If you have been following the Tilly Norwood saga, you might have noticed something odd on April 1. The AI actress and singer announced she was shutting down, effective immediately:

After careful consideration, we’ve made the difficult decision to deactivate Tilly Norwood indefinitely and retreat from the public eye.

While Tilly remains grateful for the support of her fans, the volume of negativity and criticism has become impossible to ignore. It’s clear that the world is simply not yet ready.

Thank you for being part of her journey.

As you might have guessed, Tilly’s demise was an April Fool’s joke, one aimed at the countless detractors of Norwoods particular brand of uncanny valley video.

Norwood’s creators claim that 18 humans were involved in the production of her videos, from production designers to costumers to prompters. They also claim that a human actor took part in the production, probably in some form of motion capture. It’s unclear how much of Tilly is real but we can assume that no human was involved in writing “They think I’m just a dollar dream/ But I’ve got more than they believe/ I’m not a puppet, I’m the star/ Guided by the heart, that's who we are.”

As expected, the “shutdown” was an April Fool’s joke, aimed as much at her critics as at her fans. It also served as a reminder of how strange the reaction to Norwood has been. Few synthetic performers have drawn this level of attention, or this level of irritation.

Norwood’s creators claim that as many as 18 people were involved in producing her videos, including production designers, costumers, and prompt writers. They also suggest that a human performer contributed, likely through motion capture or reference acting. What remains unclear is how much of Tilly is authored by humans and how much is generated. It is safe to assume, however, that no songwriter sat down and earnestly wrote lines like: “They think I’m just a dollar dream / But I’ve got more than they believe.”

Tilly is the product of Particle6, a London-based studio focused on building AI actors and comedians. She functions as both a proof of concept and a marketing tool, a way to show that AI video production can be structured, staffed, and, at least in theory, controlled. The goal is not just to entertain but it is to convince skeptics, particularly in film and advertising, that this kind of production is viable and not simply a form of automated plagiarism. Tilly’s creator, Eline van der Velden, even claims to have received death threats.

“The death threats and the hate … my goodness, loads,” she said. “That’s still going, but much less now. I totally understand the reaction. I made her to represent the fear – and to represent this change in AI as an art piece, in a way. I was just like, this is the zeitgeist.”

Van der Velden, who founded the production company Particle6 and its AI arm, Xicoia, debuted Norwood in a short comedy sketch called AI Commissioner last autumn. Norwood has her own social media accounts, with 141,000 followers on Instagram.

An immediate backlash came from actors including Melissa Barrera, Mara Wilson and Ralph Ineson. The actors’ unions Sag-Aftra and Equity also raised concerns. When asked about Norwood on a podcast, the actor Emily Blunt said: “Good lord, we’re screwed.”

Van der Velden said it had been a deliberately provocative exercise to shock the industry into understanding how far the technology had come. “You’re trying to provoke thoughts and discussion,” she said. “So yes, I wanted to make people sit up and go ‘holy moly’. So in a way I achieved my goal.”

The response has been uneven. When one of Norwood’s videos appeared around the time of the Oscars, critics focused on what they saw as a lack of presence. The performance felt hollow and the music felt assembled rather than written. There was a sense that something essential was missing, even if viewers struggled to define what that was.

The criticism has not been limited to aesthetics. It has also drifted into philosophy and even theology. One essay, published in a religious context, compared Tilly to the early Christian heresy of Docetism, the belief that Christ only appeared to have a physical body and did not truly exist in material form. The argument is not entirely absurd. Tilly presents the image of a human being without any of the underlying reality. She appears to feel, to perform, to exist, but there is nothing behind the surface.

The need for perfection seen in this latest AI arrival, or even in social media’s filtered, perfected photos, can be seen as yet another iteration of the gnostic heresy of Docetism. This early second-century argument saw matter and the physical world as inherently evil. Only the spiritual world had value and was redeemable. Because the physical world was evil, Jesus was therefore not incarnated into a real human body. Instead, he just appeared to have one.

Rather than sullying his heavenly self with the dirtiness of humanity, he just appeared to participate in our matter-filled world. Jesus didn’t get hurt, grow weary, feel hungry, or age. He just appeared to do all those things. Some who believed this heresy even went so far as to deny the Resurrection. If Jesus only appeared to suffer and die on a cross, they reasoned, then he wouldn’t need to rise from the dead. Similarly, Tilly Norwood just appears to have a human body, but she doesn’t. Not really. Her shiny, perfect façade has no connection to the human condition.

This may be the first time, outside of early reactions to the printing press or photography, that a new medium has been criticized not for failing to capture reality, but for simulating it too cleanly. The complaint is not that it is broken but that that it lacks friction, and without friction, it feels unreal.

Tilly Norwood is not going away. The April Fool’s post made that clear. If anything, the reaction proves the opposite. She has found a nerve, and for now, she’ll on the last of it.

News Roundup for April 2, 2026

Chuck Norris’ family left furious over AI-generated videos of star after his death

The family of Chuck Norris has spoken out after a wave of AI-generated videos spread false claims about his death. Norris died at age 86 following a medical emergency in Hawaii, and his family said online clips have circulated misleading details about his health, the circumstances of his passing, and even family relationships. In a public statement, they urged fans to rely only on official sources and not to share unverified content.

The response comes amid widespread online attention following Norris’s death, with tributes from public figures and emotional messages from his children. While the family reflected on his life and legacy, they also pushed back against what they described as fabricated narratives driven by AI content. The situation highlights how quickly misinformation can spread after a public figure’s death, especially when synthetic media is involved.

Google Faces Calls to Prohibit AI Videos for Kids on YouTube

More than 200 child development experts, advocacy groups, and educators are urging Google and its video platform YouTube to stop showing or recommending AI-generated videos to children. In a letter addressed to CEO Sundar Pichai and YouTube head Neal Mohan, the group warned that much of this content, often labeled “AI slop,” is low quality, mass-produced, and designed to capture attention rather than educate. They argue that these videos may harm children’s attention spans, blur the line between reality and fiction, and replace activities that are important for social and emotional development. The letter was signed by figures including Jonathan Haidt, along with organizations like Fairplay and the American Federation of Teachers.

YouTube responded by saying it already limits AI-generated content in YouTube Kids to select channels and requires labeling of synthetic material, while also allowing parents to block content. Still, advocates argue these safeguards are not enough, especially for very young viewers who cannot understand such labels. The debate comes as AI-generated children’s content grows rapidly, driven by creators who can produce videos cheaply at scale, and as broader scrutiny mounts over social media’s impact on young users. Google has also invested in AI-driven kids’ content through companies like Animaj, a move critics say contradicts concerns about screen time and child safety.

Sora is gone — but these 3 AI video tools are already replacing it

The TechRadar piece argues that while OpenAI’s Sora has been shut down, the broader AI video space is not slowing down, it is fragmenting and maturing. Instead of one dominant tool, users are now choosing from several specialized platforms that match or exceed Sora’s capabilities in different areas. The article highlights three main replacements: Google’s Veo 3.1, which focuses on high-quality, multi-scene video generation with integrated audio; Runway Gen-4.5, aimed at more advanced users who want deeper editing control; and Kling AI 3.0, a simpler, more accessible tool that can generate longer clips and is easier for beginners to use.

The takeaway is that Sora’s disappearance does not mark the end of AI video, but a shift in how the market is structured. Instead of a single, general-purpose tool, the space is now defined by competing systems with different strengths, pricing models, and levels of complexity. This makes the ecosystem harder to navigate, but also more flexible, with options tailored to both casual creators and professionals.