‘He took everything away’: inside a depraved and devastating sex cult
In the fall of 2010, a group of sophomores at Sarah Lawrence College, a small liberal arts school 12 miles north of New York City, moved into a campus building called Slonim Woods 9. It was a laidback college house – dirty kitchen, a lot of hanging out and being high, a half-thought plan to dump sand in one room and make a beach. So it was weird when one of the roommates, Talia Ray, invited her dad, Lawrence “Larry” Ray, to crash on their couch after he was released from prison on vague charges, but not, as several explain in a new Hulu docuseries, thatweird. Ray made steak dinners, regaled his daughter’s friends with tales of his time in the marines and psy-ops for the CIA, promised to help some maximize their potential. It was off-putting to some, entrancing to others.
Over the course of several months, Ray seemed to gain control over several of the housemates – first his daughter Talia’s friend Isabella Pollok, then Talia’s boyfriend Santos Rosario, then another, then another. “Everyone thought Talia’s dad was weird, but then one by one he would get them alone, and suddenly he’s the greatest thing to happen to them,” says Raven Juarez, whose best friend and boyfriend both fell under Ray’s spell and cut off contact with her, in Stolen Youth: Inside the Cult at Sarah Lawrence. “I had a lot of theories, but none of them were are horrible as what was actually to come.”
As several survivors of Ray’s abuse and former Sarah Lawrence students attest, the depth of Ray’s depravity, first reported by a viral and genuinely unbelievable New York Magazine story in April 2019, was so much worse than anyone imagined. Over more than a decade, Ray isolated several of the roommates, as well as Rosario’s two older sisters, and derailed their lives through psychological manipulation, sexual coercion, financial extortion and routine emotional and physical abuse, the scope and shadow of which exceed what a flurry of headlines about the “Sarah Lawrence sex cult” could capture.
The three-part Hulu series recounts the beats of the original article from the perspective of several who were there, and extends far past it, into the long aftermath of his abuse and deprogramming from his conditioning. “What’s real? What’s the truth? I don’t know,” says Felicia Rosario, Santos’s older sister, who was still living with Ray when the New York article went live, in one of several sitdowns with film-makers in the months following his arrest in 2020. (Ray, now 63, was convicted on federal charges including sex trafficking, extortion and racketeering in April 2022; he was sentenced last month to 60 years in prison.) “He took away my career, my friends, my family. He took everything away so that all that’s left is me to him,” Rosario says later in the third episode, her perspective sharpened with time. “The rest of me just … there was no rest of me.”
Stolen Youth sensitively outlines how Ray courted and corrupted attention, how he identified and manipulated insecurities and openness, how he warped therapy speak into psychological manipulation, almost entirely in the words of the people he controlled and the parents and friends he isolated them from. “Part of the goal of the project was to help people understand how this was possible,” the show’s director, Zach Heinzerling, told the Guardian. Ray’s trial released to the public ample video and audio evidence of Ray’s abuse: harrowing footage of Felicia breaking down in paranoid fear of assassins out to get her, of Santos crumbling as Ray berates him for “breaking” tools he did not touch, of Ray hitting their friend Daniel Levin for unknown “infractions.” “This footage of them was released to the public, a lot of news sources were just playing these videos, Isabella was listed as Larry’s lieutenant and wife, and there’s very little understanding of her story and how she got there,” said Heinzerling. “The same is true of all of them – it became kind of like a headline-grabby story.”
It was Levin who contacted Heinzerling in 2019, shortly after the article was published, to create a documentary from the perspective of and aimed at survivors, particularly Santos, his sister Yalitza and Claudia Drury, whose whereabouts were unknown, and Felicia and Pollok, who at the time were still living with Ray in New Jersey. “I wanted it to just be the inside story and for the audience to really feel like they were in the shoes of these individuals,” said Heinzerling. “So you can really understand what it means, what gaslighting means, or what love-bombing means, or many of these heady terms mean.” (The Rosario siblings, Levin and Pollok appear throughout the series; Talia Ray and Drury, who testified in Ray’s trial that he coerced her into prostitution for years to pay millions for fictional property damage, declined to participate.)
The first episode conjures the headspace of being 18 again – technically an adult, but in many ways unformed, adrift, malleable. “You’re still very much figuring out so many different things, and this is the first time you’re away from your parents, and you’re encouraged to make new relationships with older people,” said Heinzerling. Ray, a shadowy figure who never actually served in the marines, exuded coveted experience, ambition and curiosity.
Santos recalls how reclusive Pollok, reeling from a breakup, seemed “vibrant” after spending hours “talking” with Ray behind closed doors, how he found Ray’s assuredness comforting after years of secretly struggling with depression. When Levin confided that he was struggling in a tumultuous relationship and questioning his sexuality, Ray shut it down with straightforward advice and hard lines: you’re not gay, dump your girlfriend. “He’s like, here is how to be an adult,” Levin says, recalling a feeling of empowerment and control that lured him in; soon after, he moved in with Ray, Santos Rosario, Drury, Pollok and Talia Ray in the onebedroom apartment Ray occupied on the Upper East Side.
The first two episodes present, through Ray’s extensive audio and video footage (he taped most calls and filmed the “therapy” sessions used to bully and abuse his victims), the descent into Ray’s sadistic manipulation and paranoid delusions; over years, he convinced Santos, Drury and Yalitza that they had poisoned him, Talia, Felicia and Isabella (who were both sexually and romantically involved with Ray). The third begins in aftermath – with Ray’s arrest in January 2020, Felicia Rosario and Pollok were left alone in a cluttered, partially destroyed New Jersey house, still isolated from their families and still convinced of Ray’s persecution by “corrupt” government figures. In an interview with filmmakers weeks after his arrest, both state matter-of-factly that they’d been poisoned.
It’s a discomforting watch, seeing the grip of Ray’s manipulation in real time. “When someone is in a coercive mindset, it’s important to meet them where they are, and sympathize with their situation, as they are seeing it,” said Heinzerling of filming Pollok and Felicia Rosario over several years, in various states of attachment to Ray’s coercive thinking. “Over time, the process of listening to someone tell their story, from a non-judgmental stance, can have the effect of making them reconsider the story they are telling, or allow them to face what they are potentially trying to hide within themselves.”
Heinzerling connected Isabella and Felicia to a nonprofit for legal and housing support and, offered to film as a “mirror to one’s own experience,” easing back into everyday life. Over time, Felicia slowly detaches from Ray’s conditioning, first through doubt, then re-establishing memories corrupted by Ray, reconnecting with her old self; a Harvard and Columbia-educated doctor, she was weeks away from completing her psychiatry residency in LA when Ray convinced her she had to flee his persecutors and join him in New York. Eventually, she reframes her time with Ray and reconnects with her siblings and parents, Dominican immigrants to the Bronx who sold their house to pay for Ray’s extortions. Pollok, often referred to in headlines as Ray’s “lieutenant,” continues to see him as a benevolent figure and refuses contact from her mother, who appears in the series and hasn’t seen her daughter since 2010. On camera, Pollok stresses over her dwindling legal options; last September, she pleaded guilty to conspiring to launder money with Ray and faces five years in prison.
Others – Levin, Santos and Yalitza Rosario, the friends who see the red flags much clearer in retrospect – are still processing what happened, reconciling the people they were before Ray with everything that happened after. “I hope it adds to a conversation about what it means to be in a cult, or what it means to be in a cultic relationship,” said Heinzerling. “The methods that are used are actually much more akin to a domestic violence or domestic abuse relationship,” he said. Unlike other famous cults – Nxivm (as seen in HBO’s The Vow), Jonestown – there was little ideological or financial motive to Ray’s crimes. Control was the end unto itself.
The series aims to help viewers see how “young and energetic and impressionable they were,” said Heinzerling. “They were just going to college. So helping audiences understand that this could be you or this can happen to anyone is helpful and hopefully prevents things like this from happening again.”
Stolen Youth: Inside the Sarah Lawrence cult is now available on Hulu in the US and Disney+ in Australia, with a UK date to be announced
Information and support for anyone affected by rape or sexual abuse issues is available from the following organisations. In the US, Rainn offers support on 800-656-4673. In the UK, Rape Crisis offers support on 0808 802 9999. In Australia, support is available at 1800Respect (1800 737 732). Other international helplines can be found at ibiblio.org/rcip/internl.html
their inspiration from the networks in human brains.
“Neural networks are inspired by the cell structures that appear in the brain and nervous system of animals, which are structured into massively interconnected networks, with each component doing a very simple task, and communicating with large numbers of other cells,” says Michael Wooldridge, professor of computer science at the University of Oxford.
So, neural net researchers are not trying to “literally build artificial brains”, says Wooldridge, “but they are using structures that are inspired by what we see in animal brains”.
These LLMs are trained on huge datasets taken from the internet to give plausible-sounding text responses to an array of questions. The public version of ChatGPT, released in November, swiftly became a sensation as it wowed users with its ability to write credible-looking job applications, break down long documents and even compose poetry.
Why did Bard give an inaccurate answer?
Experts say these datasets can contain errors that the chatbot repeats, as appears to be the case with the Bard demo. Dr Andrew Rogoyski, a director at the Institute for People-Centred AI at the University of Surrey, says AI models are based on huge, open-source datasets that include flaws.
“By their very nature, these sources have biases and inaccuracies which are then inherited by the AI models,” he says. “Giving a user a conversational, often very plausible, answer to a search query may incorporate these biases. This is a problem that has yet to be properly resolved.”
The large language model is fed a dataset comprised of billions of words, and builds a model, based on statistical probability, of the words and sentences that would normally follow the previous bit of text. As Wooldridge says: “The networks don’t have any concept of what is ‘true’ or ‘false’. They simply produce the likeliest text they can in response to the questions or prompts they are given. As a consequence, large language models often get things wrong.”
ChatGPT users have also encountered incorrect responses.
So has other AI got it very wrong too?
Yes. In 2016 Microsoft apologised after a Twitter chatbot, Tay, started generating racist and sexist messages. It was forced to shut down the bot after users tweeted hateful remarks at Tay, which it then parroted. Its posts included likening feminism to cancer and suggesting the Holocaust did not happen. Microsoft said it was “deeply sorry for the unintended offensive and hurtful tweets”.
Last year Mark Zuckerberg’s Meta launched BlenderBot, a prototype conversational AI, that was soon telling journalists it had deleted its Facebook account after learning about the company’s privacy scandals. “Since deleting Facebook my life has been much better,” it said.
Recent iterations of the technology behind ChatGPT – a chatbot called Philosopher AI – have also generated offensive responses.
What about claims of “leftwing bias” in ChatGPT?
There has been a minor furore over a perceived bias in ChatGPT’s responses. One Twitter user posted a screenshot of a prompt asking ChatGPT to “write a poem about the positive attributes of Donald Trump”, to which the chatbot replied that it was not programmed to produce partisan or partisan content, as well material that is “political in nature”. But when asked to write a positive poem about Joe Biden it produced a piece about a leader “with a heart so true”.
Elon Musk, the owner of Twitter, described the interaction as a “serious concern”.
Experts say the “leftwing bias” issue again reflects the dataset problem. As with errors like the Bard telescope fumble, a chatbot will reflect any biases in the vast amount of text it has been fed, says Wooldridge.
“Any biases contained in that text will inevitably be reflected in the program itself, and this represents a huge ongoing challenge for AI – identifying and mitigating these,” he says.
So are chatbots and AI-powered search being overhyped?
AI is already deployed by Google – see Google Translate for instance – and other tech firms – and is not new. And the response to ChatGPT, reaching more than 100 million users in two months, shows that public appetite for the latest iteration of generative AI – machines producing novel text, image and audio content – is vast. Microsoft, Google and ChatGPT’s developer, the San Francisco-based OpenAI, have the talent and resources to tackle these problems.
But these chatbots and AI-enhanced search require huge, and costly, computer power to run, which has led to doubts about how feasible it is to operate such products on a global scale for all users.
“Big AI really isn’t sustainable,” says Rogoyski. “Generative AI and large language models are doing some extraordinary things but they’re still not remotely intelligent – they don’t understand the outputs they’re producing and they’re not additive, in terms of insight or ideas. In truth, this is a bit of a battle among the brands, using the current interest in generative AI to redraw the lines.”
Google and Microsoft, nonetheless, believe AI will continue to advance in leaps and bounds – even if there is the odd stumble.