Latest news with #ChatGTP
Yahoo
4 days ago
- General
- Yahoo
It was ChatGPT, says Italian who ‘hoped Meloni's daughter would meet same fate as murder victim'
An Italian teacher who called for Giorgia Meloni's eight-year-old daughter to be murdered has blamed ChatGPT for his online post. Stefano Addeo, 65, who teaches German in a school near Naples, caused fury at the weekend when he wrote on social media that he hoped Ginevra, the Italian prime minister's daughter, would 'meet the same fate' as a 14-year-old girl who was recently allegedly murdered by her jilted boyfriend. The post was condemned across the political spectrum, with opposition MPs as well as members of the governing coalition saying it was unacceptable. He is now under investigation by the authorities for inciting hatred. Mr Addeo attempted on Monday to justify the message, saying that it had been created by the artificial intelligence chatbot, albeit at his request. He told the La Repubblica newspaper: 'It wasn't me, it was artificial intelligence. I saw on television on Friday that the government was not distancing itself from the war in Israel. I asked ChatGTP to create a message critical of Meloni. This crazy thing emerged and I posted it.' Asked why he had not taken a moment to reflect on the gravity of the post, he said: 'I just wanted to post something. ChatGTP is more harmful than you think. I'm a Catholic, I'm in favour of animal rights, I'm a vegetarian. 'I'm on the Left but that does not excuse this type of message. I've had to cancel all my social media platforms, the consequences have been really ugly. I've had to shut myself away at home. 'People are throwing tomatoes at my front door. I have been depicted as a monster. I'm really shocked – people are issuing threats against me. I did a really stupid thing.' He said furore over the incident meant he had been harangued in the streets of his home town. The teacher, who has no children and lives with his 90-year-old mother, said that if he had the opportunity to meet the prime minister, he would apologise. 'I would hug her, asking her forgiveness. I wish only the best for her daughter, although I would advise her to be careful about social media,' he said. But it has emerged he has also posted messages threatening the children of other members of the Right-wing coalition – Matteo Salvini, the transport minister and head of the League party, and Antonio Tajani, the foreign minister and a former European commissioner. 'This is intolerable – you can attack me, but not my daughter,' said Mr Salvini. The teacher's social media post was a reference to the horrific case of Martina Carbonaro, a 14-year-old schoolgirl who was beaten to death with a rock a few days ago, allegedly by her 18-year-old ex-boyfriend. He was allegedly furious that she had ended their relationship. After attacking her, he attempted to hide her body beneath rubbish and debris in an abandoned farmhouse near the town of Afragola, north of Naples, it was reported. Ms Meloni, who is halfway into her five-year term as premier, has decried what she called a 'sick climate' in Italian political discourse. She said that threats made against her and her young daughter represent 'a spiral of hatred fed by an ideological fanaticism that has exceeded every limit. No political differences or ideological cause can ever justify attacks on children'. Matteo Renzi, the leader of a centre-Left party and a former prime minister, said the teacher should lose his job. Broaden your horizons with award-winning British journalism. Try The Telegraph free for 1 month with unlimited access to our award-winning website, exclusive app, money-saving offers and more.
Yahoo
21-05-2025
- Yahoo
Federal judge mulls sanctions for attorneys who used AI in court filing
The front of Hugo L Black Courthouse in Birmingham, Alabama on August 15, 2023. A U.S. District Judge is considering sanctions against attorneys representing the state for using artificial intelligence to draft a legal filing that either misstated legal authorities or cited cases that do not exist. (Jemma Stephenson/Alabama Reflector) BIRMINGHAM — A federal judge said Wednesday she would consider sanctions on attorneys who filed a motion that used artificial intelligence and cited legal authorities that do not exist. U.S. District Court Judge Anna M. Manasco told attorneys representing the state in the lawsuit — claiming corrections officers failed to protect an inmate — that lawyers continue to use artificial intelligence even after other courts have imposed corrective measures throughout the country. 'Generally, this has occurred in other cases where the courts have imposed sanctions and standing orders,' Manasco said during the hearing. 'This incident is proof-positive that those sanctions were insufficient. That causes me to consider a fuller range of sanctions.' SUBSCRIBE: GET THE MORNING HEADLINES DELIVERED TO YOUR INBOX Attorneys Matthew Reeves, the attorney who used AI; Bill Lunsford, the attorney overseeing the case, and litigants from Montgomery-based Butler Snow, the firm where Reeves and Lunsford are employed, expressed remorse and publicly apologized to both the plaintiffs and the court regarding the incident. 'I had limited use of artificial intelligence products such as Westlaw Precision with CoCounsel and ChatGTP,' Reeves said to Manasco during the hearing Wednesday. 'I first used it related to dietary matters. That is how I started to use it.' Reeves then said he began querying artificial intelligence to research colleges for his son before applying the technology professionally for research into policies and practices related to different issues. It then culminated in using it to obtain citations for the case. He said he was aware that using artificial intelligence to obtain citations did not comply with the firm's policies and that he did not verify the citations during the two instances that he relied on AI for the citations. Reeves, who according to his LinkedIn profile has been a partner with Butler Snow since April 2023, said Manasco is within her discretion to impose sanctions. 'My only request, since I am the only one responsible, is not to punish my colleagues,' he said. The incident involves five citations that Manasco characterized as 'hallucinated' across two documents filed with the court. The court was made aware of the fabricated citations after the attorneys for Frankie Johnson, the plaintiff in the case, filed a document to oppose the state's request to have their client available to give testimony on a certain date. 'As discussed below, Defendant has cited no legitimate authority to support his proposition that ordering a deposition before an anticipated document production is appropriate,' the filing said. 'Instead, Defendant appears to have wholly invented case citations in his Motion for Leave, possibly through the use of generative artificial intelligence.' Lunsford and Butler Snow have been awarded millions of dollars to represent the Alabama Department of Corrections in cases alleging abuse within Alabama's prison facilities. In December 2024, the Contract Review Committee approved contracts totaling $4.8 million to Lunsford and Butler Snow. The committee had previously approved $14.9 million in July 2023 and another $7.68 million in June of that same year. Blood Money: Alabama Department of Corrections pays to settle lawsuits alleging excessive force Manasco listed several possible consequences, from continuing education seminars and fines to referrals to the Alabama State Bar or even temporary suspensions. Johnson, currently incarcerated at William E. Donaldson Correctional Facility in Bessemer, said in the lawsuit that he was attacked by as many as 13 people and stabbed at least nine times in the prison in December 2019. Johnson said that despite the attack, staff at the prison failed to properly protect him afterward. 'As a consequence of their individual and collective failure, on or about March 5, 2020, while restrained in shackles and handcuffs during a therapy session, Mr. Johnson was stabbed repeatedly by another prisoner at Donaldson,' the lawsuit states. The attorneys for ADOC filed a motion to compel Johnson to testify the week of June 3. Johnson's attorneys said that they would not have enough time to prepare their client to give testimony. In response, Reeves and Lunsford filed another document stating that other courts have required people who are incarcerated to give their testimony during a deposition when they are given the appropriate notice, which they argue was done in this case. 'Defendant bolstered this assertion with a lengthy string citation of legal authority and parentheticals that appeared to support Defendant's proposition,' Johnson's attorneys said. 'But the entire string citation appears to have been made up out of whole cloth.' Johnson's attorneys noted four of the citations in the ADOC filing. One included an appropriate case but was not relevant to the issue that Reeves and Lunsford claimed in the document they filed. A second case, Kelley v. City of Birmingham, was said to be decided in the U.S. Northern District of Alabama in 2021, when it was actually a Alabama state case from 1939. Two other citations referred to cases that did not exist. Attorneys for Johnson began to review other documents that Lunsford and Reeves filed with the court in the case. They then found another citation that was fabricated in a different document in which 'Plaintiff's counsel found a string of similarly named opinions, none of them stood for the proposition Defendant represented, and Plaintiff's counsel could not identify any case using the citation Defendant provided,' Johnson's attorneys said in their filing. The plaintiffs in the case agreed to allow Reeves to refile the document to have Johnson testify by the original date that he and Lunsford requested, but this time with the appropriate citations. Manasco will allow the leadership at the Butler Snow law firm another 10 days to review its policies and practices pertaining to the use of artificial intelligence before considering the matter further. SUPPORT: YOU MAKE OUR WORK POSSIBLE
Forbes
27-03-2025
- Forbes
Your AI Mindset Translator: The Future Of Understanding Is Here?
In an era of hyper-connectivity and stark ideological divides, bridging the chasm of differing viewpoints is more precious than ever. When dystopian politics make the grasp of reality a shifting target, while news cycles amplify discord, and social media reinforce already harrowing echo chambers, our daily interactions can leave us bewildered by the seemingly foreign logic of others. We grapple with the frustration of miscommunication, where even shared language fails to bridge the gap in understanding. It's as if we're speaking different dialects of the same tongue, each rooted in distinct, impenetrable mindsets. Could an AI mindset translator help? While Google Translate has long facilitated linguistic cross-pollination, a more subtle translation is needed now – one that transcends words and dives into the fabric of thoughts, feelings and aspirations. Could artificial intelligence, our ubiquitous chatbot companions, evolve into sophisticated perspective interpreters, unraveling the complexities of diverse mindsets? At the heart of human connection lies empathy, the ability to walk in another's shoes. Can a machine, devoid of lived experience, truly emulate this nuanced capacity? Surprisingly, research suggests AI's immense potential in this arena. Consistently AI-powered bots are perceived as more empathetic, understanding, and friendly than their human counterparts, particularly in customer service and therapeutic contexts. Especially in mental health support, chatbots were seen as more understanding and less judgmental than human therapists. To be clear, this isn't about genuine feeling but the AI's ability to analyze and respond with patterns that mimic empathetic communication, such as acknowledging emotions and validating experiences – yet the end result, the customer experience, seems to validate that appearances sometimes prime reality. Communication extends beyond mere words. Non-verbal cues convey meaning, like tone of voice, facial expressions, and body language. A study by Mehrabian (1971) famously suggested that 7% of communication is verbal, 38% is vocal, and 55% is facial. While the precise percentages are debated, the principle remains: non-verbal cues are vital. While current AI models lack the full spectrum of human sensory perception, they rapidly advance in emotion and intent detection. Multimodal models, integrating video, audio, and text, are becoming adept at discerning subtle nuances in tonality and facial expressions, offering a more holistic understanding of human interaction. They can detect micro-expressions and slight changes in vocal intonation, indicating underlying emotions like stress or deception. Beyond the ever-advancing multimodal capacities of the big frontier models, such as ChatGTP and Gemini, companies like Smart Eye specialize in emotion AI. This progress opens a fascinating possibility: AI as a tool to navigate the intricate landscape of human thought and feelings. Imagine leveraging this technology to bridge cultural divides, comprehend opposing political viewpoints, or gain insight into different socio-economic backgrounds. Could AI become a mediator of not just words but worldviews? To effectively curate AI as a mindset interpreter, we can embrace four principles: 1. Awareness: Recognizing the inherent biases within ourselves and the AI systems we employ is the point of departure. AI models are trained on data, and that data reflects the biases present in society. For example, language models like GPT-3 have been shown to exhibit gender and racial biases, reflecting the skewed representation in their training data. We must be conscious of these limitations and actively seek to mitigate them while being candid about our ingrained personal biases. 2. Appreciation: Cultivating an appreciation for the diversity of human thought needs to happen offline, tech-less. AI can help us identify patterns and perspectives that remain hidden. By analyzing large datasets of text and speech, AI can reveal the underlying values and beliefs that shape different mindsets. For instance, AI-driven sentiment analysis can identify the emotional tone of political discourse, exposing different groups' underlying anxieties and aspirations. Furthermore, AI can create interactive visualizations that illustrate the distribution of opinions on complex issues, enabling a more nuanced understanding of public sentiment. Still, this information will only make a difference if our minds are open to acknowledging what AI is surfacing. "None so blind as those who will not see" (John 9:39-41). 3. Acceptance: Acceptance does not equate to agreement. It signifies a willingness to engage with perspectives that differ from our own without dismissing them outright. AI can facilitate this by providing neutral, objective summaries of complex arguments, allowing us to see the logic behind opposing viewpoints. This is especially useful in political discourse, where AI tools can identify common ground and facilitate constructive dialogue. For instance, AI-powered systems could be used to analyze political speeches and identify areas of potential agreement between opposing candidates, fostering more productive debate. Platforms that use AI to summarize and systematically present diverse arguments on hot-button topics may build a foundation for gradual mutual acceptance. 4. Accountability: We must hold ourselves and AI systems accountable for the information they produce and the interpretations they offer. AI-generated insights can serve mutual understanding or deepen the divide. This requires human intent for transparent and verifiable outputs. We must be mindful of the potential for AI to be used to manipulate or misrepresent information. By prioritizing ethical considerations and promoting responsible AI development, we can harness its potential for good. This means fostering transparency in how AI models make decisions, ensuring that algorithms are auditable, and establishing clear guidelines for the ethical use of AI in communication. The future of AI and NI – natural intelligence, isn't a pre-written script but an organically evolving dynamic. As our natural and artificial 'minds' are increasingly entangled, the big question is whether AI becomes a 'mirror neuron' for humanity and what we will see in that mirror. Will it reflect back our shared empathy, or are we going to face a 'hall of mirrors,' a distorted reality of personalized echo chambers that are misted by prejudice? The risk of weaponized, hyper-personalized propaganda is not a distant threat but a phantom limb, already twitching. Yet, within this uncertainty lies a promise. Beyond its use as a mindset translator, AI could be configured as a mindset cartographer, to chart the sprawling territories of human understanding. Imagine: AI not just translating words but deciphering the very music of thought. The choice is ours: to forge a future of resonant understanding or to surrender to the cacophony of fractured realities.
Forbes
25-03-2025
- Forbes
Is AI Quietly Rewiring How You Write, Or How You Think?
The use of AI is changing how we write, the question is to which extent it influences how we think ... More about the content. A shift in the transmission of ideas and knowledge is underway – and it matter to preserve the ability for mental focus. Not everything can, or should be, delegated. The very essence of writing is undergoing a revolution. Over the past two years, Artificial intelligence has begun to reshape how we compose everything from casual emails to complex reports. This technological transformation goes beyond mere mechanics; it influences our thinking in a subtle but far-reaching way. Are we witnessing an evolution in human cognition or an erosion? At first glance, the benefits are undeniable. AI writing tools, such as ChatGTP, Claude, or Gemini offer unparalleled efficiency, generating text in seconds, summarizing lengthy documents, translating languages and crafting entire reports from scratch – increasingly with impressive accuracy. This newfound ease in the production and delivery of polished materials promises to liberate us from the drudgery of drafting, allowing us to focus on higher-level conceptualization and strategic thinking. However, this has consequences beyond the output stage. One of the most intriguing shifts is the increasing emphasis on the art of questioning. As AI-powered chatbots become more sophisticated, their ability to provide answers hinges entirely on the clarity and precision of our queries. Vague or poorly formulated questions yield superficial or nonsensical results, forcing us to adopt a more rigorous approach to intellectual exploration. Potentially, this dynamic will lead us to cultivate greater internal clarity. In the pre-AI era, we often muddled through complex topics, relying on intuition and tacit understanding. Now, the need to articulate our thoughts with precision compels us to deconstruct complex ideas, identify underlying assumptions, and formulate straightforward, concise questions. In essence, AI is training us to think more like AI. We learn to think more about the question we seek to address before writing the related answers. This is positive. Another consequence of our new approach to text production is less promising. Instead of meticulously crafting arguments, exploring nuances, and forging original connections, we have begun to rely on AI to do the heavy lifting. Beyond temporary convenience, this can lead to a decline in our ability to sustain deep thought, articulate complex ideas, and engage in the type of intellectual exploration that has historically driven innovation and progress. The regular use of AI can lead to agency decay, marked by the gradual reliance on AI for even the most minor intellectual tasks. Using it regularly for writing leads to "patchwork thinking" – a new type of cognitive fragmentation where the human writer relinquishes the hard work of coherent thought and fluent penmanship, leaving it to AI to stitch together disparate elements. The implications of this shift are disconcerting. As we delegate more of the writing process to AI, are we inadvertently outsourcing our capacity for critical thinking? Research on cognitive offloading suggests that relying on external tools reduces intellectual effort, potentially leading to a decline in our abilities over time. Furthermore, studies on the impact of technology on attention spans indicate that constant multitasking and reliance on digital tools can fragment our thinking, making it harder to focus and sustain coherent thought. The brain is a muscle – do we use it or consent to lose it? AI promises to streamline our workflows and boost the final delivery's quantity and quality. And all of this comes with the bonus of freeing up time we previously spent on tasks we disliked. It seems like an all-around bonus. But there is a catch. In writing, as in sports, the how matters as much as the what. Differently put, the process itself is valuable, not just the final product. Wrestling with words, grappling with complex ideas, and meticulously weaving arguments are precious to develop and refine our thinking. By relinquishing this struggle to AI, we risk losing something essential: the cognitive calisthenics that strengthen our minds, deepen our understanding, and foster our capacity for original thought. We may gain in output but lose regarding the outcome – the gradual refining of our ideas, linguistic elasticity and verbal sophistication. It is also worth remembering that although AI is impressive, even the most advanced models on the market in March 2025 are prone to hallucination. When it comes to creativity, they remain inferior to human experts. It is a trade-off: comfort versus excellence, professional convenience versus personal growth. Unfortunately, the things that matter the most to us come at the price of discomfort; a shortcut might work temporarily, but we lose out in the long run. So, how do we navigate the complex new landscape of our mind amid AI? How do we harness the power of our expanding artificial assets without sacrificing our intellectual independence? There may be an alternative to admire, savor and preserve the cake, even if it comes with the intricacies of self-reflection. The A-Frame offers a simple framework to explore a practical path forward: Awareness: Cultivate a critical understanding of how AI influences your thinking and writing processes. Recognize when you rely too heavily on AI, and identify the aspects of your thinking that may be affected. This involves consciously monitoring your cognitive habits and how they change as you interact with AI tools. Appreciation: Value the positive contributions of AI, and your own. Appreciate its ability to enhance efficiency, streamline workflows, and provide access to vast amounts of information; and recognize the unique quirky components that make your thinking and writing special. AI can is a tool to boost your intellect, not to replace. Acceptance: AI is here to stay and will continue to evolve, fast. Resist the urge to either blindly embrace or reject it. Instead, focus on developing a balanced and nuanced understanding of its potential benefits and risks. Accountability: Take responsibility for the outcomes of your thinking and writing. Do not abdicate your intellectual agency to AI. Use AI to enhance your abilities, but always retain ownership of your thoughts, ideas, and conclusions. AI is undoubtedly changing how we write. The question remains: is it changing how we think? The answer is a complex and nuanced one. AI offers the potential to enhance our cognitive abilities, but at the same time, it jeopardizes our intellectual independence. The key lies in conscious cognition. We can deliberately develop a mindful relationship with our artificial assets. By cultivating awareness, appreciation, acceptance, and accountability, we can harness this transformative technology's power while safeguarding the essence of what makes us human: our capacity for independent, critical, and creative thought.
