1 00:00:00,160 --> 00:00:04,360 Speaker 1: This is something that popped up from Digital Musicnews dot com. 2 00:00:04,639 --> 00:00:08,880 Hackers scrape Spotify's entire library. I thought of this when 3 00:00:08,919 --> 00:00:10,839 I saw you know, we were talking earlier a little 4 00:00:10,839 --> 00:00:14,080 bit about Spotify and the importance of being on Spotify 5 00:00:14,160 --> 00:00:16,760 so that your music is discoverable, the term that we 6 00:00:16,839 --> 00:00:19,679 like to use in the industry. But hackers have scraped 7 00:00:19,800 --> 00:00:25,199 Spotify's entire library obtain three hundred terabytes worth of audio. 8 00:00:25,519 --> 00:00:31,800 Spotify says it has identified and disabled the nefarious user accounts. 9 00:00:32,759 --> 00:00:34,719 I hope they do a lot more than that. Just 10 00:00:35,039 --> 00:00:38,920 identifying and disabling those accounts is not going to fix 11 00:00:38,960 --> 00:00:40,799 the problem long term. I mean, if they did it once, 12 00:00:40,799 --> 00:00:42,240 they can do it again in theory. But it says 13 00:00:42,240 --> 00:00:47,200 here hackers say they've scraped Spotify's entire music library, compiling 14 00:00:47,240 --> 00:00:50,960 the metadata behind two hundred and fifty six million tracks 15 00:00:52,000 --> 00:00:55,280 two hundred and fifty six million tracks tied to over 16 00:00:55,359 --> 00:01:00,399 fifteen point four million artists profiles, and intend to make 17 00:01:00,439 --> 00:01:05,040 a massive amount of music available to torrent. Meanwhile, Spotify 18 00:01:05,120 --> 00:01:08,439 is acknowledged to the breach and confirmed that the culprit's 19 00:01:08,599 --> 00:01:13,159 accessed some of the platform's audio files. Some of the 20 00:01:13,200 --> 00:01:16,959 platforms audio files and quotes. So Spotify acknowledging I mean, 21 00:01:17,680 --> 00:01:22,359 you know, fifteen point actually two hundred and fifty six million, 22 00:01:22,439 --> 00:01:26,239 that's some of the platform's audio files. I wonder how 23 00:01:26,239 --> 00:01:28,760 many actual audio files do you ever wonder about that? 24 00:01:28,840 --> 00:01:34,239 Like how much actual music is on Spotify? How many 25 00:01:34,280 --> 00:01:37,359 audio files are on Spotify? Because we were talking about 26 00:01:37,400 --> 00:01:39,480 I remember a couple of years ago on the show 27 00:01:39,519 --> 00:01:41,879 now might have been two or three years ago, we 28 00:01:41,879 --> 00:01:45,239 were talking about how Spotify had announced that they were 29 00:01:45,480 --> 00:01:50,400 starting to remove some music from its platform because people 30 00:01:50,400 --> 00:01:54,519 were uploading you know, demos and really lo fi. I 31 00:01:54,519 --> 00:01:58,239 don't mean lo fi as in the style of recording. 32 00:01:58,280 --> 00:02:00,519 You know, you can you can record low fi music 33 00:02:00,599 --> 00:02:03,400 or music that sounds like it's supposed to be, you know, 34 00:02:03,920 --> 00:02:06,920 kind of like from another time even right, But I 35 00:02:06,920 --> 00:02:10,639 mean actual like low quality, like just poor recording quality music. 36 00:02:10,800 --> 00:02:14,680 People were just uploading, you know, stuff that they just 37 00:02:14,680 --> 00:02:17,120 just cheap demos that they'd made on their tape recorder 38 00:02:17,199 --> 00:02:21,599 or whatever, just uploading all of it to Spotify. And 39 00:02:21,680 --> 00:02:23,759 it got to a point where Spotify decided to go 40 00:02:23,800 --> 00:02:26,240 through and start and I don't know how they do it. 41 00:02:26,319 --> 00:02:28,199 I mean, I'm sure they have I'm sure they have 42 00:02:28,240 --> 00:02:31,360 the technology to detect these things, but actually went through 43 00:02:31,360 --> 00:02:34,360 and started removing some of that stuff. Because you know, 44 00:02:35,599 --> 00:02:39,039 if you think about how much music is being uploaded 45 00:02:39,639 --> 00:02:44,199 to these platforms like Spotify every single day, every single day, 46 00:02:44,199 --> 00:02:47,360 all over the world, people are uploading music, and it's 47 00:02:47,439 --> 00:02:49,560 not like you know, I think we tend to think 48 00:02:49,599 --> 00:02:53,319 of it as sort of this this infinite thing, like 49 00:02:53,319 --> 00:02:57,639 there's just this, you know, because it's because it's audio. 50 00:02:58,039 --> 00:03:01,759 There's just this infinite amount of space to put all 51 00:03:01,879 --> 00:03:06,599 this because it's not they're not physical objects. It's audio files. 52 00:03:06,800 --> 00:03:09,240 So you can just put as many audio files up 53 00:03:09,240 --> 00:03:11,479 as you want and it doesn't matter how much room 54 00:03:11,479 --> 00:03:13,400 it takes up. But it actually does matter how much 55 00:03:13,439 --> 00:03:15,960 rooma takes up because you have to keep building, you 56 00:03:16,039 --> 00:03:18,840 have to keep building more storage for all this music. 57 00:03:18,840 --> 00:03:23,080 It still takes up space digitally, so you can't just 58 00:03:23,199 --> 00:03:25,280 let it be a free for all. Part of what 59 00:03:25,360 --> 00:03:27,039 made me remember that is because I remember at the 60 00:03:27,039 --> 00:03:28,599 time when we talked about it on the show, some 61 00:03:28,639 --> 00:03:31,400 people were upset and it was another reason to be 62 00:03:31,439 --> 00:03:34,560 mad at Spotify. I guess people were upset saying, well, 63 00:03:34,599 --> 00:03:37,000 how can they just go through and remove stuff. Well, yeah, 64 00:03:37,000 --> 00:03:39,199 it's a little bit. First of all, a little bit 65 00:03:39,199 --> 00:03:42,000 of quality control is not a bad thing. You know, 66 00:03:42,120 --> 00:03:45,560 if people are just you know, getting the band together 67 00:03:45,800 --> 00:03:47,879 and sticking a tape recorder in the middle of the 68 00:03:47,960 --> 00:03:53,879 room and then taking that taking that recording and uploading Spotify, 69 00:03:54,280 --> 00:03:57,400 you don't necessarily want that on your platform, right, something 70 00:03:57,719 --> 00:04:01,719 something like that, you know. And also there's there's abuses 71 00:04:01,800 --> 00:04:04,319 that goes on that go on, and you know, people 72 00:04:04,400 --> 00:04:07,000 stealing other people's music and uploading it, and there's all 73 00:04:07,120 --> 00:04:09,759 kinds of things going on. So you know, I'm not 74 00:04:09,800 --> 00:04:14,199 saying that you should never be mad at Spotify about anything, 75 00:04:14,840 --> 00:04:19,160 but you know, but they have a right to police 76 00:04:19,199 --> 00:04:23,680 their own platform too a little bit and make those decisions. Anyway, 77 00:04:24,439 --> 00:04:25,879 that was a little bit of a side street, But 78 00:04:25,879 --> 00:04:30,360 the point is apparently two hundred and fifty six million tracks, 79 00:04:30,759 --> 00:04:34,959 that's only part of how much music is actually on Spotify. 80 00:04:35,720 --> 00:04:42,800 So let's see. So there's an update here again. This 81 00:04:42,959 --> 00:04:47,600 is from Digitalmusicnews dot Com. An update as of the 82 00:04:47,639 --> 00:04:50,680 twenty second It says after this piece was published. Well, no, 83 00:04:50,879 --> 00:04:52,759 let's go back, We'll go back to the update after 84 00:04:52,839 --> 00:04:55,920 let's look at the original article. First. Okay, so, the 85 00:04:56,000 --> 00:05:00,600 allegedly responsible hackers, part of a self described nonprofit project 86 00:05:00,759 --> 00:05:05,639 called Anna's Archive, themselves, disclosed the data heist in a 87 00:05:05,680 --> 00:05:10,360 blog post. In that lengthy post, drawing from the metadata, 88 00:05:10,399 --> 00:05:15,199 covers hard stats concerning duration, stream volume, popularity, genre at 89 00:05:15,240 --> 00:05:21,439 release date, and more regarding straight audio, Anna's Archive indicated 90 00:05:21,600 --> 00:05:26,480 that it had quote archived around eighty six million music files, 91 00:05:26,879 --> 00:05:32,000 representing around ninety nine point six percent of listens and 92 00:05:32,079 --> 00:05:35,759 clocking in at a little under three hundred terabytes in 93 00:05:35,839 --> 00:05:39,439 total size. A while ago, we discovered a way to 94 00:05:39,600 --> 00:05:43,680 scrape Spotify at scale. For now, this is a torrents 95 00:05:43,800 --> 00:05:47,720 only archive aimed at preservation, but if there is enough interest, 96 00:05:47,800 --> 00:05:52,720 we could add downloading of individual files to Anna's archived unquote, 97 00:05:53,000 --> 00:06:01,720 the hackers communicated. That's interesting because when they say that 98 00:06:01,920 --> 00:06:05,720 it's only again this is from the hackers, When they 99 00:06:05,800 --> 00:06:10,959 say that it's aimed at preservation, it sounds like they're 100 00:06:11,439 --> 00:06:15,040 they're trying to present this as an altruistic endeavor. We 101 00:06:15,079 --> 00:06:17,720 want to make sure that nothing ever happens to this music. 102 00:06:18,959 --> 00:06:21,519 We want to make sure again, I'm not please, I'm 103 00:06:21,560 --> 00:06:24,040 not justifying what they did. Don't misunderstand me. I'm just 104 00:06:24,079 --> 00:06:27,120 saying that this sounds like this is what they're presenting. 105 00:06:27,920 --> 00:06:30,360 You know, we want to make sure nothing happens to 106 00:06:30,399 --> 00:06:34,120 this music, you know, because in theory Spotify can remove 107 00:06:34,160 --> 00:06:36,399 anything anytime it wants, or an artists can remove anything 108 00:06:36,480 --> 00:06:39,560 anytime it wants, and then where does it go? What 109 00:06:40,079 --> 00:06:42,120 if they just delete it and then that's it. It's 110 00:06:42,120 --> 00:06:45,680 just gone. It's not available anywhere else, right, So they're 111 00:06:45,680 --> 00:06:48,680 trying to preserve all this music. That's that's what That's 112 00:06:48,720 --> 00:06:53,319 what the hackers have communicated, it says, you're Unsurprisingly, Spotify 113 00:06:53,399 --> 00:06:57,920 and especially rights holders, have plenty to say about those plans, 114 00:06:58,560 --> 00:07:06,199 as noted by Third Chair where had Yayov Zimmerman Yahov Zimmerman. However, 115 00:07:06,279 --> 00:07:12,879 whatever takedowns and illegal actions follow, the damage is already done. Technically. 116 00:07:13,000 --> 00:07:17,360 Anna's archive claims that it doesn't host any copyrighted materials, 117 00:07:17,759 --> 00:07:24,199 instead purportedly indexing metadata that is already publicly available, direct 118 00:07:24,199 --> 00:07:27,879 hosting or not. Some of the project's supporters are lamenting 119 00:07:28,279 --> 00:07:32,079 the Spotify circumvention and the possibility that will quote ruin 120 00:07:32,120 --> 00:07:37,600 the actual important literary archive unquote by encouraging aggressive litigation. 121 00:07:38,360 --> 00:07:41,879 Zimmerman wrote, quote the data is circulating on P two 122 00:07:41,879 --> 00:07:45,360 P networks, and there is no putting the back, putting 123 00:07:45,439 --> 00:07:49,480 this back in Pandora's box. Anyone can now, in theory, 124 00:07:49,879 --> 00:07:54,279 create their own personal free version of Spotify All Music 125 00:07:54,360 --> 00:07:57,079 up to twenty twenty five with enough storage and a 126 00:07:57,160 --> 00:08:01,240 personal media streaming service like Plex. The only real barriers 127 00:08:01,279 --> 00:08:05,560 are copyright law and fear of enforcement unquote you know, 128 00:08:06,240 --> 00:08:08,040 and I'm sure I'm not the only one who this 129 00:08:08,120 --> 00:08:10,920 has occurred, to to whom this has occurred, you know 130 00:08:10,920 --> 00:08:18,079 what this reminds me of. We're going back. We're going back. Jeez, 131 00:08:18,120 --> 00:08:23,439 how many years? Thirty years, maybe twenty five years? Napster. 132 00:08:24,079 --> 00:08:28,800 This reminds me of Napster when Napster first became a thing, 133 00:08:29,079 --> 00:08:34,399 file sharing, peer to peer, file sharing online, and the 134 00:08:34,480 --> 00:08:37,559 whole music industry freaked out because all of a sudden, 135 00:08:37,600 --> 00:08:42,279 everything was free online through Napster. Now, obviously a lot's 136 00:08:42,360 --> 00:08:45,360 changed since then, and the music industry had to figure 137 00:08:45,360 --> 00:08:48,440 out ways to adjust to the new reality and the 138 00:08:48,480 --> 00:08:55,200 Internet and all of it. But when I see this, 139 00:08:56,759 --> 00:08:59,360 anyone can now in theory create their own personal free 140 00:08:59,480 --> 00:09:02,519 version of Spotify All Music up to twenty twenty five 141 00:09:02,559 --> 00:09:06,480 with enough storage, and a personal media streaming server like PLEX. 142 00:09:06,519 --> 00:09:09,600 The only real barriers are copyright law and fear of enforcement. 143 00:09:12,639 --> 00:09:19,480 Let's see, there's a little do we have time. Oh yeah, 144 00:09:19,480 --> 00:09:26,799 we have time. There's a little bit more here. Oh sorry, okay, 145 00:09:26,799 --> 00:09:29,919 here we go. Perhaps more pressingly, in the AI age, 146 00:09:30,399 --> 00:09:34,000 the massive collection of audio could theoretically be used to 147 00:09:34,159 --> 00:09:40,840 train generative generative models and fuel additional unauthorized sound alike outputs, 148 00:09:41,200 --> 00:09:45,240 a particularly significant issue if the involved platforms are based 149 00:09:45,279 --> 00:09:51,759 in countries with inadequate IP protections. I'm telling you, in 150 00:09:51,759 --> 00:09:55,279 some ways, this is napster all over again. This is 151 00:09:55,399 --> 00:10:01,639 napster in the AI age. It says here. One section 152 00:10:01,679 --> 00:10:05,200 of the Anna Archives site says, quote, it is well 153 00:10:05,279 --> 00:10:10,000 understood that lms, which is large learning models like you know, CHATCH, 154 00:10:10,000 --> 00:10:14,320 GPT and others that suck up all this information. Okay, 155 00:10:14,320 --> 00:10:19,559 it is well understood that lms thrive on high quality data. 156 00:10:19,679 --> 00:10:24,559 We have the largest collection of books, papers, magazines, et 157 00:10:24,639 --> 00:10:27,679 cetera in the world, which are some of the highest 158 00:10:27,799 --> 00:10:33,360 quality text sources. Unquote. According to the same site, Anna's 159 00:10:33,440 --> 00:10:38,200 archive promptly put out the metadata and with three hundred 160 00:10:38,279 --> 00:10:43,120 terabytes worth of audio files, releasing an order of popularity. 161 00:10:44,279 --> 00:10:47,399 In other words, the full extent of the episode's fallout 162 00:10:47,519 --> 00:10:52,080 remains to be seen. And as initially mentioned as Spotify 163 00:10:52,159 --> 00:10:55,600 confirmed the unauthorized access, but not where things go from 164 00:10:55,639 --> 00:11:00,879 here in a detailed light. In a detailed light statement, Yeah, 165 00:11:00,879 --> 00:11:03,840 they didn't really say much. The Spotify spokesperson said, quote, 166 00:11:04,000 --> 00:11:08,759 an investigation into unauthorized access identified that a third party 167 00:11:08,799 --> 00:11:14,399 scraped public metadata and used illicit tactics to circumvent DRM 168 00:11:14,480 --> 00:11:17,639 to access some of the platform's audio files. We are 169 00:11:17,720 --> 00:11:23,320 actively investigating the incident quote. And again, as I pointed 170 00:11:23,360 --> 00:11:30,360 out earlier, what is missing from that statement nothing about 171 00:11:30,840 --> 00:11:33,879 how we will prevent this from happening again.