AI Insights
‘I was constantly scared of what she was going to do’: the troubled life and shocking death of Immy Nunn | Mental health
Just a few hours before she ended her life, Immy Nunn seemed happy. She and her mother, Louise, had been shopping and had lunch. It was the final day of 2022 and Immy, who was 25, appeared positive about the new year. She talked about taking her driving test and looking for a new flat. She was excited about the opportunities her profile on TikTok was bringing her; known as Deaf Immy, she had nearly 800,000 followers, attracted by her honest and often funny videos about her deafness and her mental health.
By the early hours of the next morning, Immy was dead, having taken poison she bought online, almost certainly after discovering it through an online pro-suicide forum.
On a sunny day, kitchen doors open to the garden, Louise sits at her table; every so often she glances at the photographs of her daughter. Immy’s assistance dog, Whitney, now lives with her parents, and wanders around, stopping occasionally to be stroked. Louise describes these last couple of years as: “Hell. Horrible.” The pain of losing her child, she says, “you wouldn’t wish on anyone”. She copes, she says, “day by day. I struggle with a lot of things. I don’t like doing a lot.”
For the previous 10 years, Louise had been on high alert, always terrified something would happen to her daughter. Since she was about 14, Immy had periods of severe mental illness. She had self-harmed, and attempted suicide many times, and for four years she had been an inpatient at a psychiatric hospital.
She had spent the Christmas of 2022 at her parents’ home in Bognor Regis, West Sussex, then gone back to her flat in Brighton. On 29 December, she had cut herself and gone to hospital – as far as her family knew, it was the first time she had self-harmed in ages. Immy’s dad, Ray, went straight to see her and tried to get her to come home with him, but she told him she wanted to stay, and that she had an appointment with one of her support workers the following day. On 31 December, Louise and Ray went to spend the day with her in Brighton. They returned to Bognor Regis with Whitney because Immy was going to a New Year’s Eve party at a friend’s house in nearby Shoreham-by-Sea.
Louise was woken about 5am by the mother of Immy’s friend calling to say Immy had left unexpectedly, and without her coat and shoes. They had known Immy since she was a child, and were aware of her mental health problems. Louise phoned the police straight away and kept trying to ring Immy; Ray went out to look for their daughter, eventually driving to her flat in Brighton. When he arrived, the police and an ambulance were already there.
Immy’s devastated family is one of several that appear in a two-part Channel 4 documentary, Poisoned: Killer in the Post. It is based on an investigation by the Times journalist James Beal, which started after he was contacted by David Parfett, whose son Tom also died after taking a substance he bought online. The documentary shows the impact on vulnerable people of a pro-suicide forum where methods were discussed, including signposting to a Canadian chef, Kenneth Law, who Canadian police believe shipped about 1,200 packages of poison around the world. In the UK, the National Crime Agency has identified 97 potential victims. Law is awaiting trial in Canada, charged with 14 counts of murder – the dead were in the Ontario area and between the ages of 16 and 36 – but is pleading not guilty. About five months after Immy’s death, the police told Louise and Ray that they had been given a list of names of British people linked to Law, and Immy was on it. They were doing checks, Louise says the police told her, to see who on the list was still alive.
Louise would like to see Law extradited to the UK, though she knows this is unlikely. For a decade, she and her family went through heartbreaking effort to try to keep Immy safe. “And then it’s someone online. You fear the man on the corner, don’t you, but not the man you can’t see?” And she would like to see more regulation of sites that can be harmful to vulnerable people. “The [government] are allowing them; no one’s stopping them from doing it.” The site Immy is believed to have accessed is now under investigation by Ofcom; as of 1 July, the site was no longer accessible to people in the UK.
A journalist had showed Louise the site, and she was shocked at how accessible it was. “It wasn’t even on the dark web,” she says. “I was just shocked that something like that is just there. How is it even allowed?” Vulnerable people who are struggling understandably might want to find others who are feeling the same, but the site encourages and facilitates suicide – methods are discussed and tips swapped, and the “goodbye” posts are met with congratulatory messages. As for Law, Louise says: “I hate him. Hate the sound of his name, hate seeing his face.”
Immy was always a fighter, Louise says. She had been born six weeks early and spent her first couple of weeks in hospital. The fourth of her five children, Immy had siblings who doted on her. “She was just beautiful,” Louise says of Immy as a baby. “She was so good and happy; everything about her was just perfect.” The family found out that Immy was profoundly deaf when she was 18 months old, though Louise suspected it already (one of her older children also has hearing loss, though not to the extent Immy did). Having a child with additional needs meant they spent a lot of time together. When Immy was three, she had cochlear implants, which involved trips to Great Ormond Street hospital in London every few weeks.
She was happy at school, Louise remembers. It was a mainstream school but with a unit for the several deaf children there at the time. Then, when Immy was about 13, Louise noticed a change in her. Some of her deaf friends had left, and Immy stopped seeing other friends. “You just thought: ‘Typical teenager’, until one day I saw cuts on her legs and I realised that there was something going on,” says Louise. She had been running away from school, and was clearly unhappy there. She had an appointment with the Child and Adolescent Mental Health Services but refused to go, then took her first overdose shortly before she turned 15. “I thought she was dead at that point,” says Louise. “Reality hits – this is really serious.”
The National Deaf Children’s Society helped Louise advocate for Immy at school, and find her a place at a leading school for deaf children, but it took a while, and Immy’s mental health was deteriorating. After school one day, Louise could hear her in the bathroom and became worried about what she was doing, but couldn’t get her to come out. Immy’s older sister went in and found she had cut her arm badly. “I just remember her face and her saying, ‘Mum, you need to get her to hospital straight away.’ I was constantly scared of what she was going to do.”
There were other suicide attempts. Ray is a roofer and Louise had worked part-time in a shop, around looking after the children, but she gave that up to be there for Immy. “If she was at home, you wouldn’t leave her for second,” she says.
Immy was in and out of children’s mental health units and then got a place in a unit for deaf children in London. “We would go up two, three times a week to visit and she was doing really well, but she could only stay there until she was 18,” says Louise. Once Immy was discharged, Louise says there was no follow-up care and she was instead put on unfamiliar medication, which she had a terrible reaction to. “We ended up right back where we were. She was in her room smashing things over her head, blood everywhere.”
The following year, Immy was back in psychiatric hospital, where she would be for the next four years. The family hoped it would be the start of Immy getting better, but it was also, says Louise, “four years of hell. We just didn’t know when you were going to get a phone call.” On the weekends she was allowed home, Louise would sleep in her room with her “because I was so scared of what she’d get up and do”.
Immy had been diagnosed with emotionally unstable personality disorder, PTSD and other conditions including depression and anxiety. There were periods when she was well and she seemed happy; she had a girlfriend for a while. “She’d have really good days; you’d be able to go on holiday and have fun times. But you just never knew when her mind was going to suddenly hurt herself, and she didn’t know. That was the scary thing. She’d just dissociate.”
Starting a TikTok account in 2020 helped her, Louise says. “It took her mind off things. Obviously, she was still really poorly. She’d have her good days and bad days. But I think because of the followers that grew, she felt she could help other people. As her followers grew, her confidence grew, and I think she felt as if she’d finally found something that she could do.” It helped her embrace the deaf and LGBTQ+ communities and gave her a sense of identity. “She felt as if she belonged, whereas she never really knew where she belonged.”
Immy showed her followers what life in a psychiatric hospital was like, and was open about her struggles. But she could also be joyful, and often got her family involved, usually her mum. “You’d be sat in the evenings, and she’d say, ‘Mum: I’ve got an idea – I want you to be in it.’ I loved watching her laugh.” Immy was getting brand and charity collaborations, and positive messages from people who said she’d helped them. “She just couldn’t believe it, and we were just so excited for her,” says Louise.
She was desperate to try to live more independently, even though Louise thought she wasn’t ready to leave hospital. “She was determined. She’d been in there for four years; she wanted out, she wanted a normal life.” It was a worry, she says, having Immy live an hour away in Brighton, and she would video-call her often – again and again if she didn’t pick up. “She didn’t want me to keep worrying. She was like, ‘Mum: I’m 24 – let me have my life.’” And she seemed to be doing well, though Louise could never relax.
Early in 2022, Immy took an overdose. Nine months after that, in November, she told her support worker she had been on a pro-suicide forum and had bought poison from it. Louise didn’t know about this until just before the inquest. The police went to do a welfare check on Immy, but didn’t take a British Sign Language (BSL) interpreter – something Louise was familiar with in all the years of trying to get Immy the care she needed. She would go to see doctors with her, she says, and there would be no interpreter. Louise would have to accompany Immy, even when Immy didn’t want her to, so that she could explain things to her. After that police visit, Immy wasn’t seen by a mental health professional for several weeks.
A few days after Christmas with her parents, Immy harmed herself and went to hospital but left before being seen by the mental health team. She told her parents that she’d been in hospital, and Ray immediately went to see her. “We didn’t know how bad she was,” says Louise. “The plan was that he was going to bring her home, but she said she wasn’t coming back.” Of course they were alarmed, but sadly this wasn’t out of the ordinary for Immy. “She self-harmed a lot. That was her coping mechanism. We had no clue that anything else was going on.”
Immy had sent a text to her support worker, saying she thought she needed to be admitted to psychiatric hospital and that she “could easily go to the last resort” even though she didn’t want to. In another message to her psychologist the following day, she said she planned to take poison, but also said she didn’t have any (she did – it was later discovered she had already bought some online). She agreed to be admitted to a mental health crisis facility, but that didn’t happen that day. A meeting that she was supposed to have with her care coordinator also didn’t happen. The inquest found failings in mental health care contributed to Immy’s death. The coroner also highlighted systemic challenges to deaf patients, particularly the shortage of BSL interpreters. With grim irony, the inquest itself had to be adjourned at one point because of a lack of interpreters.
Louise says the family has received no apology. The trial of Law isn’t due to start until early next year, and he has been charged only over deaths in Canada. She says she feels stuck. “I always feel as if I’m waiting for the next thing. It’s just hard.”
She likes to talk about Immy, but she finds it hard to watch her videos. “The dogs start crying when they hear her voice, especially Whitney – she still recognises Immy’s voice, and then that upsets me.” There are some lovely videos of Immy and her mum together, including the two of them singing and signing You Are My Sunshine – the first song, Immy wrote, that her mum taught her with sign language.
She touched a lot of people in her short life. It has helped to receive messages from people who were helped by Immy’s videos and her work on deaf awareness and mental health, says Louise. “I’ve had some that said: ‘She basically saved my life.’”
Poisoned: Killer in the Post is on Channel 4 at 9pm on Wednesday 9 and Thursday 10 July
For more information on online safety for young people, visit the Thomas William Parfett Foundation and the Molly Rose Foundation
AI Insights
Redefining Tomorrow: How Chatronix is Shaping the Future of Artificial Intelligence
Introduction:
Artificial Intelligence (AI) is no longer a distant concept confined to the realms of science fiction. It has become a powerful force that is transforming industries, reshaping business operations, and enhancing everyday human experiences. Among the numerous players in the AI landscape, Chatronix stands out as a beacon of innovation and practical implementation. With its strong focus on usability, integration, and real-time applications, Chatronix is helping individuals and organizations navigate the evolving digital frontier with confidence. As we enter an era defined by intelligent systems and seamless automation, understanding what makes Chatronix different provides a glimpse into how AI is set to shape the world we live in.
How AI Is Becoming a Core Part of Daily Life
Artificial Intelligence has grown far beyond theoretical models and academic research. Today, AI is embedded into nearly every part of daily life, from voice assistants that respond to simple commands to intelligent algorithms that personalize online shopping experiences. It helps automate routine tasks, improve efficiency, and analyze massive datasets in seconds. As businesses seek new ways to optimize customer interactions and internal operations, the role of AI continues to expand. Chatronix plays a central role in this transformation by offering an adaptable platform that brings AI closer to real-world applications. Visit website – Chatronix, experiencing this future is easier than ever, whether you’re a business looking to implement smart solutions or an individual interested in cutting-edge technology. What makes Chatronix unique is its ability to simplify complex AI systems and present them in a user-centric manner.
The Evolution of AI Platforms and Chatronix’s Contribution
The AI industry has witnessed a dramatic evolution over the past decade. Initially limited to niche markets and research facilities, AI has now entered mainstream business tools and digital infrastructure. Early platforms struggled with accessibility and required specialized knowledge to operate, but modern systems, like those developed by Chatronix, have overcome these limitations. Chatronix’s platform focuses on integration, flexibility, and user-friendly interfaces that allow organizations to embed intelligent decision-making into their existing operations. Unlike many legacy systems that demand considerable customization, Chatronix offers plug-and-play features that can be tailored to fit a wide range of use cases. This shift from complex infrastructure to accessible platforms has opened up AI to smaller businesses and startups, creating a more inclusive technological landscape.
The Real-World Benefits of AI Deployment with Chatronix
One of the major concerns surrounding AI adoption is whether the technology delivers measurable, practical benefits. Chatronix addresses this concern by demonstrating tangible results across various sectors. In customer service, its AI models can understand and respond to inquiries with high accuracy, significantly reducing the workload on human support teams. In logistics, Chatronix solutions can analyze supply chain data in real-time, flagging inefficiencies and predicting disruptions before they occur. Healthcare providers are using Chatronix tools to enhance diagnostics, identify patterns in medical records, and improve patient outcomes. By focusing on real-world utility rather than theoretical capabilities, Chatronix has positioned itself as a reliable partner for organizations looking to improve productivity, minimize errors, and accelerate innovation.
Integrated Intelligence: The Power Behind Chatronix’s AI Suite
What sets Chatronix apart is its deep commitment to building integrated, intelligent systems that communicate effortlessly with one another. Instead of offering isolated tools, the company has developed a comprehensive framework where each component complements the others. Chatronix’s integrated AI suite enhances collaboration between different data processes, making it easier to draw conclusions and automate actions across platforms. This is especially valuable for businesses managing multiple departments or dealing with complex workflows. By breaking down silos and ensuring smooth data transfer, the integrated suite streamlines operations, improves decision-making accuracy, and reduces the time it takes to deploy new AI applications. As more organizations turn to AI for a competitive edge, Chatronix’s unified ecosystem allows them to scale without encountering the fragmentation issues that often accompany multi-vendor solutions.
Why User Experience Matters in AI Adoption
While the technical strength of an AI system is important, its success often depends on how easily it can be adopted by users with varying levels of expertise. Chatronix prioritizes user experience in every aspect of its design, from intuitive dashboards to guided workflows that reduce the learning curve. By making AI approachable, Chatronix enables more teams to participate in digital transformation projects, fostering a culture of innovation at every level of the organization. Whether it’s a marketing professional using predictive analytics or a data analyst running machine learning models, the platform ensures that each user gets the tools they need without being overwhelmed by complexity. This focus on usability not only accelerates adoption but also increases the return on investment by ensuring that features are fully utilized. Chatronix’s emphasis on accessibility reflects its broader mission of democratizing AI and making it a tool for everyone, not just specialists.
The Future of AI: How Chatronix Is Preparing for What’s Next
The rapid pace of AI development means that platforms must not only meet current needs but also anticipate future challenges. Chatronix is actively investing in research and innovation to stay ahead of emerging trends. From generative AI models to ethical governance frameworks, the company is working on solutions that balance power with responsibility. Security, transparency, and fairness are becoming central issues in AI deployment, and Chatronix is taking a proactive approach to ensure its technologies uphold these values. In addition, the platform is exploring advancements in areas such as edge computing, real-time AI collaboration, and multilingual support. As AI continues to influence everything from education to energy management, Chatronix is positioning itself as a future-ready platform capable of supporting both broad-scale innovation and niche applications. Its roadmap is aligned with the long-term interests of both businesses and consumers, making it a valuable partner in the ongoing digital transformation.
Conclusion: A Smarter, More Connected Future with Chatronix
Artificial Intelligence holds the promise of transforming how we work, communicate, and solve problems. But realizing that promise depends on having the right tools in place—tools that are accessible, scalable, and aligned with real-world needs. Chatronix exemplifies this ideal by offering a platform that balances powerful technology with practical application. Its integrated systems, user-focused design, and commitment to innovation ensure that businesses of all sizes can benefit from AI without the usual complexity. By prioritizing interoperability, performance, and ease of use, Chatronix is helping shape a future where intelligent systems work alongside humans to unlock new possibilities. As organizations look for ways to stay competitive in a data-driven world, Chatronix emerges as a leader, offering not just tools, but a vision for what AI can achieve when thoughtfully designed and widely accessible.
AI Insights
Process and Control Today | KHS optimizes its inspection technology with the help of artificial intelligence
08/07/2025 KHS GmbH
– Patent-pending algorithms improve fault detection
– New option for KHS’ Innocheck TSI closure inspector
– Can be used on new and existing machines
Tethered caps have been mandatory for all non-returnable PET bottles in the EU since 2024. Consequently, the number of closure variants and thus the demands made of inspection technology have increased. KHS GmbH is meeting these new challenges with the help of artificial intelligence (AI). The systems provider has now equipped its proven Innocheck TSI closure inspection unit with an AI-based fault detector.
The obligatory introduction of tethered caps had far-reaching consequences for bottle and beverage producers in the EU. They were forced to find new solutions to continue to meet the high demands made of product quality and protection. This also affected inspection technology, explains Nikita Wall from Labeling and Inspection Technology Product Support at KHS. “Tethered caps were rarely used before the EU directive came into force. Conventional systems thus frequently reach the limits of their capacity where cap inspection is concerned.” What’s more, the large number of bottle and cap design variants processed and increasing line capacities present additional challenges.
Optimized cap inspection thanks to AI
With its AI-based fault detection system for tethered caps, the Dortmund turnkey supplier now provides smart engineering that identifies potential defects during cap inspection. The new module makes use of patent-pending algorithms to analyze images of bottle closures in real time. Cameras log the caps in high-resolution quality, while AI models evaluate this data. By applying deep learning, the models adjust to account for any new types of flaw found.
KHS’ AI-assisted system increases both the accuracy and efficiency of inspection. “Our customers make extremely high demands of quality assurance. In intensive field tests, our AI-based fault detector has proved that it fully meets these requirements,” emphasizes Wall.
Foundation for further projects
KHS has developed the new system specifically for its tried-and-tested Innocheck TSI cap inspector. It can be implemented on both new and existing machines.
In the future, KHS also wants to use AI for further inspection technology equipment. Says Wall, “In the development team, we’re currently discussing which systems this would make sense for. AI solutions are only practical where the requirements are complex – and if they give our customers clear benefits.”
Process and Control Today are not responsible for the content of submitted or externally produced articles and images. Click here to email us about any errors or omissions contained within this article.
AI Insights
Albo murky on Beijing’s AI bid
Anthony Albanese has kept tight-lipped on China’s offer to co-operate on artificial intelligence amid tensions between the government and the Trump administration.
Chinese ambassador Xiao Qian wrote an op-ed this week saying Beijing was keen to take a “more open attitude” to the free-trade agreement with Australia.
Mr Xiao said that included bolstering “co-operation in traditional areas such as agriculture and mining” but also exploring “new growth areas in emerging fields like artificial intelligence” – a convenient offer against a backdrop of fierce AI competition with the US.
Though, even as lacklustre Australian defence spending and US tariffs strain Canberra’s alliance with Washington, the Prime Minister was hesitant to embrace or reject the offer on Tuesday.
“We will determine our policy,” Mr Albanese told reporters in Hobart.
However, he was quick to spruik the benefits of free trade with China.
“Of course, we have a free-trade agreement that was done by the Coalition government when they were in office,” Mr Albanese said.
“What we have done is to get rid of the more than $20bn of impediments of goods that were stopped from going to China.
“It’s made an enormous difference and we have, indeed, products like wine and barley have not just bounced back – they’re back higher than they were before.”
He also spruiked his upcoming trip to China, calling it an “important trading partner for Australia”.
“Twenty-five per cent of our exports go to China,” Mr Albanese said.
“What that means is jobs and one of the things that my government prioritises is jobs.”
Pressed on whether he was open to having AI covered by the China-Australia Free Trade Agreement, he dodged it, saying his “priority is jobs”.
China is in an AI race with the US.
The launch of DeepSeek in January dealt the first serious blow to the US’ global leadership in the space, with the ensuing tech sell-off wiping about $US1 trillion ($A1.5 trillion) in value off American titan Nvidia.
The Trump administration earlier this year announced a $US500bn investment in AI, eclipsing all other countries as the clear frontrunner.
About the same time, China launched a AI investment fund with an initial $8.2bn.
But with DeepSeek free and offering capabilities comparable to American counterpart OpenAI, the AI race may not be won by those who splash the most cash.
-
Funding & Business1 week ago
Kayak and Expedia race to build AI travel agents that turn social posts into itineraries
-
Jobs & Careers1 week ago
Mumbai-based Perplexity Alternative Has 60k+ Users Without Funding
-
Mergers & Acquisitions7 days ago
Donald Trump suggests US government review subsidies to Elon Musk’s companies
-
Funding & Business7 days ago
Rethinking Venture Capital’s Talent Pipeline
-
Jobs & Careers7 days ago
Why Agentic AI Isn’t Pure Hype (And What Skeptics Aren’t Seeing Yet)
-
Funding & Business4 days ago
Sakana AI’s TreeQuest: Deploy multi-model teams that outperform individual LLMs by 30%
-
Jobs & Careers7 days ago
Astrophel Aerospace Raises ₹6.84 Crore to Build Reusable Launch Vehicle
-
Jobs & Careers7 days ago
Telangana Launches TGDeX—India’s First State‑Led AI Public Infrastructure
-
Funding & Business1 week ago
From chatbots to collaborators: How AI agents are reshaping enterprise work
-
Jobs & Careers5 days ago
Ilya Sutskever Takes Over as CEO of Safe Superintelligence After Daniel Gross’s Exit