parties will not solve that problem. It's a human problem as well. I think that's something pretty key to realize technology is not going to solve all human problems. It can help in some regards, but was that little bit like, yeah.


BP Yeah, it's not going to solve the problems it creates, yeah.


PF So wait, let me ask you a question. This will be rattling around in people's brains. I have my own answer for it. And I don't I don't think it's a fair question. But I'm gonna ask anyway. Why the hell not just get off Twitter? You know, why do people continue to invest in these platforms, they get to the point where you have to build a tool like this?


TC When people have told me in the past, like, why don't you just get off Twitter, it feels to me the same as like when I complained about street harassment, if somebody were to say, "well just don't walk on the street!" like that's not a practical solution, like this is the world we exist in, especially now with COVID. And everybody being remote and working from home, like our lives are all digital, like, these are the spaces that we inhabit now. And they're not less real for being digital and online, there's a lot of value in being in these places, being on these platforms, I personally got a ton of value being on Twitter. That's why we're there in the first place. So to say we have to cut out all the good stuff, because there's bad stuff doesn't seem reasonable.


BP Tracy, let me ask a few questions about the business itself. So how many, how big is it now? Like you're the founder, CEO, and lead engineer, who else is on your team?


TC We have a very interesting distributed team with like some fractional folks. So it's a bit harder to say like, here's a number of people, but I have a COO who does a lot of things, all these titles, I kind of only capture a tiny bit of what everyone does. You have two other engineers, right now. We're part time slash contractors, part time, contractor designer, and then I have an EA, who does also more than just EA type things.


BP Right, right. So what's the vision that you have for sort of scaling it up? Like are users going to be able to use Block Party, see the benefits and then sort of build on top of it, you know, build additions or expansions that help them uniquely them or different social network or as you gain, you know, more users are you hoping to build the company out in a traditional way?


TC So the thought around our product roadmap is like a short term, build out more functionality on Twitter itself, which is our first platform supporting, we want to go cross platform after that understanding that abuse and bad stuff often will go across multiple platforms, and people are dealing with it have to deal with it in many places. And I think some of the logic around who you want to filter can be transferred between platforms. Longer term, we don't just want to be building on top of other platforms, API's, we would love to be the ones providing API's for other folks. So in the way that like Stripe builds payments APIs, and makes it easier for anybody who needs to take payments to integrate with them. And because stripe does a really good job defining the constructs and telling you like, here's what your model should look like, you should have customers products, prices, like here's how you should model the system, it actually pushes people integrating with them to think more crisply about what they're doing. And what block party can be is sort of like those safety and moderation API's that also help to push people who are building new platforms and thinking about these problems in a different way, hopefully, in a more considered thoughtful way.


PF So sort of the client platform for people building new social experiences, like just like, we should plug into Block Party.


TC Yeah, here's how you should be thinking about user preferences and safety and how they want to control their experience online.


BP I was just gonna say when I was thinking even about the little dog park app I was gonna make and I was talking to a few people, you know, how do we want people to auth in? And how would we do safety? They were sort of saying, Oh, yeah, you know, you could give them these various options. Do you want to log in with Facebook and Google and you know, what do you want to do? And so right, if you're building for the beginning, your show saying, like, I'd love trust and safety to be a part of this, you could have you know, that in the background is one option that the user selects. You know, I want to make sure from the beginning, I'm using this for my like, personal filtering.


PF But Ben that is a new conversation that we really like where you're learning to code and getting your—like five years ago, no one would have said trust and safety in a typical conversation. Lots of people would have, but they didn't have big voices at that moment, they would have said, how do you create a dog park app that grows to every dog park in the world? And so this is a new conversation brought on by human behavior? What if we rebooted, right, what if we could start Twitter again today in a magical wonderful world in which the last 1314 years didn't happen? What would you start with? How would you How would you build a platform from scratch, a social platform?


TC That's a really great question. I have not thought about this very deeply before, I kind of just like assuming that this is the world we're stuck with. I think a lot of it had to center around user control and being being able to say, here's who I want to be able to hear from, and a lot more of these protections around, like what can generate notifications for me, what types of stuff people be allowed to send me and maybe actually creating some more friction in the UI. So I think one of the downsides it's a good thing, but also a downside with Twitter is that you just get touch real time responses and everything is so prompt to like quick reply to anything you see. But that lack of friction also reduces for like thoughtfulness of interactions. So increasing some of the friction probably be part of it.


BP And then you know, it's so hard because right, you know, to sort of gather resources and importance you got to scale really fast. So you make it frictionless, but then if you actually wanted it to be like safe and usable for long term, you'd introduce the friction. And when I see, you know, I go to retweet something. It's like, "have you read this article?" And I just, I laugh, I laugh. "Are you sure you want to, quote tweet before you click the link?" It's like agh, are you kidding?


PF They're doing that now. I like when it's an article I've written. And I'm just like, yeah, I've read the article. That's just so this is the this is the great paradox, right? We wouldn't live in this world if that friction had been there, because they wouldn't have hit that scale of that sort of that infinite explosion of everyone yelling at each other. And I think for a lot of people, it's just so stimulating and fun early days, and they're not getting yelled at. And they're, you know, they're sort of like just kind of watching. And, you know, the worst thing in the world is watching new users come on, and realize they can just be terrible, and kind of get away with it. And then, you know, and it's sometimes their 14 year olds, and you're just like, Oh, god, what have we created?


BP I had another question, which was, you know, you're sort of like imagining what could it be like, if we built the back from the beginning? I mean, one thing that is really kind of up in the air now, which I'm very interested in is Section 230, sort of being rethought, you know, at a high level and whether or not that was the right decision whether or not that should continue to be the law of the land. Do you have a perspective on that like, to what degree internet platform should be accountable for that you know, what their users post?


TC I think there should be more accountability and better governance structures. I don't think Section 230 is the way we're going to solve this problem. So my take on it is kind of like we've evolved, this whole new society online interacts intersects with offline world without having a system of governance for it. And so everything that's happening online is kind of in this lawlessness. And what we're trying to do is use offline world regulation to create some accountability for online behavior. But it's really only in the most extreme cases, like a murderous insurrection at the Capitol, where people are held to some kind of account, but it's trying to shoehorn, like a totally different type of interaction society into old school of law. And what we, I think need to do is completely reimagine what system governance looks like where you have to understand that you can't have a system of governance per platform, because people are on all these platforms simultaneously. They have to intersect with offline worldviews. You can't just do some online and then organize offline and do bad things offline. That's okay to have to intersect understand these intersections. It is also tricky though, because if you look at offline world governance, and what laws apply to you, usually it's based on where you physically are. And there's only one place you can physically be at any time. There are some things around like extradition. So if you're trying to like flee justice, there's there's ways to handle that. But online, you can be present so many places simultaneously. So what are the rules that apply? What are the those intersections? We can already see people who are pretty clever at navigating Terms of Service from like one platform to the next. Or they'll do things that are not quite bad enough to get them banned from one but they'll point to bad stuff happening on another platform and organize around that. And so I think our our thinking around what governance looks like it's just too early and not grand enough.


PF How could internet jail work? Right? Like, this is the hard part. Where do you—meta filter always had the concept of like the timeout. Like it was just like, you're gonna go away for a week. You need a little break.


TC Yeah, I think these different levels of accountability are important. So it's not just like all or nothing like you're completely banned, or it's all fine. And in the offline world. That's true, right? You don't just get put into prison for life for a small infraction, like there are these different levels of like fines or jail time or like—


PF It depends but yeah, that's a different podcast. We'll do that one after this one!


TC Yeah it's a big topic of conversation with the criminal justice system and all the issues there.


PF But this is what's tricky. It's all really blurry. And it got incredibly blurry in the last four years, when the most powerful person in the world was the greatest Terms of Service violator also in the entire world. Just not to be elliptical. Like I'm talking about Donald Trump like it was just like he did he did everything Twitter said you weren't allowed to do but they couldn't kick them because of the existing power structures in the world that they were trying to reflect on. I mean, it's like they should have you know, I like we I think we all knew it like four years ago, like whoa, this is gotta end and it didn't. So you can see like how messy and big and real world this is. And we're just I think what's so tricky with law is it's slow and the internet is, the metabolism of the internet is very, very fast and there is no reactive mechanism that is quick enough to contain harm at scale with Internet Governance like I don't I don't think I've even seen one proposed, right, like what is that? Because you almost need like a big red like 'whoa, emergency shut it all down' button because they figured out some new way to make the memes racist and then how are you going to like wind that back to some kind of governance and that's also—


BP Legislation on the blockchain, just structured the incentives, right Paul. Smart contract. Society is just a smart contract between us. I got it Tracy. I got it.


PF Why don't we just put laws in into GitHub, and then we can do pull requests. I mean, it's just this stuff, this stuff is muddy. Nobody likes to fund it. Like it's just, we're just stumbling into the dark. But luckily, there are the new—


TC The other thing is there's there's law. And then there's also like social norms. This is true in the offline world as well, there's things that you're not allowed to because they're violation of law, and then there's things that you don't do, because people would look at you funny, or you would feel like that was not appropriate. And that's another area where online society is kind of failing, like, not only are there not really any laws, besides in theory, these terms of service rules, which are not really that enforced, and then they kind of get ignored anyways. But also, like, the social norms are totally broken, right, the norm is that you can be a complete [censored], and it's okay. So that's a hard problem, as well as human behavior. How do you fix social norms? But I do think we can try to shift that faster than laws or enforcement of laws.


PF The structure that you're choosing to address this right is a business. And I'm curious about that, like I've seen other efforts, which are sort of not for profit, and so forth. Why a like classic startup, to address these issues, which are social?


TC So the way I think about what we're building right now is the harms are being perpetuated on consumer web scale, because we've connected everybody online. And so harassment can happen at this really wide scale. If we only tried to address those harms, in a non digital non scaled way, like we will never be able to actually protect people. And so what we're trying to do block party is like, use the power of technology, to protect against technology, wielded in a negative way. And I think there is a legitimate business here. If you think about just more consumer protection, like consumers want more safety, there's things that are better for people, and there's value to them, they'll want to pay for it. But also just thinking about like what we're building technology, it takes capital to sustain a technology product, like you need engineers, it's not an easy problem to solve, just with scaling. Like if you just think about what we have to do, right? We have to ingest so much data from Twitter to be able to run the protections that we're doing. We have to like, pull all of your mentions, mutes, blocks to be constantly running, like just maintaining all that.


PF They make that a pain. Oh my god. Yeah. So what did the big platform companies make of you? Like I'm sure people have have talked to you and like, are they like, why are you in our world, Tracy? Or are they like, Oh, just please fix it.


TC So far, they've been really supportive. So we're primarily on Twitter right now. So most of our competition with Twitter, but they've been really supportive, like they've spotlighted us in their API announcements to say, like, "oh, we're working with companies like Block Party to design, the next version of our API is to make sure that these things are supported in our ecosystem." And they like the idea there's more people working on solving this problem, I think there's been a sea change in thinking about solving the harms of harassment, where they'll recognize that their stuff they need to do on the health side, but also, there's space, there is a really big problem. So there's lots of space for other people to work on it with them. So that's like a good change.


PF Their product roadmap just seems different than it was a couple years ago. Like they're just trying more things and communicating a little more openly, which good, they would they really need to. So you're building this particular startup. What are the other things you think people should be working on? Like, what if you could scale yourself? What would be the other things that you would be doing right now?


TC There's a lot of social justice, climate justice, issues that need to be addressed. I think a lot of the things that have become super clear, because of the pandemic or all things, we should be putting more resources towards all these kind of like cracks in society. I do think also, there's a lot more room for public private partnership, then that kind of like recent history has suggested is good. There's just so much more room for the government to have like organized efforts that also bring in private sector to address some of these really big problems. So I previously did spend some time with US Digital Service. Like there's a lot of room for tech, civic tech people who have technology backgrounds, be working in government, lots of things.


PF If people want to see an experience this product, what can they do? And also, how can people be helpful? What kind of people are you looking for in the world to get in touch?


TC Sure. So we're a And on Twitter, we're @blockpartyapp_, there are many other block parties by the way, we are not the blockchain one. We're the anti harassment one. The types of folks would be great to get in touch, we really need engineers like every other startup. But yeah, if you think this is important problem, Python, dev, React dev, love to hear from you.


PF And what's the what's the best way for people to get in touch?


TC If you go to we have a contact page. So it's very easy to message us through there or works too.




BP Alright, it's that time of the episode I'm going to shout out a lifeboat badge winner. Somebody who came on Stack Overflow found a question with a score of negative three or less, gave it an answer that got up to a score of 20 or more. Today's winner awarded May 7 is Bryan Oakley, "How to redirect print statements to Tkinter text widget." Okay, if you've ever had trouble redirecting your print statements, we got an answer for you in the show notes.


PF Goooood lifeboatin'.


BP I'm Ben Popper, Director of Content here at Stack Overflow. You can always find me on Twitter @BenPopper, you can always email us


PF I'm Paul Ford, friend of Stack Overflow, check out our company Postlight, and if you're an engineer looking for work, apply to Tracy first, go check out Block Party. But if you'd also like to work at an agency, or you're a product manager or designer, whoo, boy, I'd love to talk to you just get in touch check out that website. Click on some links.


BP Alright.


TC I'm Tracy, you could find me on Twitter @triketora and Block Party on Twitter @blockpartyapp_ underscore and our website is


PF Oh it's trike—okay, I thought it was triktora. Alright, you learn something every time you're on this podcast.


TC It's a made up word. It's okay.


PF Still, now there's an official pronunciation. You heard it here in the Stack Overflow Podcast.


[outro music]