October 6, 2025 • 41 mins
Dive into "The Gray Files" and uncover the frightening reality of your digital footprint. In this eye-opening episode, we expose the dark underbelly of data surveillance, from U.S. government agencies like ICE using social media to track individuals, to employers, and insurance companies analyzing your every click. We pull back the curtain on proprietary algorithms and black-box AI systems that build secret profiles on you, scoring your life for risk, and behavioral patterns. Discover how companies like Palantir, Clearview AI, and LexisNexis turn your posts, likes, and shares into intelligence used to deny you jobs, insurance, and even freedom. Learn about the "Right to Be Forgotten", the myth of online deletion, and why the future of privacy isn't dead, it's just a commodity for the wealthy. This isn't science fiction; it's a deep dive into the real-world consequences of living in the age of algorithmic scrutiny. Tune in to understand your own digital shadow and the urgent need for a new conversation on data privacy.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker:
Welcome back to episode eighteen
of The Gray Files, where we peel (00:04):
undefined
Speaker: back the layers of technology, economics, data science, and (00:09):
undefined
Speaker: even the human condition itself, all in an effort to try and (00:15):
undefined
Speaker: understand this vast and often perplexing world we live in. (00:20):
undefined
Speaker: I'm your host, Erika Barker, and tonight we are talking about how (00:26):
undefined
Speaker: you and everyone you know are being watched, not from a dark (00:32):
undefined
Speaker: van down the street with sophisticated surveillance (00:38):
undefined
Speaker: equipment, but by algorithms. (00:42):
undefined
Speaker: We're examining how the everyday (00:46):
undefined
Speaker: act of posting, liking, and (00:48):
undefined
Speaker: scrolling turns into something (00:51):
undefined
Speaker: else once machines start keeping (00:53):
undefined
Speaker: score. (00:57):
undefined
Speaker: It was a bright cold day in (01:06):
undefined
Speaker: April and the clocks were (01:09):
undefined
Speaker: striking thirteen. (01:12):
undefined
Speaker: That's the famous opening line (01:15):
undefined
Speaker: to Orwell's nineteen eighty (01:17):
undefined
Speaker: four. (01:19):
undefined
Speaker: Now, in twenty twenty five, I (01:21):
undefined
Speaker: just read a disturbing story in (01:24):
undefined
Speaker: wired magazine. (01:26):
undefined
Speaker: It was about documents that show (01:28):
undefined
Speaker: how the US government plans to (01:30):
undefined
Speaker: hire dozens of contractors to (01:33):
undefined
Speaker: scan ex Facebook, TikTok and (01:36):
undefined
Speaker: other platforms to target people (01:40):
undefined
Speaker: for deportation. (01:43):
undefined
Speaker: This is not some simple think (01:45):
undefined
Speaker: tank memo or request for (01:47):
undefined
Speaker: information. (01:50):
undefined
Speaker: It has requirements and deadlines. (01:51):
undefined
Speaker: This is a detailed plan for (01:55):
undefined
Speaker: around the clock social media (01:58):
undefined
Speaker: monitoring. (02:01):
undefined
Speaker: And here's the part that caught my attention. (02:03):
undefined
Speaker: The system is built to sort, rank and predict. (02:07):
undefined
Speaker: It turns what we say in public (02:13):
undefined
Speaker: into scores that travel (02:15):
undefined
Speaker: categories we never agreed to, (02:18):
undefined
Speaker: but that still shape how we are (02:20):
undefined
Speaker: treated tonight. (02:24):
undefined
Speaker: Not in April of nineteen eighty (02:26):
undefined
Speaker: four, but in October of twenty (02:29):
undefined
Speaker: twenty five. (02:33):
undefined
Speaker: The stakes change again. (02:34):
undefined
Speaker: Immigration and Customs Enforcement is seeking vendors (02:37):
undefined
Speaker: for a social media surveillance program designed to turn public (02:41):
undefined
Speaker: posts and to operational leads. (02:46):
undefined
Speaker: That is the visible tip of something larger. (02:50):
undefined
Speaker: Every employer. (02:56):
undefined
Speaker: Lawyer. (02:57):
undefined
Speaker: Every insurer, every agency is learning to read your digital (02:58):
undefined
Speaker: shadow and they are getting really, really good at it. (03:05):
undefined
Speaker: Big brother is indeed watching you. (03:12):
undefined
Speaker: Part one. (03:18):
undefined
Speaker: The watchers have names and addresses. (03:20):
undefined
Speaker: Two buildings, one in Williston, (03:25):
undefined
Speaker: Vermont, population eight (03:28):
undefined
Speaker: thousand. (03:31):
undefined
Speaker: The kind of place where everyone knows your coffee order. (03:33):
undefined
Speaker: The other in Santa Ana, (03:37):
undefined
Speaker: California, in the shadow of (03:39):
undefined
Speaker: Disneyland, where dreams come (03:42):
undefined
Speaker: true. (03:44):
undefined
Speaker: Inside these buildings, nearly (03:45):
undefined
Speaker: thirty contractors will sit at (03:48):
undefined
Speaker: screens watching, not watching (03:51):
undefined
Speaker: television, not watching movies, (03:54):
undefined
Speaker: watching you. (03:57):
undefined
Speaker: The contract documents are public, filed October third. (04:00):
undefined
Speaker: They want artificial intelligence that can process ex (04:06):
undefined
Speaker: Facebook, Instagram, TikTok, YouTube, Reddit, every major (04:10):
undefined
Speaker: platform where humans gather to be, well, human. (04:16):
undefined
Speaker: Thirty minutes. (04:23):
undefined
Speaker: That's the turnaround time for (04:25):
undefined
Speaker: urgent cases from your post to (04:27):
undefined
Speaker: their report. (04:31):
undefined
Speaker: Half an hour to transform a tweet into intelligence. (04:33):
undefined
Speaker: They're calling it Osint, which (04:39):
undefined
Speaker: stands for Open Source (04:42):
undefined
Speaker: Intelligence. (04:44):
undefined
Speaker: It sounds technical, neutral, almost boring. (04:46):
undefined
Speaker: So let me tell you what these systems actually do. (04:52):
undefined
Speaker: Imagine you post a photo at your cousin's wedding. (04:57):
undefined
Speaker: Beautiful day. (05:01):
undefined
Speaker: Everyone's happy. (05:03):
undefined
Speaker: You're holding a beer. (05:05):
undefined
Speaker: Laughing at something off camera. (05:06):
undefined
Speaker: The location tag says El Paso. (05:09):
undefined
Speaker: The system sees the subject was mobile on October fifteenth. (05:12):
undefined
Speaker: Subject consumes alcohol. (05:17):
undefined
Speaker: Subject has family connections in border region. (05:20):
undefined
Speaker: Subject associates with it (05:24):
undefined
Speaker: starts checking your cousin's (05:27):
undefined
Speaker: profile. (05:28):
undefined
Speaker: Individual who posted anti enforcement content in March of (05:30):
undefined
Speaker: twenty twenty three. (05:34):
undefined
Speaker: Your moment of joy becomes data (05:36):
undefined
Speaker: points and someone else's (05:40):
undefined
Speaker: database. (05:42):
undefined
Speaker: The algorithm doesn't see the wedding. (05:43):
undefined
Speaker: It sees patterns and patterns, and the wrong light can look (05:46):
undefined
Speaker: like anything you want them to. (05:53):
undefined
Speaker: Part two. (05:57):
undefined
Speaker: The infrastructure was already here. (05:59):
undefined
Speaker: This isn't new. (06:04):
undefined
Speaker: Since twenty fourteen, Ice has operated something called Gost (06:06):
undefined
Speaker: or possibly ghost. (06:11):
undefined
Speaker: I'm not quite sure how to pronounce it, but it stands for (06:13):
undefined
Speaker: Giant Oak Search Technology. (06:17):
undefined
Speaker: Sounds like something from a spy novel, doesn't it? (06:21):
undefined
Speaker: Well, it's real, and it's been watching ghost crawls social (06:25):
undefined
Speaker: media looking for what the system calls derogatory (06:31):
undefined
Speaker: information, posts critical of America, photos at protest, (06:35):
undefined
Speaker: connections to people already flagged in other databases. (06:41):
undefined
Speaker: The interface is almost insultingly simple. (06:47):
undefined
Speaker: Thumbs up. (06:52):
undefined
Speaker: This person seems fine. (06:53):
undefined
Speaker: Thumbs down flag for review. (06:55):
undefined
Speaker: Like Tinder, but for human freedom. (07:00):
undefined
Speaker: Meanwhile, Ice maintains (07:03):
undefined
Speaker: contracts worth millions with (07:05):
undefined
Speaker: data brokers LexisNexis, Thomson (07:08):
undefined
Speaker: Reuters. (07:12):
undefined
Speaker: Companies that aggregate everything that's technically (07:13):
undefined
Speaker: public about you your address, history, your relatives. (07:16):
undefined
Speaker: That speeding ticket from twenty (07:22):
undefined
Speaker: nineteen, the house you looked (07:25):
undefined
Speaker: at but didn't buy, your voter (07:27):
undefined
Speaker: registration. (07:30):
undefined
Speaker: All of it compiled, cross-referenced and scored a (07:32):
undefined
Speaker: company called Palantir, which we did an episode on in episode (07:39):
undefined
Speaker: seven, named after the seeing Stones in Lord of the rings. (07:43):
undefined
Speaker: Just got thirty million dollars (07:47):
undefined
Speaker: to build something called (07:50):
undefined
Speaker: Emigration OS, an operating (07:52):
undefined
Speaker: system like windows, but for (07:55):
undefined
Speaker: tracking humans. (07:57):
undefined
Speaker: Peter Thiel, Palantir's founder, once said privacy was dead. (07:59):
undefined
Speaker: He was wrong. (08:06):
undefined
Speaker: Privacy isn't dead, per se. (08:07):
undefined
Speaker: It's just expensive. (08:11):
undefined
Speaker: The wealthy can afford lawyers, (08:14):
undefined
Speaker: PR firms, reputation management (08:17):
undefined
Speaker: services. (08:20):
undefined
Speaker: The rest of us, well, we are just raw data. (08:22):
undefined
Speaker: Part three. (08:29):
undefined
Speaker: The profile you'll never see. (08:31):
undefined
Speaker: Everyone knows employers check social media. (08:36):
undefined
Speaker: Not stop being news in twenty ten. (08:39):
undefined
Speaker: What's happening now is (08:43):
undefined
Speaker: algorithmic behavioral (08:44):
undefined
Speaker: profiling. (08:46):
undefined
Speaker: And you have no idea what story your patterns Hatton's tell (08:48):
undefined
Speaker: Resume Builders twenty twenty three survey confirms seventy (08:52):
undefined
Speaker: three percent of hiring managers screen social media. (08:56):
undefined
Speaker: But that statistic misses the evolution. (09:01):
undefined
Speaker: They're not reading your post anymore. (09:05):
undefined
Speaker: They're feeding them into systems that analyze patterns. (09:08):
undefined
Speaker: Think about what an algorithm might see. (09:13):
undefined
Speaker: Ten photos with drinks across two months. (09:16):
undefined
Speaker: You see normal social life. (09:21):
undefined
Speaker: The algorithm might flag potential substance dependency (09:24):
undefined
Speaker: risk, poor judgment patterns, health insurance liability. (09:27):
undefined
Speaker: Posting regularly at two a m. Well, you're a night owl. (09:33):
undefined
Speaker: The algorithm might interpret poor self-regulation likely to (09:38):
undefined
Speaker: miss morning meetings. (09:43):
undefined
Speaker: Potential productivity issues. (09:44):
undefined
Speaker: Do you delete posts frequently? (09:48):
undefined
Speaker: You're just editing yourself, right? (09:50):
undefined
Speaker: Well, the algorithm might code. (09:53):
undefined
Speaker: Impulsive decision making, (09:55):
undefined
Speaker: potential PR risk, emotional (09:56):
undefined
Speaker: instability indicators. (09:59):
undefined
Speaker: These systems are being sold right now. (10:02):
undefined
Speaker: HR tech companies advertise (10:06):
undefined
Speaker: behavioral prediction, cultural (10:09):
undefined
Speaker: fit analysis, risk assessment, (10:12):
undefined
Speaker: modeling. (10:15):
undefined
Speaker: They promise to predict who will (10:17):
undefined
Speaker: quit, who will cause problems, (10:19):
undefined
Speaker: who will cost more in health (10:22):
undefined
Speaker: insurance. (10:24):
undefined
Speaker: How do they calculate this? (10:26):
undefined
Speaker: We don't know. (10:28):
undefined
Speaker: These algorithms are proprietary trade secrets. (10:30):
undefined
Speaker: Black boxes. (10:34):
undefined
Speaker: But consider that's technically possible. (10:37):
undefined
Speaker: Sentiment analysis. (10:40):
undefined
Speaker: Tracking emotional volatility across post network analysis. (10:42):
undefined
Speaker: Measuring professional versus personal content ratios. (10:47):
undefined
Speaker: Temporal patterns. (10:51):
undefined
Speaker: Identifying focus and self-control indicators. (10:53):
undefined
Speaker: Language processing. (10:57):
undefined
Speaker: Detecting aggression. (10:58):
undefined
Speaker: Negativity. (11:00):
undefined
Speaker: Risk taking. (11:02):
undefined
Speaker: Terminology. (11:03):
undefined
Speaker: Every like. (11:05):
undefined
Speaker: Share. (11:07):
undefined
Speaker: Comment becomes a data point. (11:08):
undefined
Speaker: Your three am political rants, (11:11):
undefined
Speaker: your deleted post, your weakened (11:14):
undefined
Speaker: location tags at bars, your (11:18):
undefined
Speaker: complaint to positivity ratio, (11:20):
undefined
Speaker: the gaps in your posting that (11:23):
undefined
Speaker: might indicate depression or (11:25):
undefined
Speaker: instability. (11:28):
undefined
Speaker: You are not being judged by a (11:30):
undefined
Speaker: person who might understand (11:33):
undefined
Speaker: context. (11:35):
undefined
Speaker: You are being scored by software that sees patterns. (11:36):
undefined
Speaker: And here's the trap forty seven (11:42):
undefined
Speaker: percent of employers are (11:45):
undefined
Speaker: suspicious if you have no online (11:47):
undefined
Speaker: presence whatsoever. (11:50):
undefined
Speaker: But maintaining one means feeding the profiling system. (11:51):
undefined
Speaker: Maybe the algorithm is sophisticated enough to (11:56):
undefined
Speaker: accurately predict behavior from social media patterns. (11:59):
undefined
Speaker: Maybe it's digital astrology dressed up as a science. (12:04):
undefined
Speaker: You'll never know. (12:08):
undefined
Speaker: The company won't tell you why you weren't selected. (12:10):
undefined
Speaker: They might not even know themselves. (12:14):
undefined
Speaker: The algorithm generated a risk score. (12:17):
undefined
Speaker: That's all they needed. (12:20):
undefined
Speaker: Your digital shadow tells a story. (12:22):
undefined
Speaker: You just don't get to read it. (12:25):
undefined
Speaker: And you definitely do not get to edit it. (12:28):
undefined
Speaker: Part four the insurance detective and your phone. (12:35):
undefined
Speaker: A ski trip posted to Instagram. (12:43):
undefined
Speaker: Black diamond run. (12:47):
undefined
Speaker: Pure joy. (12:49):
undefined
Speaker: Six weeks later, a back injury from moving furniture. (12:50):
undefined
Speaker: It's legitimate. (12:54):
undefined
Speaker: The MRI confirmed it. (12:56):
undefined
Speaker: Your doctor prescribed treatment, but the insurance (12:57):
undefined
Speaker: denies the claim. (13:02):
undefined
Speaker: Inconsistent with reported physical limitations. (13:04):
undefined
Speaker: They don't mention the ski video. (13:08):
undefined
Speaker: They don't have to. (13:11):
undefined
Speaker: This is how modern insurance investigation works. (13:13):
undefined
Speaker: The ski video was from before the injury. (13:17):
undefined
Speaker: But the algorithm doesn't care about temporal context. (13:20):
undefined
Speaker: It sees athletic capability, (13:25):
undefined
Speaker: risk taking behavior, active (13:28):
undefined
Speaker: lifestyle. (13:30):
undefined
Speaker: It calculates lower payout probability. (13:31):
undefined
Speaker: Property casualty three hundred sixty. (13:36):
undefined
Speaker: An insurance industry (13:38):
undefined
Speaker: publication reported that in (13:40):
undefined
Speaker: twenty twenty five, a social (13:42):
undefined
Speaker: media evidence factors into (13:44):
undefined
Speaker: forty two percent of disputed (13:46):
undefined
Speaker: claims. (13:49):
undefined
Speaker: Investigations. (13:50):
undefined
Speaker: Almost half insurance companies (13:51):
undefined
Speaker: now contract with specialized (13:55):
undefined
Speaker: firms that scan social media for (13:57):
undefined
Speaker: claim verification. (14:00):
undefined
Speaker: They have names like Risk (14:02):
undefined
Speaker: Mitigation Services and Claim (14:04):
undefined
Speaker: Integrity Solutions. (14:08):
undefined
Speaker: What they're looking for depends on the claim for injury cases. (14:11):
undefined
Speaker: It's any physical movement exercise activity for disability (14:17):
undefined
Speaker: claims, it's evidence of unreported work or income. (14:22):
undefined
Speaker: Health insurance applications for lifestyle indicators that (14:27):
undefined
Speaker: suggest higher risk. (14:30):
undefined
Speaker: That marathon you just ran three (14:33):
undefined
Speaker: years ago and posted about to (14:35):
undefined
Speaker: the algorithm. (14:37):
undefined
Speaker: You're either low risk because you're healthy or high risk (14:39):
undefined
Speaker: because athletes get injuries. (14:45):
undefined
Speaker: It depends on what they're trying to prove. (14:48):
undefined
Speaker: The same data tells opposite (14:51):
undefined
Speaker: stories, depending on who's (14:53):
undefined
Speaker: interpreting it. (14:55):
undefined
Speaker: Think about wedding videos on TikTok. (14:56):
undefined
Speaker: Someone with chronic pain having one good day dancing for 30s at (15:00):
undefined
Speaker: their sister's wedding. (15:06):
undefined
Speaker: The investigators seized physical capacity inconsistent (15:08):
undefined
Speaker: with claim the algorithm flags potential fraud indicator. (15:12):
undefined
Speaker: The claim gets denied. (15:18):
undefined
Speaker: They don't see the three hours of pain before the dance. (15:21):
undefined
Speaker: The two days in bed after the (15:25):
undefined
Speaker: context that makes 30s of joy (15:28):
undefined
Speaker: possible and months of (15:30):
undefined
Speaker: suffering. (15:33):
undefined
Speaker: Industry training materials (15:34):
undefined
Speaker: publicly available from (15:36):
undefined
Speaker: investigation certification (15:39):
undefined
Speaker: programs. (15:40):
undefined
Speaker: Teach techniques like Cross, (15:41):
undefined
Speaker: referencing Geotags with claimed (15:44):
undefined
Speaker: limitations. (15:46):
undefined
Speaker: Analyzing Venmo transactions for undisclosed income. (15:48):
undefined
Speaker: Searching tagged photos where (15:52):
undefined
Speaker: the subject appears in other (15:54):
undefined
Speaker: posts and using facial (15:56):
undefined
Speaker: recognition to find unclaimed (15:58):
undefined
Speaker: social profiles. (16:01):
undefined
Speaker: You document your life thinking you're sharing with friends. (16:04):
undefined
Speaker: You're actually building an evidence file that can be read (16:08):
undefined
Speaker: against you by anyone willing to pay for it. (16:13):
undefined
Speaker: Part five the tools they use. (16:19):
undefined
Speaker: Can link cobwebs. (16:26):
undefined
Speaker: Shadow Dragon. (16:29):
undefined
Speaker: They sound like hacker handles from a nineties movie. (16:31):
undefined
Speaker: Well, they're actually companies (16:35):
undefined
Speaker: with government contracts and (16:38):
undefined
Speaker: corporate clients. (16:40):
undefined
Speaker: Shadow Dragon Social Net (16:42):
undefined
Speaker: monitors over two hundred (16:45):
undefined
Speaker: platforms, not just Facebook and (16:47):
undefined
Speaker: Twitter. (16:50):
undefined
Speaker: We're talking about niche forums, regional social (16:51):
undefined
Speaker: networks, places where you forgot you had accounts. (16:56):
undefined
Speaker: It maps relationships, finds alternate accounts, builds what (17:01):
undefined
Speaker: they call patterns of life. (17:07):
undefined
Speaker: Texas Department of Public (17:12):
undefined
Speaker: Safety gave a five million (17:14):
undefined
Speaker: dollars contract for a tool (17:16):
undefined
Speaker: called tangles. (17:19):
undefined
Speaker: It builds relationship webs from social media, finding (17:21):
undefined
Speaker: connections between people who've never met in person but (17:26):
undefined
Speaker: liked the same post. (17:31):
undefined
Speaker: Clearview AI just got a nine (17:34):
undefined
Speaker: million dollars ceiling contract (17:37):
undefined
Speaker: from Ice. (17:39):
undefined
Speaker: This is the company that scraped billions of photos from social (17:41):
undefined
Speaker: media without permission. (17:45):
undefined
Speaker: They built a facial recognition system so powerful it's actually (17:48):
undefined
Speaker: illegal in Europe. (17:54):
undefined
Speaker: You know that photo from your friends Instagram where you're (17:57):
undefined
Speaker: in the background? (18:00):
undefined
Speaker: Clearview has it and it knows it's you. (18:01):
undefined
Speaker: But the newest tool is even more invasive. (18:06):
undefined
Speaker: Location data not from your (18:10):
undefined
Speaker: post, but from your phone (18:13):
undefined
Speaker: itself. (18:16):
undefined
Speaker: A company called Babble Street had to cancel their contract (18:18):
undefined
Speaker: after journalists exposed they were selling location data. (18:23):
undefined
Speaker: So Ice just bought the same capability from Penlink instead. (18:28):
undefined
Speaker: Billions of location signals every day from hundreds of (18:35):
undefined
Speaker: millions of phones. (18:40):
undefined
Speaker: No warrant needed. (18:42):
undefined
Speaker: They can see where you've been, (18:45):
undefined
Speaker: who you've been near, how long (18:47):
undefined
Speaker: you stayed. (18:50):
undefined
Speaker: And most disturbingly, they can look back in time. (18:52):
undefined
Speaker: Part six when context dies. (18:58):
undefined
Speaker: December of twenty thirteen (19:05):
undefined
Speaker: Justine Sacco boards a plane to (19:08):
undefined
Speaker: South Africa. (19:11):
undefined
Speaker: She tweets going to Africa. (19:13):
undefined
Speaker: Hope I don't get Aids. (19:16):
undefined
Speaker: Haha just kidding, I'm white. (19:18):
undefined
Speaker: It was sarcasm. (19:21):
undefined
Speaker: A really bad joke about privilege and inequality at the (19:23):
undefined
Speaker: time of the tweet. (19:27):
undefined
Speaker: She only had one hundred and seventy followers. (19:29):
undefined
Speaker: She then turned off her phone for the eleven hour flight. (19:32):
undefined
Speaker: Fell asleep when she landed. (19:36):
undefined
Speaker: She was the number one trending topic worldwide. (19:39):
undefined
Speaker: Hashtag has just been landed yet. (19:45):
undefined
Speaker: People were waiting at the (19:48):
undefined
Speaker: airport to photograph her (19:50):
undefined
Speaker: reaction. (19:52):
undefined
Speaker: She had been fired while in the air. (19:54):
undefined
Speaker: Death threats filled her inbox. (19:58):
undefined
Speaker: Her life as she knew it was over. (20:00):
undefined
Speaker: Context. (20:05):
undefined
Speaker: Died at thirty thousand feet. (20:06):
undefined
Speaker: This is what happens when human (20:09):
undefined
Speaker: communication meets algorithmic (20:12):
undefined
Speaker: interpretation. (20:14):
undefined
Speaker: Sarcasm becomes statement. (20:16):
undefined
Speaker: Jokes become evidence. (20:20):
undefined
Speaker: Mistakes become permanent. (20:23):
undefined
Speaker: And twenty seventeen ten students had their admissions (20:27):
undefined
Speaker: revoked from Harvard University. (20:31):
undefined
Speaker: Someone leaked screenshots from a private Facebook group where (20:34):
undefined
Speaker: they shared offensive memes. (20:39):
undefined
Speaker: These kids and they were kids (20:42):
undefined
Speaker: eighteen years old thought they (20:45):
undefined
Speaker: were being edgy in a private (20:47):
undefined
Speaker: space. (20:50):
undefined
Speaker: Harvard disagreed. (20:51):
undefined
Speaker: Four years of perfect grades, extracurriculars, SAT prep gone (20:54):
undefined
Speaker: one hundred and forty characters can end a career. (21:01):
undefined
Speaker: A private group chat can derail a future. (21:06):
undefined
Speaker: A photo from five years ago can deny an insurance claim. (21:10):
undefined
Speaker: The algorithm doesn't understand context and strangely enough, (21:16):
undefined
Speaker: increasingly, neither do we. (21:22):
undefined
Speaker: Part seven the authentication of everything. (21:27):
undefined
Speaker: Every platform wants to verify you. (21:33):
undefined
Speaker: Now, blue checks on Twitter, ID verification on Facebook, real (21:36):
undefined
Speaker: names on LinkedIn. (21:42):
undefined
Speaker: They say it's about authenticity. (21:45):
undefined
Speaker: Fighting bots, creating trust. (21:47):
undefined
Speaker: Well it's not. (21:51):
undefined
Speaker: It's about making you survivable. (21:53):
undefined
Speaker: An anonymous account can speak truth to power. (21:57):
undefined
Speaker: A verified account has a home address. (22:01):
undefined
Speaker: China's social credit system (22:05):
undefined
Speaker: seemed dystopian when it (22:08):
undefined
Speaker: launched. (22:10):
undefined
Speaker: Bad social score. (22:11):
undefined
Speaker: Well, you might not be able to buy plane tickets, get a loan, (22:13):
undefined
Speaker: or you might not be able to send your kids to good schools. (22:18):
undefined
Speaker: And we here in the States laughed. (22:22):
undefined
Speaker: That could never happen here. (22:26):
undefined
Speaker: Well it's happening. (22:28):
undefined
Speaker: We just call it different names. (22:31):
undefined
Speaker: Background checks, credit scores. (22:34):
undefined
Speaker: Social media screening. (22:38):
undefined
Speaker: Algorithmic risk assessment. (22:41):
undefined
Speaker: Insurance evaluation. (22:44):
undefined
Speaker: The Chinese were just honest about centralizing it. (22:47):
undefined
Speaker: Pharma or pharma is a major (22:51):
undefined
Speaker: player in social media (22:55):
undefined
Speaker: screening. (22:57):
undefined
Speaker: They advertise that it searches (22:58):
undefined
Speaker: ten thousand online public (23:00):
undefined
Speaker: sources for what it calls (23:03):
undefined
Speaker: behavioral intelligence. (23:05):
undefined
Speaker: According to their own marketing materials. (23:08):
undefined
Speaker: They screen for nine types of (23:11):
undefined
Speaker: workplace misconduct, including (23:14):
undefined
Speaker: fraud, harassment, threats, and (23:17):
undefined
Speaker: violence. (23:20):
undefined
Speaker: They claim a ninety nine point nine five percent accuracy. (23:22):
undefined
Speaker: And they say their AI can (23:27):
undefined
Speaker: identify extremist symbols, (23:29):
undefined
Speaker: violent imagery, memes and (23:32):
undefined
Speaker: gestures. (23:36):
undefined
Speaker: They use what they call avatar (23:38):
undefined
Speaker: recognition to find your (23:41):
undefined
Speaker: accounts, even when you use (23:43):
undefined
Speaker: different names. (23:46):
undefined
Speaker: This is actually from their (23:48):
undefined
Speaker: marketing materials, not (23:50):
undefined
Speaker: speculation. (23:52):
undefined
Speaker: Again, this is from their pitch decks. (23:53):
undefined
Speaker: Apparently another screening (23:57):
undefined
Speaker: company analyzes what they call (23:59):
undefined
Speaker: eleven different behaviors using (24:02):
undefined
Speaker: a billion profile Osint (24:06):
undefined
Speaker: database. (24:08):
undefined
Speaker: They advertise the ability to (24:10):
undefined
Speaker: spot bias threats, political, (24:12):
undefined
Speaker: disparaging and prejudiced (24:16):
undefined
Speaker: speech across text and visual (24:19):
undefined
Speaker: content. (24:21):
undefined
Speaker: Aquasource HR combines AI driven search tools with what they call (24:23):
undefined
Speaker: expert human analysis from trained social anthropologist. (24:31):
undefined
Speaker: Social anthropologist studying (24:37):
undefined
Speaker: your tweets like you're an (24:40):
undefined
Speaker: undiscovered tribe. (24:42):
undefined
Speaker: And here is the number that should terrify everyone. (24:45):
undefined
Speaker: A sterling survey found that sixty eight percent of employers (24:50):
undefined
Speaker: admitted to using social media to find answers to illegal (24:55):
undefined
Speaker: interview questions, questions they can't legally ask you. (25:01):
undefined
Speaker: Age. (25:08):
undefined
Speaker: Religion. (25:09):
undefined
Speaker: Sexual orientation. (25:10):
undefined
Speaker: Health conditions. (25:12):
undefined
Speaker: Sixty eight percent admitted it. (25:14):
undefined
Speaker: These companies defend themselves with the same line. (25:18):
undefined
Speaker: It's all public data. (25:22):
undefined
Speaker: What's public is your entire life, and they've built an (25:26):
undefined
Speaker: industry around reading it. (25:31):
undefined
Speaker: Part eight the deletion myth. (25:36):
undefined
Speaker: You can't delete anything. (25:41):
undefined
Speaker: Not really. (25:44):
undefined
Speaker: Sure, you can remove posts from (25:47):
undefined
Speaker: your timeline, but the data (25:49):
undefined
Speaker: persist. (25:51):
undefined
Speaker: And backups. (25:52):
undefined
Speaker: Caches. (25:53):
undefined
Speaker: Archives. (25:55):
undefined
Speaker: Screenshots. (25:56):
undefined
Speaker: Proliferate. (25:57):
undefined
Speaker: Aggregators. (25:59):
undefined
Speaker: Scrape and store. (25:59):
undefined
Speaker: The European Union has the right to be forgotten. (26:02):
undefined
Speaker: Between twenty fourteen and twenty seventeen, two point four (26:07):
undefined
Speaker: million people requested Google remove their information. (26:12):
undefined
Speaker: Google approved forty three percent, less than half. (26:18):
undefined
Speaker: In America, we don't even have that. (26:23):
undefined
Speaker: No federal right to deletion. (26:26):
undefined
Speaker: No right to see what's in your life. (26:28):
undefined
Speaker: No right to correct errors when you delete a post. (26:31):
undefined
Speaker: Here's what actually happens. (26:36):
undefined
Speaker: The platform removes it from public view. (26:38):
undefined
Speaker: That's all. (26:43):
undefined
Speaker: Server logs remain. (26:44):
undefined
Speaker: Backup systems retain copies. (26:46):
undefined
Speaker: Law enforcement can still access through legal request. (26:50):
undefined
Speaker: Meanwhile, the Internet Archive captures pages. (26:54):
undefined
Speaker: Google caches search results. (26:58):
undefined
Speaker: Third party monitoring services, (27:01):
undefined
Speaker: the ones selling to HR (27:04):
undefined
Speaker: departments and insurance (27:05):
undefined
Speaker: companies have already scraped (27:06):
undefined
Speaker: and stored other user's (27:09):
undefined
Speaker: screenshot before you think to (27:12):
undefined
Speaker: delete. (27:13):
undefined
Speaker: Data brokers operate on a simple (27:15):
undefined
Speaker: principle accumulation without (27:17):
undefined
Speaker: expiration. (27:20):
undefined
Speaker: Companies like LexisNexis and (27:22):
undefined
Speaker: Thomson Reuters aggregate public (27:24):
undefined
Speaker: records social media purchase (27:27):
undefined
Speaker: histories. (27:30):
undefined
Speaker: They create profiles on hundreds of millions of Americans. (27:31):
undefined
Speaker: Profiles you can't see. (27:37):
undefined
Speaker: You can't correct. (27:39):
undefined
Speaker: You can't delete. (27:41):
undefined
Speaker: They sell access to these (27:44):
undefined
Speaker: profiles tens of thousands of (27:45):
undefined
Speaker: times per year to employers, (27:48):
undefined
Speaker: insurers, landlords, anyone with (27:52):
undefined
Speaker: a business account and a credit (27:56):
undefined
Speaker: card. (27:58):
undefined
Speaker: The data multiplies. (27:59):
undefined
Speaker: Every search someone runs on you, generates metadata about (28:01):
undefined
Speaker: who's interested and new. (28:06):
undefined
Speaker: Every analysis creates derived data. (28:09):
undefined
Speaker: Risk scores, behavioral predictions, network maps. (28:13):
undefined
Speaker: You exist in databases you've never heard of, scored by (28:18):
undefined
Speaker: algorithms you'll never see sold to buyers you'll never know. (28:24):
undefined
Speaker: Deletion is theatre. (28:30):
undefined
Speaker: The data persist, and in America you have no right to even know (28:33):
undefined
Speaker: what story it tells. (28:40):
undefined
Speaker: Part nine The Great Quieting. (28:44):
undefined
Speaker: Something is happening to public discourse. (28:51):
undefined
Speaker: People are self-censoring. (28:54):
undefined
Speaker: Not because the government told them to. (28:57):
undefined
Speaker: Because they know everything is watched. (29:00):
undefined
Speaker: Recorded. (29:03):
undefined
Speaker: Analyzed. (29:04):
undefined
Speaker: Scored. (29:06):
undefined
Speaker: A Pew Research study found that seventy two percent of Americans (29:08):
undefined
Speaker: believe their online activities are being tracked by companies. (29:12):
undefined
Speaker: And they are right. (29:18):
undefined
Speaker: But it's what they do with that knowledge that matters. (29:20):
undefined
Speaker: They become careful, calculated, boring. (29:25):
undefined
Speaker: The phrases repeat across social media. (29:30):
undefined
Speaker: I don't post about politics anymore. (29:33):
undefined
Speaker: I deleted all my party photos linked and voice only made a (29:37):
undefined
Speaker: separate account for work. (29:45):
undefined
Speaker: The same story. (29:48):
undefined
Speaker: Different voices. (29:49):
undefined
Speaker: Self-censorship as survival strategy. (29:51):
undefined
Speaker: The chilling effect is real and measurable. (29:56):
undefined
Speaker: When Cambridge Analytica's data harvesting was exposed, Facebook (30:00):
undefined
Speaker: lost two point eight million US users under twenty five. (30:06):
undefined
Speaker: Not just inactive. (30:11):
undefined
Speaker: Gone. (30:13):
undefined
undefined
Speaker: When people learn how their data is used, they don't just change (30:14):
undefined
Speaker: passwords, they change behavior. (30:19):
undefined
Speaker: We are watching democracy's immune system shut down. (30:22):
undefined
Speaker: Descent requires the possibility (30:29):
undefined
Speaker: of having the freedom to be (30:32):
undefined
Speaker: anonymous. (30:34):
undefined
Speaker: Innovation requires the freedom to be wrong. (30:35):
undefined
Speaker: Growth requires the space to be foolish. (30:39):
undefined
Speaker: When everything is permanent and (30:44):
undefined
Speaker: searchable, nothing important (30:47):
undefined
Speaker: gets said. (30:50):
undefined
Speaker: The result is a generation learning to perform stability (30:51):
undefined
Speaker: rather than experience growth. (30:56):
undefined
Speaker: The cure rate for an algorithmic audience that never forgets and (30:59):
undefined
Speaker: never forgives everyone. (31:03):
undefined
Speaker: Self edits now. (31:06):
undefined
Speaker: Not for friends. (31:08):
undefined
Speaker: For the machine that's watching. (31:10):
undefined
Speaker: And the machine is always watching. (31:13):
undefined
Speaker: Part ten the resistance that isn't. (31:20):
undefined
Speaker: People think they are fighting back. (31:26):
undefined
Speaker: Private accounts. (31:29):
undefined
Speaker: Fake names. (31:31):
undefined
Speaker: Deleted apps. (31:34):
undefined
Speaker: Digital detoxes. (31:36):
undefined
Speaker: It doesn't work. (31:39):
undefined
Speaker: Your phone's advertising ID (31:40):
undefined
Speaker: connects everything your credit (31:43):
undefined
Speaker: card links, your purchases, your (31:45):
undefined
Speaker: location. (31:48):
undefined
Speaker: Data tells the real story. (31:49):
undefined
Speaker: Your contacts upload their address book with your number (31:52):
undefined
Speaker: and them you exist and other people's data shadows. (31:57):
undefined
Speaker: Even if you never, ever created accounts. (32:03):
undefined
Speaker: Ghost profiles. (32:08):
undefined
Speaker: That's what Facebook calls them internally. (32:09):
undefined
Speaker: According to leaked documents, they are profiles of people who (32:13):
undefined
Speaker: never signed up. (32:19):
undefined
Speaker: Built from photos of other posts. (32:21):
undefined
Speaker: Contact list. (32:24):
undefined
Speaker: Others. (32:25):
undefined
Speaker: Upload and relationships others document. (32:26):
undefined
Speaker: Now you of course can opt out of Facebook, but you can't opt out (32:30):
undefined
Speaker: of being in Facebook. (32:36):
undefined
Speaker: The real resistance is political. (32:39):
undefined
Speaker: The only real resistance is political. (32:42):
undefined
Speaker: Force with teeth. (32:47):
undefined
Speaker: Regulations that bite. (32:49):
undefined
Speaker: Rights that can't be waived in terms of service. (32:51):
undefined
Speaker: Europe has GDPR. (32:57):
undefined
Speaker: California has CcpA. (33:00):
undefined
Speaker: They're imperfect, but they are (33:03):
undefined
Speaker: something the rest of America (33:06):
undefined
Speaker: has. (33:09):
undefined
Speaker: Thoughts and prayers. (33:10):
undefined
Speaker: Federal privacy legislation has been proposed repeatedly. (33:12):
undefined
Speaker: It never passes. (33:17):
undefined
Speaker: The tech lobby warns it would break the internet. (33:19):
undefined
Speaker: The security lobby says it would help terrorists. (33:22):
undefined
Speaker: First, the business lobby claims economic catastrophe. (33:25):
undefined
Speaker: We can't even agree on basic transparency, whether people (33:31):
undefined
Speaker: should know what data companies collect, whether they should see (33:36):
undefined
Speaker: their own files. (33:41):
undefined
Speaker: The answer is always the same (33:43):
undefined
Speaker: too complicated, too expensive, (33:46):
undefined
Speaker: too dangerous. (33:49):
undefined
Speaker: Translation transparency would reveal how bad it really is. (33:52):
undefined
Speaker: Part eleven the score you can't see. (34:01):
undefined
Speaker: You have a score. (34:08):
undefined
Speaker: Multiple scores. (34:11):
undefined
Speaker: Actually. (34:12):
undefined
Speaker: Credit score. (34:13):
undefined
Speaker: You know that one already. (34:14):
undefined
Speaker: But also customer lifetime value score. (34:16):
undefined
Speaker: Insurance risk score. (34:20):
undefined
Speaker: Employment Probability score. (34:23):
undefined
Speaker: Social influence score. (34:26):
undefined
Speaker: Fraud. (34:29):
undefined
Speaker: Likelihood score. (34:30):
undefined
Speaker: These are not conspiracy theories. (34:33):
undefined
Speaker: These are products sold at industry conferences. (34:36):
undefined
Speaker: Data analytics companies openly (34:41):
undefined
Speaker: advertise predictive scoring (34:44):
undefined
Speaker: systems. (34:46):
undefined
Speaker: Know your customer's true value. (34:48):
undefined
Speaker: Predictive employee churn before it happens. (34:51):
undefined
Speaker: Identify risk before they manifest. (34:55):
undefined
Speaker: HR technology firms claim their (34:59):
undefined
Speaker: algorithms can predict with high (35:02):
undefined
Speaker: accuracy which employees will (35:05):
undefined
Speaker: quit, which will get sick, which (35:08):
undefined
Speaker: will become what they call (35:13):
undefined
Speaker: problematic. (35:15):
undefined
Speaker: How do they do this? (35:18):
undefined
Speaker: Well, they analyzed digital footprints, email patterns, (35:20):
undefined
Speaker: social media activity, even physical workplace data. (35:24):
undefined
Speaker: When available. (35:30):
undefined
Speaker: Badge swipes, parking times, (35:31):
undefined
Speaker: cafeteria purchases the privacy (35:34):
undefined
Speaker: policies are buried in (35:37):
undefined
Speaker: employment agreements. (35:39):
undefined
Speaker: Page forty seven, section three. (35:41):
undefined
Speaker: The part no one reads. (35:44):
undefined
Speaker: You consented without knowing what you consented to. (35:48):
undefined
Speaker: Part twelve tomorrow's crime. (35:55):
undefined
Speaker: Today's punishment. (35:59):
undefined
Speaker: Predictive policing was supposed (36:03):
undefined
Speaker: to stop crime before it (36:05):
undefined
Speaker: happened. (36:07):
undefined
Speaker: Instead, it criminalizes probability. (36:08):
undefined
Speaker: Chicago's heat list algorithm, (36:12):
undefined
Speaker: officially called the Strategic (36:16):
undefined
Speaker: Subject List, identified people (36:19):
undefined
Speaker: likely to be involved in (36:22):
undefined
Speaker: shootings, either as victims or (36:25):
undefined
Speaker: perpetrators. (36:28):
undefined
Speaker: The algorithm did not (36:30):
undefined
Speaker: distinguish if you made the (36:32):
undefined
Speaker: list. (36:34):
undefined
Speaker: Police visited your home, (36:35):
undefined
Speaker: knocked on your door, told you (36:38):
undefined
Speaker: they were watching the (36:41):
undefined
Speaker: algorithm, considered arrest, (36:44):
undefined
Speaker: not convictions. (36:47):
undefined
Speaker: Social networks. (36:49):
undefined
Speaker: Geography. (36:51):
undefined
Speaker: Age. (36:52):
undefined
Speaker: Being young, black, and living in certain neighborhoods was (36:54):
undefined
Speaker: enough to score high. (36:58):
undefined
Speaker: The program was discontinued after a Rand Corporation study (37:01):
undefined
Speaker: found it ineffective, and civil rights groups proved it was (37:05):
undefined
Speaker: algorithmic racial profiling. (37:10):
undefined
Speaker: But the concept didn't die. (37:13):
undefined
Speaker: It evolved. (37:16):
undefined
Speaker: Now it's person based predictive analytics. (37:18):
undefined
Speaker: It's the same idea, but better branding. (37:23):
undefined
Speaker: Police departments nationwide use variations. (37:27):
undefined
Speaker: Ice also uses it as well as border patrol. (37:31):
undefined
Speaker: Your social media feeds these (37:35):
undefined
Speaker: systems that protest you (37:38):
undefined
Speaker: attended, that article you (37:41):
undefined
Speaker: shared. (37:43):
undefined
Speaker: That friend who got arrested all became inputs to algorithms (37:44):
undefined
Speaker: calculating threat probability. (37:50):
undefined
Speaker: The companies building these systems claim accuracy rates (37:54):
undefined
Speaker: above eighty five percent. (37:58):
undefined
Speaker: They say the algorithms identify patterns humans miss. (38:01):
undefined
Speaker: What patterns? (38:07):
undefined
Speaker: Well, they won't say trade secrets. (38:08):
undefined
Speaker: Proprietary methods. (38:12):
undefined
Speaker: The algorithm generates a score. (38:14):
undefined
Speaker: The score generates an action, and no one knows exactly why. (38:18):
undefined
Speaker: Not even the people using it. (38:25):
undefined
Speaker: Part thirteen the Future that's already here. (38:30):
undefined
Speaker: William Gibson said. (38:37):
undefined
Speaker: The future is already here. (38:39):
undefined
Speaker: It's just unevenly distributed. (38:42):
undefined
Speaker: He was right, but he was also wrong. (38:45):
undefined
Speaker: The surveillance future isn't unevenly distributed. (38:49):
undefined
Speaker: It's universal. (38:55):
undefined
Speaker: We all live in it. (38:56):
undefined
Speaker: We just experience it differently based on our scores. (38:58):
undefined
Speaker: If your scores are good, (39:03):
undefined
Speaker: employed, insured, documented, (39:05):
undefined
Speaker: compliant, you might never (39:08):
undefined
Speaker: notice the cage doors are open (39:10):
undefined
Speaker: for you. (39:13):
undefined
Speaker: Services appear. (39:14):
undefined
Speaker: Life just feels frictionless. (39:15):
undefined
Speaker: If your scores are bad, every interaction is friction. (39:19):
undefined
Speaker: Every application is denied. (39:24):
undefined
Speaker: Every benefit requires proof. (39:27):
undefined
Speaker: Every movement is questioned. (39:30):
undefined
Speaker: The architecture of surveillance is complete. (39:34):
undefined
Speaker: The only question is how it will be used in the future. (39:38):
undefined
Speaker: ICE's social media monitoring program isn't the beginning. (39:43):
undefined
Speaker: It's the formalization of what's already happening. (39:48):
undefined
Speaker: The government is just catching up to what corporations have (39:52):
undefined
Speaker: been doing for a decade. (39:57):
undefined
Speaker: Your posts are being watched, (39:59):
undefined
Speaker: your patterns analyzed, your (40:02):
undefined
Speaker: future predicted, your worth (40:05):
undefined
Speaker: calculated. (40:08):
undefined
Speaker: Right now, as you listen to this podcast, an algorithm somewhere (40:10):
undefined
Speaker: is updating your file. (40:17):
undefined
Speaker: You can't see it. (40:20):
undefined
Speaker: You can't correct it. (40:21):
undefined
Speaker: You can't escape it. (40:23):
undefined
Speaker: But you should know it exists. (40:25):
undefined
Speaker: Because in a world where everything is remembered and (40:28):
undefined
Speaker: nothing is forgiven, the most radical act isn't revolution. (40:32):
undefined
Speaker: It's remembering that you are (40:39):
undefined
Speaker: more than the sum of your data (40:42):
undefined
Speaker: points. (40:44):
undefined
Speaker: You are a human being. (40:45):
undefined
Speaker: Complex. (40:48):
undefined
Speaker: Contradictory. (40:49):
undefined
Speaker: Capable of change. (40:51):
undefined
Speaker: The algorithm doesn't know that. (40:54):
undefined
Speaker: But you do. (40:57):
undefined
Speaker: And maybe, just maybe, that's enough to start with. (40:59):
undefined
undefined
Speaker: back the layers of technology, economics, data science, and (00:09):
undefined
Speaker: even the human condition itself, all in an effort to try and (00:15):
undefined
Speaker: understand this vast and often perplexing world we live in. (00:20):
undefined
Speaker: I'm your host, Erika Barker, and tonight we are talking about how (00:26):
undefined
Speaker: you and everyone you know are being watched, not from a dark (00:32):
undefined
Speaker: van down the street with sophisticated surveillance (00:38):
undefined
Speaker: equipment, but by algorithms. (00:42):
undefined
Speaker: We're examining how the everyday (00:46):
undefined
Speaker: act of posting, liking, and (00:48):
undefined
Speaker: scrolling turns into something (00:51):
undefined
Speaker: else once machines start keeping (00:53):
undefined
Speaker: score. (00:57):
undefined
Speaker: It was a bright cold day in (01:06):
undefined
Speaker: April and the clocks were (01:09):
undefined
Speaker: striking thirteen. (01:12):
undefined
Speaker: That's the famous opening line (01:15):
undefined
Speaker: to Orwell's nineteen eighty (01:17):
undefined
Speaker: four. (01:19):
undefined
Speaker: Now, in twenty twenty five, I (01:21):
undefined
Speaker: just read a disturbing story in (01:24):
undefined
Speaker: wired magazine. (01:26):
undefined
Speaker: It was about documents that show (01:28):
undefined
Speaker: how the US government plans to (01:30):
undefined
Speaker: hire dozens of contractors to (01:33):
undefined
Speaker: scan ex Facebook, TikTok and (01:36):
undefined
Speaker: other platforms to target people (01:40):
undefined
Speaker: for deportation. (01:43):
undefined
Speaker: This is not some simple think (01:45):
undefined
Speaker: tank memo or request for (01:47):
undefined
Speaker: information. (01:50):
undefined
Speaker: It has requirements and deadlines. (01:51):
undefined
Speaker: This is a detailed plan for (01:55):
undefined
Speaker: around the clock social media (01:58):
undefined
Speaker: monitoring. (02:01):
undefined
Speaker: And here's the part that caught my attention. (02:03):
undefined
Speaker: The system is built to sort, rank and predict. (02:07):
undefined
Speaker: It turns what we say in public (02:13):
undefined
Speaker: into scores that travel (02:15):
undefined
Speaker: categories we never agreed to, (02:18):
undefined
Speaker: but that still shape how we are (02:20):
undefined
Speaker: treated tonight. (02:24):
undefined
Speaker: Not in April of nineteen eighty (02:26):
undefined
Speaker: four, but in October of twenty (02:29):
undefined
Speaker: twenty five. (02:33):
undefined
Speaker: The stakes change again. (02:34):
undefined
Speaker: Immigration and Customs Enforcement is seeking vendors (02:37):
undefined
Speaker: for a social media surveillance program designed to turn public (02:41):
undefined
Speaker: posts and to operational leads. (02:46):
undefined
Speaker: That is the visible tip of something larger. (02:50):
undefined
Speaker: Every employer. (02:56):
undefined
Speaker: Lawyer. (02:57):
undefined
Speaker: Every insurer, every agency is learning to read your digital (02:58):
undefined
Speaker: shadow and they are getting really, really good at it. (03:05):
undefined
Speaker: Big brother is indeed watching you. (03:12):
undefined
Speaker: Part one. (03:18):
undefined
Speaker: The watchers have names and addresses. (03:20):
undefined
Speaker: Two buildings, one in Williston, (03:25):
undefined
Speaker: Vermont, population eight (03:28):
undefined
Speaker: thousand. (03:31):
undefined
Speaker: The kind of place where everyone knows your coffee order. (03:33):
undefined
Speaker: The other in Santa Ana, (03:37):
undefined
Speaker: California, in the shadow of (03:39):
undefined
Speaker: Disneyland, where dreams come (03:42):
undefined
Speaker: true. (03:44):
undefined
Speaker: Inside these buildings, nearly (03:45):
undefined
Speaker: thirty contractors will sit at (03:48):
undefined
Speaker: screens watching, not watching (03:51):
undefined
Speaker: television, not watching movies, (03:54):
undefined
Speaker: watching you. (03:57):
undefined
Speaker: The contract documents are public, filed October third. (04:00):
undefined
Speaker: They want artificial intelligence that can process ex (04:06):
undefined
Speaker: Facebook, Instagram, TikTok, YouTube, Reddit, every major (04:10):
undefined
Speaker: platform where humans gather to be, well, human. (04:16):
undefined
Speaker: Thirty minutes. (04:23):
undefined
Speaker: That's the turnaround time for (04:25):
undefined
Speaker: urgent cases from your post to (04:27):
undefined
Speaker: their report. (04:31):
undefined
Speaker: Half an hour to transform a tweet into intelligence. (04:33):
undefined
Speaker: They're calling it Osint, which (04:39):
undefined
Speaker: stands for Open Source (04:42):
undefined
Speaker: Intelligence. (04:44):
undefined
Speaker: It sounds technical, neutral, almost boring. (04:46):
undefined
Speaker: So let me tell you what these systems actually do. (04:52):
undefined
Speaker: Imagine you post a photo at your cousin's wedding. (04:57):
undefined
Speaker: Beautiful day. (05:01):
undefined
Speaker: Everyone's happy. (05:03):
undefined
Speaker: You're holding a beer. (05:05):
undefined
Speaker: Laughing at something off camera. (05:06):
undefined
Speaker: The location tag says El Paso. (05:09):
undefined
Speaker: The system sees the subject was mobile on October fifteenth. (05:12):
undefined
Speaker: Subject consumes alcohol. (05:17):
undefined
Speaker: Subject has family connections in border region. (05:20):
undefined
Speaker: Subject associates with it (05:24):
undefined
Speaker: starts checking your cousin's (05:27):
undefined
Speaker: profile. (05:28):
undefined
Speaker: Individual who posted anti enforcement content in March of (05:30):
undefined
Speaker: twenty twenty three. (05:34):
undefined
Speaker: Your moment of joy becomes data (05:36):
undefined
Speaker: points and someone else's (05:40):
undefined
Speaker: database. (05:42):
undefined
Speaker: The algorithm doesn't see the wedding. (05:43):
undefined
Speaker: It sees patterns and patterns, and the wrong light can look (05:46):
undefined
Speaker: like anything you want them to. (05:53):
undefined
Speaker: Part two. (05:57):
undefined
Speaker: The infrastructure was already here. (05:59):
undefined
Speaker: This isn't new. (06:04):
undefined
Speaker: Since twenty fourteen, Ice has operated something called Gost (06:06):
undefined
Speaker: or possibly ghost. (06:11):
undefined
Speaker: I'm not quite sure how to pronounce it, but it stands for (06:13):
undefined
Speaker: Giant Oak Search Technology. (06:17):
undefined
Speaker: Sounds like something from a spy novel, doesn't it? (06:21):
undefined
Speaker: Well, it's real, and it's been watching ghost crawls social (06:25):
undefined
Speaker: media looking for what the system calls derogatory (06:31):
undefined
Speaker: information, posts critical of America, photos at protest, (06:35):
undefined
Speaker: connections to people already flagged in other databases. (06:41):
undefined
Speaker: The interface is almost insultingly simple. (06:47):
undefined
Speaker: Thumbs up. (06:52):
undefined
Speaker: This person seems fine. (06:53):
undefined
Speaker: Thumbs down flag for review. (06:55):
undefined
Speaker: Like Tinder, but for human freedom. (07:00):
undefined
Speaker: Meanwhile, Ice maintains (07:03):
undefined
Speaker: contracts worth millions with (07:05):
undefined
Speaker: data brokers LexisNexis, Thomson (07:08):
undefined
Speaker: Reuters. (07:12):
undefined
Speaker: Companies that aggregate everything that's technically (07:13):
undefined
Speaker: public about you your address, history, your relatives. (07:16):
undefined
Speaker: That speeding ticket from twenty (07:22):
undefined
Speaker: nineteen, the house you looked (07:25):
undefined
Speaker: at but didn't buy, your voter (07:27):
undefined
Speaker: registration. (07:30):
undefined
Speaker: All of it compiled, cross-referenced and scored a (07:32):
undefined
Speaker: company called Palantir, which we did an episode on in episode (07:39):
undefined
Speaker: seven, named after the seeing Stones in Lord of the rings. (07:43):
undefined
Speaker: Just got thirty million dollars (07:47):
undefined
Speaker: to build something called (07:50):
undefined
Speaker: Emigration OS, an operating (07:52):
undefined
Speaker: system like windows, but for (07:55):
undefined
Speaker: tracking humans. (07:57):
undefined
Speaker: Peter Thiel, Palantir's founder, once said privacy was dead. (07:59):
undefined
Speaker: He was wrong. (08:06):
undefined
Speaker: Privacy isn't dead, per se. (08:07):
undefined
Speaker: It's just expensive. (08:11):
undefined
Speaker: The wealthy can afford lawyers, (08:14):
undefined
Speaker: PR firms, reputation management (08:17):
undefined
Speaker: services. (08:20):
undefined
Speaker: The rest of us, well, we are just raw data. (08:22):
undefined
Speaker: Part three. (08:29):
undefined
Speaker: The profile you'll never see. (08:31):
undefined
Speaker: Everyone knows employers check social media. (08:36):
undefined
Speaker: Not stop being news in twenty ten. (08:39):
undefined
Speaker: What's happening now is (08:43):
undefined
Speaker: algorithmic behavioral (08:44):
undefined
Speaker: profiling. (08:46):
undefined
Speaker: And you have no idea what story your patterns Hatton's tell (08:48):
undefined
Speaker: Resume Builders twenty twenty three survey confirms seventy (08:52):
undefined
Speaker: three percent of hiring managers screen social media. (08:56):
undefined
Speaker: But that statistic misses the evolution. (09:01):
undefined
Speaker: They're not reading your post anymore. (09:05):
undefined
Speaker: They're feeding them into systems that analyze patterns. (09:08):
undefined
Speaker: Think about what an algorithm might see. (09:13):
undefined
Speaker: Ten photos with drinks across two months. (09:16):
undefined
Speaker: You see normal social life. (09:21):
undefined
Speaker: The algorithm might flag potential substance dependency (09:24):
undefined
Speaker: risk, poor judgment patterns, health insurance liability. (09:27):
undefined
Speaker: Posting regularly at two a m. Well, you're a night owl. (09:33):
undefined
Speaker: The algorithm might interpret poor self-regulation likely to (09:38):
undefined
Speaker: miss morning meetings. (09:43):
undefined
Speaker: Potential productivity issues. (09:44):
undefined
Speaker: Do you delete posts frequently? (09:48):
undefined
Speaker: You're just editing yourself, right? (09:50):
undefined
Speaker: Well, the algorithm might code. (09:53):
undefined
Speaker: Impulsive decision making, (09:55):
undefined
Speaker: potential PR risk, emotional (09:56):
undefined
Speaker: instability indicators. (09:59):
undefined
Speaker: These systems are being sold right now. (10:02):
undefined
Speaker: HR tech companies advertise (10:06):
undefined
Speaker: behavioral prediction, cultural (10:09):
undefined
Speaker: fit analysis, risk assessment, (10:12):
undefined
Speaker: modeling. (10:15):
undefined
Speaker: They promise to predict who will (10:17):
undefined
Speaker: quit, who will cause problems, (10:19):
undefined
Speaker: who will cost more in health (10:22):
undefined
Speaker: insurance. (10:24):
undefined
Speaker: How do they calculate this? (10:26):
undefined
Speaker: We don't know. (10:28):
undefined
Speaker: These algorithms are proprietary trade secrets. (10:30):
undefined
Speaker: Black boxes. (10:34):
undefined
Speaker: But consider that's technically possible. (10:37):
undefined
Speaker: Sentiment analysis. (10:40):
undefined
Speaker: Tracking emotional volatility across post network analysis. (10:42):
undefined
Speaker: Measuring professional versus personal content ratios. (10:47):
undefined
Speaker: Temporal patterns. (10:51):
undefined
Speaker: Identifying focus and self-control indicators. (10:53):
undefined
Speaker: Language processing. (10:57):
undefined
Speaker: Detecting aggression. (10:58):
undefined
Speaker: Negativity. (11:00):
undefined
Speaker: Risk taking. (11:02):
undefined
Speaker: Terminology. (11:03):
undefined
Speaker: Every like. (11:05):
undefined
Speaker: Share. (11:07):
undefined
Speaker: Comment becomes a data point. (11:08):
undefined
Speaker: Your three am political rants, (11:11):
undefined
Speaker: your deleted post, your weakened (11:14):
undefined
Speaker: location tags at bars, your (11:18):
undefined
Speaker: complaint to positivity ratio, (11:20):
undefined
Speaker: the gaps in your posting that (11:23):
undefined
Speaker: might indicate depression or (11:25):
undefined
Speaker: instability. (11:28):
undefined
Speaker: You are not being judged by a (11:30):
undefined
Speaker: person who might understand (11:33):
undefined
Speaker: context. (11:35):
undefined
Speaker: You are being scored by software that sees patterns. (11:36):
undefined
Speaker: And here's the trap forty seven (11:42):
undefined
Speaker: percent of employers are (11:45):
undefined
Speaker: suspicious if you have no online (11:47):
undefined
Speaker: presence whatsoever. (11:50):
undefined
Speaker: But maintaining one means feeding the profiling system. (11:51):
undefined
Speaker: Maybe the algorithm is sophisticated enough to (11:56):
undefined
Speaker: accurately predict behavior from social media patterns. (11:59):
undefined
Speaker: Maybe it's digital astrology dressed up as a science. (12:04):
undefined
Speaker: You'll never know. (12:08):
undefined
Speaker: The company won't tell you why you weren't selected. (12:10):
undefined
Speaker: They might not even know themselves. (12:14):
undefined
Speaker: The algorithm generated a risk score. (12:17):
undefined
Speaker: That's all they needed. (12:20):
undefined
Speaker: Your digital shadow tells a story. (12:22):
undefined
Speaker: You just don't get to read it. (12:25):
undefined
Speaker: And you definitely do not get to edit it. (12:28):
undefined
Speaker: Part four the insurance detective and your phone. (12:35):
undefined
Speaker: A ski trip posted to Instagram. (12:43):
undefined
Speaker: Black diamond run. (12:47):
undefined
Speaker: Pure joy. (12:49):
undefined
Speaker: Six weeks later, a back injury from moving furniture. (12:50):
undefined
Speaker: It's legitimate. (12:54):
undefined
Speaker: The MRI confirmed it. (12:56):
undefined
Speaker: Your doctor prescribed treatment, but the insurance (12:57):
undefined
Speaker: denies the claim. (13:02):
undefined
Speaker: Inconsistent with reported physical limitations. (13:04):
undefined
Speaker: They don't mention the ski video. (13:08):
undefined
Speaker: They don't have to. (13:11):
undefined
Speaker: This is how modern insurance investigation works. (13:13):
undefined
Speaker: The ski video was from before the injury. (13:17):
undefined
Speaker: But the algorithm doesn't care about temporal context. (13:20):
undefined
Speaker: It sees athletic capability, (13:25):
undefined
Speaker: risk taking behavior, active (13:28):
undefined
Speaker: lifestyle. (13:30):
undefined
Speaker: It calculates lower payout probability. (13:31):
undefined
Speaker: Property casualty three hundred sixty. (13:36):
undefined
Speaker: An insurance industry (13:38):
undefined
Speaker: publication reported that in (13:40):
undefined
Speaker: twenty twenty five, a social (13:42):
undefined
Speaker: media evidence factors into (13:44):
undefined
Speaker: forty two percent of disputed (13:46):
undefined
Speaker: claims. (13:49):
undefined
Speaker: Investigations. (13:50):
undefined
Speaker: Almost half insurance companies (13:51):
undefined
Speaker: now contract with specialized (13:55):
undefined
Speaker: firms that scan social media for (13:57):
undefined
Speaker: claim verification. (14:00):
undefined
Speaker: They have names like Risk (14:02):
undefined
Speaker: Mitigation Services and Claim (14:04):
undefined
Speaker: Integrity Solutions. (14:08):
undefined
Speaker: What they're looking for depends on the claim for injury cases. (14:11):
undefined
Speaker: It's any physical movement exercise activity for disability (14:17):
undefined
Speaker: claims, it's evidence of unreported work or income. (14:22):
undefined
Speaker: Health insurance applications for lifestyle indicators that (14:27):
undefined
Speaker: suggest higher risk. (14:30):
undefined
Speaker: That marathon you just ran three (14:33):
undefined
Speaker: years ago and posted about to (14:35):
undefined
Speaker: the algorithm. (14:37):
undefined
Speaker: You're either low risk because you're healthy or high risk (14:39):
undefined
Speaker: because athletes get injuries. (14:45):
undefined
Speaker: It depends on what they're trying to prove. (14:48):
undefined
Speaker: The same data tells opposite (14:51):
undefined
Speaker: stories, depending on who's (14:53):
undefined
Speaker: interpreting it. (14:55):
undefined
Speaker: Think about wedding videos on TikTok. (14:56):
undefined
Speaker: Someone with chronic pain having one good day dancing for 30s at (15:00):
undefined
Speaker: their sister's wedding. (15:06):
undefined
Speaker: The investigators seized physical capacity inconsistent (15:08):
undefined
Speaker: with claim the algorithm flags potential fraud indicator. (15:12):
undefined
Speaker: The claim gets denied. (15:18):
undefined
Speaker: They don't see the three hours of pain before the dance. (15:21):
undefined
Speaker: The two days in bed after the (15:25):
undefined
Speaker: context that makes 30s of joy (15:28):
undefined
Speaker: possible and months of (15:30):
undefined
Speaker: suffering. (15:33):
undefined
Speaker: Industry training materials (15:34):
undefined
Speaker: publicly available from (15:36):
undefined
Speaker: investigation certification (15:39):
undefined
Speaker: programs. (15:40):
undefined
Speaker: Teach techniques like Cross, (15:41):
undefined
Speaker: referencing Geotags with claimed (15:44):
undefined
Speaker: limitations. (15:46):
undefined
Speaker: Analyzing Venmo transactions for undisclosed income. (15:48):
undefined
Speaker: Searching tagged photos where (15:52):
undefined
Speaker: the subject appears in other (15:54):
undefined
Speaker: posts and using facial (15:56):
undefined
Speaker: recognition to find unclaimed (15:58):
undefined
Speaker: social profiles. (16:01):
undefined
Speaker: You document your life thinking you're sharing with friends. (16:04):
undefined
Speaker: You're actually building an evidence file that can be read (16:08):
undefined
Speaker: against you by anyone willing to pay for it. (16:13):
undefined
Speaker: Part five the tools they use. (16:19):
undefined
Speaker: Can link cobwebs. (16:26):
undefined
Speaker: Shadow Dragon. (16:29):
undefined
Speaker: They sound like hacker handles from a nineties movie. (16:31):
undefined
Speaker: Well, they're actually companies (16:35):
undefined
Speaker: with government contracts and (16:38):
undefined
Speaker: corporate clients. (16:40):
undefined
Speaker: Shadow Dragon Social Net (16:42):
undefined
Speaker: monitors over two hundred (16:45):
undefined
Speaker: platforms, not just Facebook and (16:47):
undefined
Speaker: Twitter. (16:50):
undefined
Speaker: We're talking about niche forums, regional social (16:51):
undefined
Speaker: networks, places where you forgot you had accounts. (16:56):
undefined
Speaker: It maps relationships, finds alternate accounts, builds what (17:01):
undefined
Speaker: they call patterns of life. (17:07):
undefined
Speaker: Texas Department of Public (17:12):
undefined
Speaker: Safety gave a five million (17:14):
undefined
Speaker: dollars contract for a tool (17:16):
undefined
Speaker: called tangles. (17:19):
undefined
Speaker: It builds relationship webs from social media, finding (17:21):
undefined
Speaker: connections between people who've never met in person but (17:26):
undefined
Speaker: liked the same post. (17:31):
undefined
Speaker: Clearview AI just got a nine (17:34):
undefined
Speaker: million dollars ceiling contract (17:37):
undefined
Speaker: from Ice. (17:39):
undefined
Speaker: This is the company that scraped billions of photos from social (17:41):
undefined
Speaker: media without permission. (17:45):
undefined
Speaker: They built a facial recognition system so powerful it's actually (17:48):
undefined
Speaker: illegal in Europe. (17:54):
undefined
Speaker: You know that photo from your friends Instagram where you're (17:57):
undefined
Speaker: in the background? (18:00):
undefined
Speaker: Clearview has it and it knows it's you. (18:01):
undefined
Speaker: But the newest tool is even more invasive. (18:06):
undefined
Speaker: Location data not from your (18:10):
undefined
Speaker: post, but from your phone (18:13):
undefined
Speaker: itself. (18:16):
undefined
Speaker: A company called Babble Street had to cancel their contract (18:18):
undefined
Speaker: after journalists exposed they were selling location data. (18:23):
undefined
Speaker: So Ice just bought the same capability from Penlink instead. (18:28):
undefined
Speaker: Billions of location signals every day from hundreds of (18:35):
undefined
Speaker: millions of phones. (18:40):
undefined
Speaker: No warrant needed. (18:42):
undefined
Speaker: They can see where you've been, (18:45):
undefined
Speaker: who you've been near, how long (18:47):
undefined
Speaker: you stayed. (18:50):
undefined
Speaker: And most disturbingly, they can look back in time. (18:52):
undefined
Speaker: Part six when context dies. (18:58):
undefined
Speaker: December of twenty thirteen (19:05):
undefined
Speaker: Justine Sacco boards a plane to (19:08):
undefined
Speaker: South Africa. (19:11):
undefined
Speaker: She tweets going to Africa. (19:13):
undefined
Speaker: Hope I don't get Aids. (19:16):
undefined
Speaker: Haha just kidding, I'm white. (19:18):
undefined
Speaker: It was sarcasm. (19:21):
undefined
Speaker: A really bad joke about privilege and inequality at the (19:23):
undefined
Speaker: time of the tweet. (19:27):
undefined
Speaker: She only had one hundred and seventy followers. (19:29):
undefined
Speaker: She then turned off her phone for the eleven hour flight. (19:32):
undefined
Speaker: Fell asleep when she landed. (19:36):
undefined
Speaker: She was the number one trending topic worldwide. (19:39):
undefined
Speaker: Hashtag has just been landed yet. (19:45):
undefined
Speaker: People were waiting at the (19:48):
undefined
Speaker: airport to photograph her (19:50):
undefined
Speaker: reaction. (19:52):
undefined
Speaker: She had been fired while in the air. (19:54):
undefined
Speaker: Death threats filled her inbox. (19:58):
undefined
Speaker: Her life as she knew it was over. (20:00):
undefined
Speaker: Context. (20:05):
undefined
Speaker: Died at thirty thousand feet. (20:06):
undefined
Speaker: This is what happens when human (20:09):
undefined
Speaker: communication meets algorithmic (20:12):
undefined
Speaker: interpretation. (20:14):
undefined
Speaker: Sarcasm becomes statement. (20:16):
undefined
Speaker: Jokes become evidence. (20:20):
undefined
Speaker: Mistakes become permanent. (20:23):
undefined
Speaker: And twenty seventeen ten students had their admissions (20:27):
undefined
Speaker: revoked from Harvard University. (20:31):
undefined
Speaker: Someone leaked screenshots from a private Facebook group where (20:34):
undefined
Speaker: they shared offensive memes. (20:39):
undefined
Speaker: These kids and they were kids (20:42):
undefined
Speaker: eighteen years old thought they (20:45):
undefined
Speaker: were being edgy in a private (20:47):
undefined
Speaker: space. (20:50):
undefined
Speaker: Harvard disagreed. (20:51):
undefined
Speaker: Four years of perfect grades, extracurriculars, SAT prep gone (20:54):
undefined
Speaker: one hundred and forty characters can end a career. (21:01):
undefined
Speaker: A private group chat can derail a future. (21:06):
undefined
Speaker: A photo from five years ago can deny an insurance claim. (21:10):
undefined
Speaker: The algorithm doesn't understand context and strangely enough, (21:16):
undefined
Speaker: increasingly, neither do we. (21:22):
undefined
Speaker: Part seven the authentication of everything. (21:27):
undefined
Speaker: Every platform wants to verify you. (21:33):
undefined
Speaker: Now, blue checks on Twitter, ID verification on Facebook, real (21:36):
undefined
Speaker: names on LinkedIn. (21:42):
undefined
Speaker: They say it's about authenticity. (21:45):
undefined
Speaker: Fighting bots, creating trust. (21:47):
undefined
Speaker: Well it's not. (21:51):
undefined
Speaker: It's about making you survivable. (21:53):
undefined
Speaker: An anonymous account can speak truth to power. (21:57):
undefined
Speaker: A verified account has a home address. (22:01):
undefined
Speaker: China's social credit system (22:05):
undefined
Speaker: seemed dystopian when it (22:08):
undefined
Speaker: launched. (22:10):
undefined
Speaker: Bad social score. (22:11):
undefined
Speaker: Well, you might not be able to buy plane tickets, get a loan, (22:13):
undefined
Speaker: or you might not be able to send your kids to good schools. (22:18):
undefined
Speaker: And we here in the States laughed. (22:22):
undefined
Speaker: That could never happen here. (22:26):
undefined
Speaker: Well it's happening. (22:28):
undefined
Speaker: We just call it different names. (22:31):
undefined
Speaker: Background checks, credit scores. (22:34):
undefined
Speaker: Social media screening. (22:38):
undefined
Speaker: Algorithmic risk assessment. (22:41):
undefined
Speaker: Insurance evaluation. (22:44):
undefined
Speaker: The Chinese were just honest about centralizing it. (22:47):
undefined
Speaker: Pharma or pharma is a major (22:51):
undefined
Speaker: player in social media (22:55):
undefined
Speaker: screening. (22:57):
undefined
Speaker: They advertise that it searches (22:58):
undefined
Speaker: ten thousand online public (23:00):
undefined
Speaker: sources for what it calls (23:03):
undefined
Speaker: behavioral intelligence. (23:05):
undefined
Speaker: According to their own marketing materials. (23:08):
undefined
Speaker: They screen for nine types of (23:11):
undefined
Speaker: workplace misconduct, including (23:14):
undefined
Speaker: fraud, harassment, threats, and (23:17):
undefined
Speaker: violence. (23:20):
undefined
Speaker: They claim a ninety nine point nine five percent accuracy. (23:22):
undefined
Speaker: And they say their AI can (23:27):
undefined
Speaker: identify extremist symbols, (23:29):
undefined
Speaker: violent imagery, memes and (23:32):
undefined
Speaker: gestures. (23:36):
undefined
Speaker: They use what they call avatar (23:38):
undefined
Speaker: recognition to find your (23:41):
undefined
Speaker: accounts, even when you use (23:43):
undefined
Speaker: different names. (23:46):
undefined
Speaker: This is actually from their (23:48):
undefined
Speaker: marketing materials, not (23:50):
undefined
Speaker: speculation. (23:52):
undefined
Speaker: Again, this is from their pitch decks. (23:53):
undefined
Speaker: Apparently another screening (23:57):
undefined
Speaker: company analyzes what they call (23:59):
undefined
Speaker: eleven different behaviors using (24:02):
undefined
Speaker: a billion profile Osint (24:06):
undefined
Speaker: database. (24:08):
undefined
Speaker: They advertise the ability to (24:10):
undefined
Speaker: spot bias threats, political, (24:12):
undefined
Speaker: disparaging and prejudiced (24:16):
undefined
Speaker: speech across text and visual (24:19):
undefined
Speaker: content. (24:21):
undefined
Speaker: Aquasource HR combines AI driven search tools with what they call (24:23):
undefined
Speaker: expert human analysis from trained social anthropologist. (24:31):
undefined
Speaker: Social anthropologist studying (24:37):
undefined
Speaker: your tweets like you're an (24:40):
undefined
Speaker: undiscovered tribe. (24:42):
undefined
Speaker: And here is the number that should terrify everyone. (24:45):
undefined
Speaker: A sterling survey found that sixty eight percent of employers (24:50):
undefined
Speaker: admitted to using social media to find answers to illegal (24:55):
undefined
Speaker: interview questions, questions they can't legally ask you. (25:01):
undefined
Speaker: Age. (25:08):
undefined
Speaker: Religion. (25:09):
undefined
Speaker: Sexual orientation. (25:10):
undefined
Speaker: Health conditions. (25:12):
undefined
Speaker: Sixty eight percent admitted it. (25:14):
undefined
Speaker: These companies defend themselves with the same line. (25:18):
undefined
Speaker: It's all public data. (25:22):
undefined
Speaker: What's public is your entire life, and they've built an (25:26):
undefined
Speaker: industry around reading it. (25:31):
undefined
Speaker: Part eight the deletion myth. (25:36):
undefined
Speaker: You can't delete anything. (25:41):
undefined
Speaker: Not really. (25:44):
undefined
Speaker: Sure, you can remove posts from (25:47):
undefined
Speaker: your timeline, but the data (25:49):
undefined
Speaker: persist. (25:51):
undefined
Speaker: And backups. (25:52):
undefined
Speaker: Caches. (25:53):
undefined
Speaker: Archives. (25:55):
undefined
Speaker: Screenshots. (25:56):
undefined
Speaker: Proliferate. (25:57):
undefined
Speaker: Aggregators. (25:59):
undefined
Speaker: Scrape and store. (25:59):
undefined
Speaker: The European Union has the right to be forgotten. (26:02):
undefined
Speaker: Between twenty fourteen and twenty seventeen, two point four (26:07):
undefined
Speaker: million people requested Google remove their information. (26:12):
undefined
Speaker: Google approved forty three percent, less than half. (26:18):
undefined
Speaker: In America, we don't even have that. (26:23):
undefined
Speaker: No federal right to deletion. (26:26):
undefined
Speaker: No right to see what's in your life. (26:28):
undefined
Speaker: No right to correct errors when you delete a post. (26:31):
undefined
Speaker: Here's what actually happens. (26:36):
undefined
Speaker: The platform removes it from public view. (26:38):
undefined
Speaker: That's all. (26:43):
undefined
Speaker: Server logs remain. (26:44):
undefined
Speaker: Backup systems retain copies. (26:46):
undefined
Speaker: Law enforcement can still access through legal request. (26:50):
undefined
Speaker: Meanwhile, the Internet Archive captures pages. (26:54):
undefined
Speaker: Google caches search results. (26:58):
undefined
Speaker: Third party monitoring services, (27:01):
undefined
Speaker: the ones selling to HR (27:04):
undefined
Speaker: departments and insurance (27:05):
undefined
Speaker: companies have already scraped (27:06):
undefined
Speaker: and stored other user's (27:09):
undefined
Speaker: screenshot before you think to (27:12):
undefined
Speaker: delete. (27:13):
undefined
Speaker: Data brokers operate on a simple (27:15):
undefined
Speaker: principle accumulation without (27:17):
undefined
Speaker: expiration. (27:20):
undefined
Speaker: Companies like LexisNexis and (27:22):
undefined
Speaker: Thomson Reuters aggregate public (27:24):
undefined
Speaker: records social media purchase (27:27):
undefined
Speaker: histories. (27:30):
undefined
Speaker: They create profiles on hundreds of millions of Americans. (27:31):
undefined
Speaker: Profiles you can't see. (27:37):
undefined
Speaker: You can't correct. (27:39):
undefined
Speaker: You can't delete. (27:41):
undefined
Speaker: They sell access to these (27:44):
undefined
Speaker: profiles tens of thousands of (27:45):
undefined
Speaker: times per year to employers, (27:48):
undefined
Speaker: insurers, landlords, anyone with (27:52):
undefined
Speaker: a business account and a credit (27:56):
undefined
Speaker: card. (27:58):
undefined
Speaker: The data multiplies. (27:59):
undefined
Speaker: Every search someone runs on you, generates metadata about (28:01):
undefined
Speaker: who's interested and new. (28:06):
undefined
Speaker: Every analysis creates derived data. (28:09):
undefined
Speaker: Risk scores, behavioral predictions, network maps. (28:13):
undefined
Speaker: You exist in databases you've never heard of, scored by (28:18):
undefined
Speaker: algorithms you'll never see sold to buyers you'll never know. (28:24):
undefined
Speaker: Deletion is theatre. (28:30):
undefined
Speaker: The data persist, and in America you have no right to even know (28:33):
undefined
Speaker: what story it tells. (28:40):
undefined
Speaker: Part nine The Great Quieting. (28:44):
undefined
Speaker: Something is happening to public discourse. (28:51):
undefined
Speaker: People are self-censoring. (28:54):
undefined
Speaker: Not because the government told them to. (28:57):
undefined
Speaker: Because they know everything is watched. (29:00):
undefined
Speaker: Recorded. (29:03):
undefined
Speaker: Analyzed. (29:04):
undefined
Speaker: Scored. (29:06):
undefined
Speaker: A Pew Research study found that seventy two percent of Americans (29:08):
undefined
Speaker: believe their online activities are being tracked by companies. (29:12):
undefined
Speaker: And they are right. (29:18):
undefined
Speaker: But it's what they do with that knowledge that matters. (29:20):
undefined
Speaker: They become careful, calculated, boring. (29:25):
undefined
Speaker: The phrases repeat across social media. (29:30):
undefined
Speaker: I don't post about politics anymore. (29:33):
undefined
Speaker: I deleted all my party photos linked and voice only made a (29:37):
undefined
Speaker: separate account for work. (29:45):
undefined
Speaker: The same story. (29:48):
undefined
Speaker: Different voices. (29:49):
undefined
Speaker: Self-censorship as survival strategy. (29:51):
undefined
Speaker: The chilling effect is real and measurable. (29:56):
undefined
Speaker: When Cambridge Analytica's data harvesting was exposed, Facebook (30:00):
undefined
Speaker: lost two point eight million US users under twenty five. (30:06):
undefined
Speaker: Not just inactive. (30:11):
undefined
Speaker: Gone. (30:13):
undefined
undefined
Speaker: When people learn how their data is used, they don't just change (30:14):
undefined
Speaker: passwords, they change behavior. (30:19):
undefined
Speaker: We are watching democracy's immune system shut down. (30:22):
undefined
Speaker: Descent requires the possibility (30:29):
undefined
Speaker: of having the freedom to be (30:32):
undefined
Speaker: anonymous. (30:34):
undefined
Speaker: Innovation requires the freedom to be wrong. (30:35):
undefined
Speaker: Growth requires the space to be foolish. (30:39):
undefined
Speaker: When everything is permanent and (30:44):
undefined
Speaker: searchable, nothing important (30:47):
undefined
Speaker: gets said. (30:50):
undefined
Speaker: The result is a generation learning to perform stability (30:51):
undefined
Speaker: rather than experience growth. (30:56):
undefined
Speaker: The cure rate for an algorithmic audience that never forgets and (30:59):
undefined
Speaker: never forgives everyone. (31:03):
undefined
Speaker: Self edits now. (31:06):
undefined
Speaker: Not for friends. (31:08):
undefined
Speaker: For the machine that's watching. (31:10):
undefined
Speaker: And the machine is always watching. (31:13):
undefined
Speaker: Part ten the resistance that isn't. (31:20):
undefined
Speaker: People think they are fighting back. (31:26):
undefined
Speaker: Private accounts. (31:29):
undefined
Speaker: Fake names. (31:31):
undefined
Speaker: Deleted apps. (31:34):
undefined
Speaker: Digital detoxes. (31:36):
undefined
Speaker: It doesn't work. (31:39):
undefined
Speaker: Your phone's advertising ID (31:40):
undefined
Speaker: connects everything your credit (31:43):
undefined
Speaker: card links, your purchases, your (31:45):
undefined
Speaker: location. (31:48):
undefined
Speaker: Data tells the real story. (31:49):
undefined
Speaker: Your contacts upload their address book with your number (31:52):
undefined
Speaker: and them you exist and other people's data shadows. (31:57):
undefined
Speaker: Even if you never, ever created accounts. (32:03):
undefined
Speaker: Ghost profiles. (32:08):
undefined
Speaker: That's what Facebook calls them internally. (32:09):
undefined
Speaker: According to leaked documents, they are profiles of people who (32:13):
undefined
Speaker: never signed up. (32:19):
undefined
Speaker: Built from photos of other posts. (32:21):
undefined
Speaker: Contact list. (32:24):
undefined
Speaker: Others. (32:25):
undefined
Speaker: Upload and relationships others document. (32:26):
undefined
Speaker: Now you of course can opt out of Facebook, but you can't opt out (32:30):
undefined
Speaker: of being in Facebook. (32:36):
undefined
Speaker: The real resistance is political. (32:39):
undefined
Speaker: The only real resistance is political. (32:42):
undefined
Speaker: Force with teeth. (32:47):
undefined
Speaker: Regulations that bite. (32:49):
undefined
Speaker: Rights that can't be waived in terms of service. (32:51):
undefined
Speaker: Europe has GDPR. (32:57):
undefined
Speaker: California has CcpA. (33:00):
undefined
Speaker: They're imperfect, but they are (33:03):
undefined
Speaker: something the rest of America (33:06):
undefined
Speaker: has. (33:09):
undefined
Speaker: Thoughts and prayers. (33:10):
undefined
Speaker: Federal privacy legislation has been proposed repeatedly. (33:12):
undefined
Speaker: It never passes. (33:17):
undefined
Speaker: The tech lobby warns it would break the internet. (33:19):
undefined
Speaker: The security lobby says it would help terrorists. (33:22):
undefined
Speaker: First, the business lobby claims economic catastrophe. (33:25):
undefined
Speaker: We can't even agree on basic transparency, whether people (33:31):
undefined
Speaker: should know what data companies collect, whether they should see (33:36):
undefined
Speaker: their own files. (33:41):
undefined
Speaker: The answer is always the same (33:43):
undefined
Speaker: too complicated, too expensive, (33:46):
undefined
Speaker: too dangerous. (33:49):
undefined
Speaker: Translation transparency would reveal how bad it really is. (33:52):
undefined
Speaker: Part eleven the score you can't see. (34:01):
undefined
Speaker: You have a score. (34:08):
undefined
Speaker: Multiple scores. (34:11):
undefined
Speaker: Actually. (34:12):
undefined
Speaker: Credit score. (34:13):
undefined
Speaker: You know that one already. (34:14):
undefined
Speaker: But also customer lifetime value score. (34:16):
undefined
Speaker: Insurance risk score. (34:20):
undefined
Speaker: Employment Probability score. (34:23):
undefined
Speaker: Social influence score. (34:26):
undefined
Speaker: Fraud. (34:29):
undefined
Speaker: Likelihood score. (34:30):
undefined
Speaker: These are not conspiracy theories. (34:33):
undefined
Speaker: These are products sold at industry conferences. (34:36):
undefined
Speaker: Data analytics companies openly (34:41):
undefined
Speaker: advertise predictive scoring (34:44):
undefined
Speaker: systems. (34:46):
undefined
Speaker: Know your customer's true value. (34:48):
undefined
Speaker: Predictive employee churn before it happens. (34:51):
undefined
Speaker: Identify risk before they manifest. (34:55):
undefined
Speaker: HR technology firms claim their (34:59):
undefined
Speaker: algorithms can predict with high (35:02):
undefined
Speaker: accuracy which employees will (35:05):
undefined
Speaker: quit, which will get sick, which (35:08):
undefined
Speaker: will become what they call (35:13):
undefined
Speaker: problematic. (35:15):
undefined
Speaker: How do they do this? (35:18):
undefined
Speaker: Well, they analyzed digital footprints, email patterns, (35:20):
undefined
Speaker: social media activity, even physical workplace data. (35:24):
undefined
Speaker: When available. (35:30):
undefined
Speaker: Badge swipes, parking times, (35:31):
undefined
Speaker: cafeteria purchases the privacy (35:34):
undefined
Speaker: policies are buried in (35:37):
undefined
Speaker: employment agreements. (35:39):
undefined
Speaker: Page forty seven, section three. (35:41):
undefined
Speaker: The part no one reads. (35:44):
undefined
Speaker: You consented without knowing what you consented to. (35:48):
undefined
Speaker: Part twelve tomorrow's crime. (35:55):
undefined
Speaker: Today's punishment. (35:59):
undefined
Speaker: Predictive policing was supposed (36:03):
undefined
Speaker: to stop crime before it (36:05):
undefined
Speaker: happened. (36:07):
undefined
Speaker: Instead, it criminalizes probability. (36:08):
undefined
Speaker: Chicago's heat list algorithm, (36:12):
undefined
Speaker: officially called the Strategic (36:16):
undefined
Speaker: Subject List, identified people (36:19):
undefined
Speaker: likely to be involved in (36:22):
undefined
Speaker: shootings, either as victims or (36:25):
undefined
Speaker: perpetrators. (36:28):
undefined
Speaker: The algorithm did not (36:30):
undefined
Speaker: distinguish if you made the (36:32):
undefined
Speaker: list. (36:34):
undefined
Speaker: Police visited your home, (36:35):
undefined
Speaker: knocked on your door, told you (36:38):
undefined
Speaker: they were watching the (36:41):
undefined
Speaker: algorithm, considered arrest, (36:44):
undefined
Speaker: not convictions. (36:47):
undefined
Speaker: Social networks. (36:49):
undefined
Speaker: Geography. (36:51):
undefined
Speaker: Age. (36:52):
undefined
Speaker: Being young, black, and living in certain neighborhoods was (36:54):
undefined
Speaker: enough to score high. (36:58):
undefined
Speaker: The program was discontinued after a Rand Corporation study (37:01):
undefined
Speaker: found it ineffective, and civil rights groups proved it was (37:05):
undefined
Speaker: algorithmic racial profiling. (37:10):
undefined
Speaker: But the concept didn't die. (37:13):
undefined
Speaker: It evolved. (37:16):
undefined
Speaker: Now it's person based predictive analytics. (37:18):
undefined
Speaker: It's the same idea, but better branding. (37:23):
undefined
Speaker: Police departments nationwide use variations. (37:27):
undefined
Speaker: Ice also uses it as well as border patrol. (37:31):
undefined
Speaker: Your social media feeds these (37:35):
undefined
Speaker: systems that protest you (37:38):
undefined
Speaker: attended, that article you (37:41):
undefined
Speaker: shared. (37:43):
undefined
Speaker: That friend who got arrested all became inputs to algorithms (37:44):
undefined
Speaker: calculating threat probability. (37:50):
undefined
Speaker: The companies building these systems claim accuracy rates (37:54):
undefined
Speaker: above eighty five percent. (37:58):
undefined
Speaker: They say the algorithms identify patterns humans miss. (38:01):
undefined
Speaker: What patterns? (38:07):
undefined
Speaker: Well, they won't say trade secrets. (38:08):
undefined
Speaker: Proprietary methods. (38:12):
undefined
Speaker: The algorithm generates a score. (38:14):
undefined
Speaker: The score generates an action, and no one knows exactly why. (38:18):
undefined
Speaker: Not even the people using it. (38:25):
undefined
Speaker: Part thirteen the Future that's already here. (38:30):
undefined
Speaker: William Gibson said. (38:37):
undefined
Speaker: The future is already here. (38:39):
undefined
Speaker: It's just unevenly distributed. (38:42):
undefined
Speaker: He was right, but he was also wrong. (38:45):
undefined
Speaker: The surveillance future isn't unevenly distributed. (38:49):
undefined
Speaker: It's universal. (38:55):
undefined
Speaker: We all live in it. (38:56):
undefined
Speaker: We just experience it differently based on our scores. (38:58):
undefined
Speaker: If your scores are good, (39:03):
undefined
Speaker: employed, insured, documented, (39:05):
undefined
Speaker: compliant, you might never (39:08):
undefined
Speaker: notice the cage doors are open (39:10):
undefined
Speaker: for you. (39:13):
undefined
Speaker: Services appear. (39:14):
undefined
Speaker: Life just feels frictionless. (39:15):
undefined
Speaker: If your scores are bad, every interaction is friction. (39:19):
undefined
Speaker: Every application is denied. (39:24):
undefined
Speaker: Every benefit requires proof. (39:27):
undefined
Speaker: Every movement is questioned. (39:30):
undefined
Speaker: The architecture of surveillance is complete. (39:34):
undefined
Speaker: The only question is how it will be used in the future. (39:38):
undefined
Speaker: ICE's social media monitoring program isn't the beginning. (39:43):
undefined
Speaker: It's the formalization of what's already happening. (39:48):
undefined
Speaker: The government is just catching up to what corporations have (39:52):
undefined
Speaker: been doing for a decade. (39:57):
undefined
Speaker: Your posts are being watched, (39:59):
undefined
Speaker: your patterns analyzed, your (40:02):
undefined
Speaker: future predicted, your worth (40:05):
undefined
Speaker: calculated. (40:08):
undefined
Speaker: Right now, as you listen to this podcast, an algorithm somewhere (40:10):
undefined
Speaker: is updating your file. (40:17):
undefined
Speaker: You can't see it. (40:20):
undefined
Speaker: You can't correct it. (40:21):
undefined
Speaker: You can't escape it. (40:23):
undefined
Speaker: But you should know it exists. (40:25):
undefined
Speaker: Because in a world where everything is remembered and (40:28):
undefined
Speaker: nothing is forgiven, the most radical act isn't revolution. (40:32):
undefined
Speaker: It's remembering that you are (40:39):
undefined
Speaker: more than the sum of your data (40:42):
undefined
Speaker: points. (40:44):
undefined
Speaker: You are a human being. (40:45):
undefined
Speaker: Complex. (40:48):
undefined
Speaker: Contradictory. (40:49):
undefined
Speaker: Capable of change. (40:51):
undefined
Speaker: The algorithm doesn't know that. (40:54):
undefined
Speaker: But you do. (40:57):
undefined
Speaker: And maybe, just maybe, that's enough to start with. (40:59):
undefined
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Are You A Charlotte?
In 1997, actress Kristin Davis’ life was forever changed when she took on the role of Charlotte York in Sex and the City. As we watched Carrie, Samantha, Miranda and Charlotte navigate relationships in NYC, the show helped push once unacceptable conversation topics out of the shadows and altered the narrative around women and sex. We all saw ourselves in them as they searched for fulfillment in life, sex and friendships. Now, Kristin Davis wants to connect with you, the fans, and share untold stories and all the behind the scenes. Together, with Kristin and special guests, what will begin with Sex and the City will evolve into talks about themes that are still so relevant today. "Are you a Charlotte?" is much more than just rewatching this beloved show, it brings the past and the present together as we talk with heart, humor and of course some optimism.