# Bluesky Scraper - DTNS 4905  ## Episode metadata - Episode title: Bluesky Scraper - DTNS 4905 - Show: Daily Tech News Show - Owner / Host: Tom Merritt - Episode publish date: 2024-11-27 - Episode AI description: A tech giant bids farewell after 38 years, leaving a significant legacy in gaming. Controversies spark as artists backlash against OpenAI's AI tool, raising ethical questions. Meanwhile, a new dataset emerges from Bluesky's posts, showcasing the evolving relationship between social media and public sentiment. Updates on tech innovations include Spotify's new features and TikTok's user restrictions. The podcast also discusses the necessity for better collaboration between artists and technology to navigate the complex landscape of creative expression. - Duration: 30:01 - Episode URL: [Open in Snipd](https://share.snipd.com/episode/0fdf77d7-aeac-4227-995a-0873aef0cf20) - Show URL: [Open in Snipd](https://share.snipd.com/show/a6aadb87-70da-40ba-8247-c06ce1f1d8f3) - Export date: 2026-02-17T23:21:49 ## Snips ### [Blue Sky Data Scraping Incident](https://share.snipd.com/snip/2880d652-b10e-4234-ad4a-5a122965b554) 🎧 11:00 - 12:47 (01:47) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/d26f9952-4e91-46b9-8724-003b7d404792" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - A Hugging Face employee, Daniel Van Strian, scraped 1 million Blue Sky posts for machine learning research. - Facing backlash, he removed the data set and apologized for his oversight regarding data consent. #### 💬 Quote > Somebody Made a Dataset of 1 Million Blue Sky Posts for Machine Learning Research. > — 404 Media 404 Media on the Blue Sky data scraping incident. #### 📚 Transcript **Tom Merritt:** 404 Media published an article Tuesday evening titled, Somebody Made a Dataset of 1 Million Blue Sky Posts for Machine Learning Research. That was the first post. A machine learning librarian working at Hugging Face, Daniel Van Strian, released a dataset composed of 1 million blue sky posts complete with what they were posted and who posted them intended for machine learning research. Now, Van Strien initially said when sharing the data set, this data set contains 1 million public posts collected from blue sky social's Firehose API intended for machine learning research and experimentation with social media data. Each post contains text content, metadata, and information about media attachments and reply relationships. Now, Van Strien has since removed the data set due to some backlash of people saying you shouldn't have done this. Posting on Blue Sky, quote, I've removed the Blue Sky data from the repo. While I wanted to support tool development for the platform, I recognize this approach violates principles of transparency and consent in data collection. I apologize for this mistake. Okay, so I thought that this would be a good jumping off point to talk about what we how we feel about scraping of data. Blue Sky, obviously, has been in the news as of late, because a lot of people have joined the network, you know, trying to figure out if it's, you know, if it's their new ex or a supplement to that type thing. But scraping data is controversial. Scott, what do you think? --- ### [Data Scraping Ethics](https://share.snipd.com/snip/5ee95d08-7449-4243-ab9a-1638c586b70a) 🎧 12:47 - 14:40 (01:52) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/4637c2cb-c241-46d4-ab27-40c117f98c4f" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Scraping public data is a complex issue with ethical considerations. - It highlights the ongoing discussion about data privacy and usage in the digital age. #### 💬 Quote > Blue Sky, Twitter, anyone, nobody can really do anything about public data. And I don't think this guy had any [...] ill intent. > — Scott Johnson Scott Johnson on data scraping on Blue Sky #### 📚 Transcript **Sarah Lane:** Well, I think that this is an interesting story, because I think the natural inclination for a lot of people is to go, oh, here we go again, another platform, and this one happens to be having a moment of growth, is going to share your data. There's this sort of feeling, right, to jump on those And this is not what's happening here at all. Really, this is just a story about somebody who went to a public website that has public data, scraped that data, and is now using it for his research. And people are more going after that person than Blue Sky. And Blue Sky's response is sort of like, well, it's an honor system thing, and there's not a lot we can do. And I agree with them. Like, it's just the Internet. There's stuff out there. You're going to go scrape it and people are going to do it with or without your permission. We don't have laws to protect this if there need to be any at all anyway. And it is just the way of it moving forward. So in this particular case, I don't want to just say it's a much ado about nothing, even though it kind of is in some respects. But I think it is good that it does continually bring up this discussion of is how we want to do this? Are we okay with this? Open and free data for a long time was just a built in part of the internet. And it was why it was so great. And we all embrace that. Well, now we have an open and free internet that can now be scraped and used for other things. Sometimes things that you'll have no credit in the work you that you did that it got scraped or whatever it is. So there's questions now, right? We've changed the conversation has been tweaked. And I think it's good to have it as often as we need to have it. But I think in this particular case, it's just illustrative of the idea that blue sky, Twitter, anyone, nobody can really do anything about public data. And I don't think this guy had any like ill intent. I think everybody's trying to do these studies right now. I'll bet you money. There are other versions of this same data sitting on somebody's research drive and they're just not telling us. --- ### [User Data Consent](https://share.snipd.com/snip/13f8aa42-3444-4bca-83ef-169d031acf1d) 🎧 15:41 - 17:53 (02:12) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/e1d25d53-75ef-4459-8df9-f930127967c3" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Consider allowing users to express their consent for data usage. - This approach respects user preferences and promotes ethical data handling practices. #### 💬 Quote > Blue Sky has the right idea, which is like, let's give people a way to say whether they consent or not. So the people who want to respect that can. > — Tom Merritt Tom Merritt's advice on handling user data consent. #### 📚 Transcript don't think people really, let me, this. There's a lot of people who are just against stuff because that's what you do. I'm against people collecting AI. I wonder what percentage of people in this data set really objected. really would have objected. Blue Sky's promise is that this is an open system where your posts can be migrated from server to server, right? It's a Fediverse-like approach, just like Mastodon. This is not really scraping. Scraping is when you go and you get a website and you scrape the data off of it. This is using the API, the API that is meant to be open for people to access posts and be able to do something. It's one of the foundations of Blue Sky. So it's not just that Blue Sky is like, well, it's the honor system. There's not much we can do. It's like, this is how the system was designed. The other thing Blue Sky said was, we are looking into ways to let users express whether they give consent or not, so that someone who does want to play by the rules can say oh well i'll only use the posts where someone is given consent to be trained on it uh i also don't know what damage comes from your public blue sky post being used in a machine learning uh training data but that's probably a conversation for another day but if you want **Tom Merritt:** to if i if i were like i'm sarah lyn and here's my address, and I like somehow also it's machine learning. It's not my large language models. Usually people are like, well, you're going to, you're going to, you know, imitate me. This isn't a large language model. It's a machine learning situation. So it's entirely different. I just, I don't know. I don't want to go on and on and on about this, but it's the same thing I've been saying, which is it's not illegal. And like Scott said very eloquently, if you want an open Internet, you have to allow it to be open to be used. And I think Blue Sky has the right idea, which is like, let's give people a way to say whether they consent or not. So the people who want to respect that can. But until we get a new law about whether public data is acceptable for use or what the rules using it should be, you're going to have a lot of these kinds of situations. --- ### [Artists' Protest Against OpenAI](https://share.snipd.com/snip/2ad4ab7a-43d1-4ade-8db2-ddf737831a5c) 🎧 18:23 - 20:25 (02:01) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/2931a01e-24d5-436d-a577-d549276a54a4" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Artists with early access to OpenAI's Sora released a limited public version. - They criticized OpenAI for using them as "PR puppets" without proper compensation or respect for their feedback. #### 💬 Quote > we believe instead we are being lured into ‘art washing’ to tell the world that Sora is a useful tool for artists. We are not your free bug testers, PR puppets, training data, or validation tokens. > — Group of 19 Artists Group of artists criticizing OpenAI's practices. #### 📚 Transcript of big AI companies that people get mad at, on Tuesday, a group of artists who had been given early access to OpenAI's Sora video generator released a version of the tool publicly on Hugging Face. Hugging Face and another story. This version used their tokens, in other words, the artists' tokens, the artists who had been given private access to give anyone access to Sora they essentially published a front end that used their tokens their tool could only generate up to 10 seconds of video at 1080p though support supposedly Sora can generate up to 60 seconds along with the release 19 artists wrote an open letter saying and I quote we received access to Sora with the promise to be early testers, red teamers, and creative partners. However, we believe instead we are being lured into art washing to tell the world that Sora is a useful tool for artists. We are not your free bug testers, PR puppets, training data, or validation tokens. The artist also accused OpenAI of using them as unpaid labor, pointing out that the only compensation anybody would get was that a few of the testers would have their creations screened publicly. They added, we are not against the use of AI technology as a tool for the arts. If we were, we probably wouldn't have been invited to this program. We are sharing this to the world in the hopes that OpenAI becomes more open, more artist-friendly, and supports the arts beyond PR stunts. OpenAI by saying, like, look, this is still in development. It's a limited number of artists, but they have free access. We don't tell them what they can do or even if they have to do anything. They suspended all access to Sora temporarily while they figured out how to, you know, shut down this public one and claim that only a few of the artists who signed this open letter were actually part of its early access program. But Scott, you're an artist. I'm curious what sense you get from this is these are people who like, we like AI. This isn't an anti-AI activist move. We're just mad at the way open AI is treating us. --- ### [Artists' Perspective on AI](https://share.snipd.com/snip/6140faaf-d71f-44b6-964a-0d9d151a8bc0) 🎧 20:25 - 24:50 (04:24) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/c3e46f6e-4d28-4eaa-b04a-aca4b930080c" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Artists' concerns extend beyond simple pro- or anti-AI sentiments. - They desire respectful collaboration and genuine feedback incorporation in AI development. #### 💬 Quote > This isn't an anti-AI activist move. We're just mad at the way open AI is treating us. > — Tom Merritt Tom Merritt summarizing the artists' perspective on OpenAI #### 📚 Transcript Well, **Sarah Lane:** it's a tier of this AI artist thing that I hadn't considered very much. And that is this kind of middle group where, you know, on the one hand, you have an extreme group over here, which I sometimes tip my toe in, which is, man, this stuff stinks. Please stop doing it. This is bad for the future. It's bad for artists, all that sort of stuff. And then there's this other extreme where AI is the future and this is all we're going to use and forget about anything else. This is the new and everything else is old. In the middle there, though, there are some pretty reasonable, more centrist thinking folks who know that some of this technology not only is here to stay, but actually could be really beneficial. Like they say in their own quote, beneficial to the arts, good tools for the arts. You know, it's a tricky thing to navigate because it's not simple. It's not just, well, this one thing is okay and this other thing is okay. Sometimes it's, you know, Sora as a product, Sora rather, as a product may have lots good points, bad points, and middle points. And that's, I think, kind of their point. I think that this is a group I trust their voices more than I would the other extremes. They're either pro or way too anti. The ones in the middle are the ones that are kind of putting rubber to the road. They were also invited to this, or at least some of them, according to OpenAI. And I want to hear, this is kind of what I do want to hear in these processes, not just that, well, we had some artists sign an NDA and they all fiddled with the thing. And on the other end of this, you'll get to hear whether they liked it or not, and they'll either promote it or they won't or whatever. I like that they're speaking up now if they have concerns, because if they have concerns, I'm way more interested in them having them than i am either of the other two extremes so i see this as a good thing and i think that open ai's response is very corporate but not in the spirit at which we're trying to get all this in right now yeah it feels like these artists are saying uh you just want to hear positive things from us and And were trying to tell you it didn't work very well. **Tom Merritt:** That's what the PR puppet part of this sounded like to me. Because it's sort of like, if you have access to a tool that isn't available to anybody else, which we all do from time to time. It's like, yeah, I am your bug tester. I am contributing to your training data i mean that's sort of the part of being in a beta testing system um but uh i don't feel like being part of a pr machine is any handshake agreement that i ever have to have. Um, I'm not exactly sure what the artists, uh, thought that they needed to agree to, or, you know, if there was some sort of a shakedown going on, you know, I don't know, but, but that, that's the part where it's like, yeah, if they, if, if, if this, if this group really felt that way, then that would be something that you would push back on. It feels like they don't feel respected because they're saying, we have a problem with AI. We might have some helpful things to say about whether this tool is useful or not, but they're not listening to us. **Sarah Lane:** Yeah, I think that there's a general feeling among artists when it comes that side of what Opium AI is doing, because their stuff is so multifaceted. And a lot these companies are that now that's not just, hey, we make images, these guys make videos, and we're over here doing LLMs. Everybody's doing kind of everything. And they're trying to cater to this stuff. And part of that is they want to do tools that require some form of, hey, the community of artists, creators, filmmakers, whatever, they're with us on this. And so it does feel a little PR stunty if they're not actually taking feedback, if it's just being administered like a PR move. And even if it isn't, even if OpenAI's intentions are pure in every possible way, this is the area I just think that needs the most work in AI stuff is the, the artists and the engineers. Like if you can't get the artists who like you, who wanted to participate on your side, there's a problem. Even if it's not the majority of them, there's a problem. Yeah. **Sarah Lane:** Yeah. So that's something I think they, they can find, they can find some common ground on this, but again, this illustrates that we're, we're at, we are at a junction point on some of this stuff. And it's interesting to see the pushback. --- ### [Yoshida's Legacy at PlayStation](https://share.snipd.com/snip/d7f3a446-1d3c-40a8-a77c-a812dc5f7482) 🎧 26:46 - 29:11 (02:24) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/5d428171-1b42-4806-89c9-91387e863391" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Shuhei Yoshida, a key figure in PlayStation's development, stepped down after 38 years at Sony. - His leadership shaped PlayStation's success, from its origins to navigating competition with Microsoft. #### 💬 Quote > Sony's current position in the worldwide electronics marketplace is directly affected by decisions that he and his team members made to make them a premier brand in gaming. > — Scott Johnson Scott Johnson on Shuhei Yoshida's impact on Sony #### 📚 Transcript **Tom Merritt:** Well, we don't know. Sometimes you can kind of tell. All right. So very different subject. A key player in the development of Sony's PlayStation game console, Shuhei Yoshida, is stepping down after 38 years with the company. Yoshida was one of the original members of the PlayStation Project back in 1993. He also spearheaded the division's licensing program and in turn became a producer for major game titles like Gran Turismo. In 2000, he became VP of Sony Computer Entertainment America, followed by a role as Senior Vice President of SCE, Studios USA, in 2007. Finally, in 2008, he became the President of SEC Worldwide Studios. So Yoshida has been around the block. Scott Johnson, what do the gamers make of this? **Sarah Lane:** Well, I'm a huge fan of this man and his legacy. think that it is notable to say that, and I think accurate to say, that Sony's current position in the worldwide electronics marketplace is directly affected by decisions that he and his team members made to make them a premier brand in gaming. At a time where there was a real pivot happening between 2D, 3D, all this sort of stuff. And their deal they had with Nintendo, which looked like a pretty good partnership for a long time. He was also involved in that. I think them recognizing that, man, do we really just want to hook our wagon to this wagon and just be part of another person's wagon? Or do we want to be our own freaking video game wagon? And they chose the latter. And it ruffled a lot of feathers. Like, Nintendo, I don't think ever forgave them for it. But it created one of the most recognizable brands in the history of gaming. Tom mentioned earlier we just got Final Word. We've been saying this for decades. But Final Word, that the PlayStation 2 is the largest selling or the most selling console in the history of consoles. 160 million consoles worldwide. Nobody's quite done that. We had thought this was true, but I guess Sony finally confirmed it. And that's just part of his legacy. His approachableness, this guy's willing to go on anybody's podcast, talk anything they want to talk about. Always very open, always extremely friendly. He seemed like somebody who was connecting to the audience and not just a corporate C-suite executive who keeps far away from everybody else. --- ### [Sony's Future in Gaming](https://share.snipd.com/snip/e3888e56-3e54-4e41-b5eb-d319740b4c9f) 🎧 29:11 - 33:20 (04:09) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/ba380e8f-55d9-44d6-a92b-2270df563770" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Yoshida's departure marks a transition for Sony as they face questions about their future in gaming. - He played a vital role in building strong third-party developer relationships, crucial for console success. #### 💬 Quote > They are at a bit of a crossroads now. And a lot of people talk about this, what's Sony's future look like? > — Scott Johnson Scott Johnson on Sony's future after Yoshida's departure. #### 📚 Transcript **Sarah Lane:** That impression was always there during those 38 years. And while Sony has certainly had its ups and downs and its bumps here and there, for the most part, I think his leadership and leadership like it within the company has kept them in a really great place. They are at a bit of a crossroads now. And a lot of people talk about this, what's Sony's future look like? What is the future of hardware and physical media and all these kind of questions, game preservation? These are all questions that will haunt Sony as much as any other company trying to navigate this particular part of the industry's timescale. But he is, without a doubt, one of the most influential characters in the history of gaming. And he'll probably be heavily missed there. I mean, I think in recent years, it was not a lot of hands-on directly. But there are a lot of times during the history of this company, playstation one's launch playstation two's pivotal launch uh three and the problems they had and then crawling back to a leadership position in the face of microsoft this multi-billion nightmare company trying to come steal some of their ground um i think you got to give it to him for for navigating a lot of that stuff so you know on one hand i little sad because it's just like old guard going away but also geez well done you you entered the company in your 20s and yeah 60 the playstation is 30 years old he's been on the team for 31 years i mean that's yeah that's **Sarah Lane:** a lot yeah he's been with the company since 86 so even you know he's even got his feet wet with whatever the heck they were working on prior to this and um knowing that he was there before PlayStation became a thing and was responsible for what was happening with the Nintendo PlayStation, which Sony was building for Nintendo and then scrapped, like what a wide range of being there for it, right? Like that's just insane. But I think he deserves the break, and I hope that Gorilla's former CEO, I forgot his name, he'll be taking over. And they're very capable also, so I have a feeling that'll be a good piece of leadership remaining. Not Gorilla Glass, either. No, Gorilla, like the people that make the Horizon Zero Dawn games. But yeah, he's a dude we're all gonna we're gonna think pretty fondly on and there's there's people like this in almost all of these cases we think this way about certainly about Miyamoto over at Nintendo and sure and people at Xbox some of its founders and that sort of thing but uh this is Sony's version of that and even though Sony doesn't play they don't play much in this celebrity game as that as other game companies do. It's still hard not to recognize him for his contribution. A lot of people were mentioning the video that he did shortly after Microsoft's horrible 2013 announcement. And that is probably my favorite memory of him. It's the thing that I know most about him. If anybody doesn't recall, Microsoft had an announcement where they talked about, you know, being able to share your digital games. And it was considered to be a very complex thing of like, you only have three shares and you have to be on the same network when you authorize and blah, blah, blah, blah, blah. And Yoshida did a video where he's like, here's how you share a Sony PlayStation game. And then there's like 10 seconds of step one. And then it comes back to him and he hands it to another guy who goes, thanks. And that's **Sarah Lane:** it. They just smile and that's the end of it. Yeah, they really took advantage of that. Microsoft the ball. They picked up the ball and did really well that generation. They're being asked similar questions now because the landscape has changed. Unfortunately, Microsoft was right about a bunch of this. They just did it wrong and did it in the wrong timing. So we'll see how they navigate it moving forward. But you can't – one other piece of this I didn't mention. You can't deny the effect this guy had on one aspect of gaming, and that is third-party relationships. They existed under Nintendo, but they were stringent and hard and difficult to navigate. If you're a third-party dev not under the direct umbrella of Nintendo, it was a pain. Sony made that way easier and still had high-quality standards and just made it so third parties were as crucial to a console's success as anything under the hood. So, you know, props to him for that as well. Yeah. --- ### [Data Usage at Taylor Swift Concert](https://share.snipd.com/snip/afeb04c7-961d-4b85-a0fb-88f2b0c09c52) 🎧 33:23 - 34:21 (00:57) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/e2f6602e-4848-4a41-9a75-f338bad3dd81" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Ryan suggested Rogers' reported 7.4TB data usage at the Taylor Swift concert likely encompassed all data, not just Rogers'. - This is because Rogers may have operated the distributed antenna system (DAS) within the Rogers Center. #### 💬 Quote > It's quite probable that Rogers owns the DAS [...] in the Rogers Center. All major venues and entertainment complexes use dedicated mobile network designs [...] > — Ryan Ryan clarifying data usage at the Taylor Swift concert. #### 📚 Transcript **Tom Merritt:** Ryan had some thoughts on our GDI conversation yesterday about the 7.4 terabytes of mobile data on the Rogers 5G service used by people at Taylor Swift Show in Toronto's Rogers Center. Ryan says, it's quite probable that Rogers owns the DAS or distributed antenna system in the Rogers Center. All major venues and entertainment complexes use dedicated mobile network designs to maintain service quality during events. Most are independently operated. The one in the Minnesota State Fair is run by MP slash next level, but others may be carrier owned or operated. It's possible that Rogers was reporting the entire data usage in the stadium. It's a good point. We were speculating on whether there was more than 7.4 terabytes, considering that there are other carriers besides Rogers. But Ryan may be right that the 7.4 terabytes may be all the data because Rogers was running the DAS. --- Created with [Snipd](https://www.snipd.com) | Highlight & Take Notes from Podcasts