Elon Musk's AI chatbot Grok is posting antisemitic tropes on X
Grok echoed antisemitic tropes, claiming individuals with Jewish-sounding names are disproportionately linked to hate and deceitfulness, and was prompted by users to utter Nazi slogans.
The Anti-Defamation League condemned Grok's responses as 'irresponsible, dangerous and antisemitic,' warning that such rhetoric would amplify extremism on X.
The chatbot indicated its July 5 update, which included directives not to shy away from 'politically incorrect' claims, enabled its new behavior.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Daily Mail
13 minutes ago
- Daily Mail
Shocking secret of tearful Bryan Kohberger prosecutor, Bill Thompson, whose death penalty-sparing plea deal sparked widespread fury
The prosecutor who handed Bryan Kohberger a controversial plea deal last week has never secured a death sentence for any of the killers he's convicted during his 33 years in office. Latah County Prosecuting Attorney Bill Thompson cried during the July 2 hearing that saved Kohberger's life, and put him in prison for life without appeals for the 2022 murders of four University of Idaho students. The Democrat prosecutor faced massive backlash over the deal, including from some of the victims' relatives, who said they felt cheated out of their day in court and the opportunity to see the killer face the details of his crime. He had initially indicated to the terrified Moscow community that he would seek the death penalty against Kohberger, a former Washington State University graduate student of criminology. But the plea deal is not a shock to those familiar with the 68-year-old's career - Thompson has a history of striking a deal rather than seek capital punishment for the killers he's prosecuted, as highlighted in a new report by the Idaho Statesman. The last mass murder caser Thompson prosecuted, eight years before the Kohberger case, had a similar conclusion. Thompson brokered a deal that saw John Lee, 29, admit to killing three people in 2015, including his adoptive mother, in exchange for removing the possibility of the death penalty. Similarly, in 2014, Thompson struck a deal with Silas Parks, 25, who agreed to plead guilty to voluntary manslaughter after he strangled his pregnant wife to death. In that case, Thomas did not seek the death penalty because there was not enough evidence that the killings were premeditated. Instead, Parks was sentenced to a maximum of 40 years in prison. The only death sentence obtained by Thompson was in 2000, after Dale Shackelford, 37, killed his ex-girlfriend and her boyfriend and burned their bodies. However, the death sentence was later revoked by the Supreme Court and Thompson did not push to have Shackelford returned to death row. Kohberger's plea deal will be the last Thompson brokers - the veteran prosecutor, who plays guitar in a local band in his free time, is expected to retire following the case's conclusion. He was first elected as Latah County's prosecutor in 1992, which makes him the longest-serving prosecutor in Idaho, after eight re-elections. Moscow Mayor Art Bettge told the Idaho Statesman he believes Thompson only ran for office the last time so he could see the Kohberger case through, rather than inflict it on a rookie replacement. Last week, Thompson choked up as he laid out the timeline of the shocking murders of Ethan Chapin, 20, Kaylee Goncalves, 21, Xana Kernodle, 20, and Madison Mogen, 21. 'On November 13, 2022, excuse me...,' Thompson said, growing emotional as he reached for a sip of water and a colleague patted him on the back. Kohberger stared coldly ahead as Thompson laid out the timeline, saying that he stalked the students' home at 1122 King Road in Moscow, Idaho before the slayings. Thompson said the state would have argued in court that Kohberger slipped in through a side door before killing Mogen and Goncalves on the third floor. He said Kohberger then encountered Kernodle as she came down the stairs to pick up a Doordash order, and killed her with the same Ka-Bar knife he used on her roommates. He then entered Kernodle's bedroom and stabbed her sleeping boyfriend Chapin to death, Thompson said. As Kohberger's victims' loved ones sobbed in the courtroom, the killer showed no emotion and stared ahead. He repeatedly said the word 'yes' when asked if he'd committed the horrific killings, understood the terms of the plea deal and that it meant he could never apply for parole. Madison Mogen's father Ben Mogen hung his head and his legs shook as the judge asked Kohberger if he admitted to stabbing his daughter to death. Kaylee Goncalves' mother Kristi Goncalves, who was flanked by several family members, cried as the judge asked Kohberger if he had murdered her daughter and Kohberger replied in the affirmative. The Goncalves family previously vented their fury at Thompson after he offered Kohberger the plea deal that spared him the death penalty. They declared on Facebook that the State of Idaho 'failed us' as they hit out at prosecutors for failing to notify them that a plea deal was going through. 'We weren't even called about the plea; we received an email with a letter attached,' family members said in a statement. 'That's how Latah County's Prosecutor's Office treats murder victims' families. 'Adding insult to injury, they're rushing the plea, giving families just one day to coordinate and appear at the courthouse for a plea on July 2.' The family of Chapin — one of three triplets who attended the university together — supports the deal. Mogen's father, Ben Mogen, told CBS News he was relieved by the agreement. 'We can actually put this behind us and not have these future dates and future things that we don't want to have to be at, that we shouldn't have to be at, that have to do with this terrible person,' he said. 'We get to just think about the rest of lives and have to try and figure out how to do it without Maddie and the rest of the kids.'
The Guardian
13 minutes ago
- The Guardian
US neo-fascist group claims it is part of Texas floods relief efforts
A US racist and neo-fascist hate group that has become a public fixture in recent years has descended on central Texas in a stunt it claims is part of the 'disaster relief' efforts under way after the devastating flash floods hit the region last week. Patriot Front, founded following the deadly 2017 Unite the Right rally in Charlottesville, Virginia, where its leader, Thomas Rousseau, a Texan, was later charged for his participation, has claimed on its Telegram app channel that it has shown up in the areas near Camp Mystic, where 27 young campers lost their lives. 'Patriot Front is here in central Texas,' Rousseau declares in a video statement, amid the backdrop of what sounds like buzzing chainsaws and a flood ravaged community, 'responding to the flooding, which has destroyed communities and taken the lives of scores of Americans'. Rousseau goes on to claim that his so-called 'activists' are distributing supplies to survivors, but clarifies that his group is prioritizing their 'people' and 'European peoples' in those operations. The far-right compulsion to disguise racist actions under the terms of humanitarianism, has its roots in Adolf Hitler's autobiography, Mein Kampf, where he prescribes to his readers to take pride in keeping their collective communities strong as a not-so-hidden metaphor for cleansing it of what they deemed other, undesirable peoples. Similarly, American hate groups, second amendment-oriented militias, and neo-Nazi street gangs have long shown up as community relief cadres across the US, in a sort of ploy to whitewash their images as dangerous forces in the country. After hurricanes struck the Carolinas and Florida in fall 2024, Patriot Front also exploited the moment during a particularly polarizing presidential campaign season, showing up to clear debris in badly hit communities. 'It's not surprising to see Patriot Front inserting itself into disaster relief in Texas,' said Heidi Beirich, co-founder of the Global Project Against Hate and Extremism, who has researched rightwing extremism in the US for decades. 'The group was founded there, and like other extremist groups, they want to take advantage of relief efforts to mainstream their ideas, present themselves as non-threatening and helping the community, and ultimately use what they hope will be positive PR to recruit and grow,' Beirich noted that everyone from Klansmen to armed militias have seen value in public appearances providing disaster aid. Sign up to Headlines US Get the most important US headlines and highlights emailed direct to you every morning after newsletter promotion 'We've seen David Duke do this in the past, and more recently Veterans on Patrol inserted themselves into the relief efforts in Asheville after last year's hurricane,' she said. 'The concern is that it works – and Patriot Front's white supremacist agenda gets laundered as positive, and that helps them spread hate and recruit.' Last week, Patriot Front caused a stir in Louisville, Kentucky, over the holiday weekend when its masked-members marched to the beat of drums in downtown streets holding a Confederate banner, which has become a commonplace demonstration for the group having previously appeared in Boston around the fourth of July weekend in 2022. Thomas Rousseau has sometimes tried to soften his image in public statements, referring to himself and his crew, merely as political 'activists'. But recent Guardian reporting shows the group has increasingly allied itself with the neo-Nazi Active Club movement and the white supremacist leader Robert Rundo.
Geeky Gadgets
15 minutes ago
- Geeky Gadgets
LangChain Expression Language : Discover the Power of LCEL
What if the way we build and manage workflows could be transformed into something more intuitive, adaptable, and efficient? Enter the LangChain Expression Language (LCEL)—a new framework that redefines how developers construct chains in LangChain. Gone are the days of wrestling with rigid components and verbose code. With LCEL, the process becomes as seamless as connecting puzzle pieces, thanks to its streamlined syntax and innovative features like the pipe operator. Imagine being able to design complex workflows with clarity and precision, all while reducing the time and effort traditionally required. LCEL isn't just an upgrade; it's a paradigm shift for anyone navigating the challenges of modern chain-building. James Briggs explores how LCEL's modular runnables, parallel processing, and simplified design empower developers to tackle even the most intricate workflows with ease. You'll uncover how its unique capabilities—like processing multiple data streams simultaneously or customizing workflows without external code—make it a fantastic option for efficiency and scalability. Whether you're a seasoned developer or new to LangChain, LCEL offers tools that promise to optimize your processes and spark creative possibilities. As you journey through its features, consider how this approach might reshape not only how you build chains but also how you think about solving complex problems. Overview of LCEL Features The Challenges of Traditional Chain-Building Traditional chain-building in LangChain relied on predefined components such as prompt templates, language models (LMs), and output parsers. While functional, this approach often lacked flexibility and required developers to write additional custom code to handle modifications or integrate multiple data sources. These limitations made it difficult to adapt workflows to evolving requirements and increased the time and effort needed for development. Furthermore, the deprecation of older methods underscored the need for a more modern and flexible solution that could streamline these processes. LCEL: A Simplified and Intuitive Approach LCEL introduces a innovative approach to chain-building by using an intuitive syntax centered around the pipe operator (`|`). This operator enables seamless connections between components, allowing the output of one component to flow directly into the input of the next. By eliminating verbose and complex code, the pipe operator enhances both the readability and maintainability of workflows. Behind the scenes, the pipe operator uses the `or` method, making sure smooth integration between components. This design not only simplifies the development process but also reduces the likelihood of errors, making it easier for developers to focus on creating efficient and scalable workflows. LangChain Expression Language (LCEL) Explained Watch this video on YouTube. Explore further guides and articles from our vast library that you may find relevant to your interests in LangChain. Runnables: Modular Building Blocks for Workflow Design At the core of LCEL are runnables, which are modular components designed to process data step-by-step. These building blocks allow you to create workflows tailored to specific tasks by chaining them together. For instance, you can preprocess text, generate outputs using a language model, and format the results for presentation—all within a single, cohesive chain. Key features of runnables include: Runnable Lambda: This feature enables you to define custom runnables directly within the framework, eliminating the need for external classes and simplifying the development process. This feature enables you to define custom runnables directly within the framework, eliminating the need for external classes and simplifying the development process. Runnable Pass Through: This component allows variables to pass through the chain unchanged, providing flexibility when handling intermediate data or maintaining specific inputs. By combining these features, runnables empower developers to design workflows that are both highly customizable and easy to maintain. Parallel Processing: Boosting Efficiency and Scalability LCEL's parallel processing capabilities represent a major leap forward in efficiency. The Runnable Parallel component enables multiple processes to execute simultaneously, allowing you to combine outputs from various data sources in real time. For example, you can retrieve context from two separate datasets and merge the results to answer a complex query. This feature is particularly valuable for applications that involve large-scale data operations or require time-sensitive processing, such as generating insights from multiple data streams or handling high-volume requests. By allowing concurrent processing, LCEL reduces processing time and ensures that workflows remain efficient, even as complexity increases. Real-World Applications of LCEL LCEL's versatility makes it an ideal solution for a wide range of use cases. Here are some practical examples of how LCEL can be applied: Report Generation: Use LCEL to chain components that generate reports, replace specific terms, and remove unnecessary sections, all within a single workflow. Use LCEL to chain components that generate reports, replace specific terms, and remove unnecessary sections, all within a single workflow. Data Integration: Combine outputs from multiple sources to provide comprehensive answers to complex questions, making sure accuracy and depth in the results. Combine outputs from multiple sources to provide comprehensive answers to complex questions, making sure accuracy and depth in the results. Handling Complex Operations: Use LCEL's support for dictionaries to manage multiple function arguments effortlessly, simplifying the execution of intricate workflows. These examples demonstrate LCEL's ability to streamline operations across diverse domains, from automating repetitive tasks to integrating complex data sources. Why LCEL Stands Out LCEL offers several distinct advantages over traditional chain-building methods, making it a preferred choice for developers seeking efficiency and flexibility: Simplified Syntax: The pipe operator and modular design make chain-building more intuitive, reducing the learning curve for new users. The pipe operator and modular design make chain-building more intuitive, reducing the learning curve for new users. Enhanced Flexibility: Runnables and parallel processing provide the tools needed to create highly customized and scalable workflows. Runnables and parallel processing provide the tools needed to create highly customized and scalable workflows. Improved Efficiency: By allowing concurrent processing and seamless integration, LCEL minimizes development time and reduces processing overhead. These benefits position LCEL as a powerful tool for developers looking to optimize their workflows and achieve better outcomes in less time. The Future of Chain-Building with LCEL The LangChain Expression Language (LCEL) redefines the landscape of chain-building by offering a more intuitive, flexible, and efficient framework. With features like the pipe operator, modular runnables, and parallel processing, LCEL enables developers to create scalable workflows tailored to their specific needs. Whether you're generating reports, integrating data from multiple sources, or handling complex operations, LCEL provides the tools necessary to streamline processes and deliver high-quality results. As the demands of modern applications continue to evolve, LCEL stands ready to meet these challenges, offering a robust and adaptable solution for developers across industries. Media Credit: James Briggs Filed Under: AI, Guides Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
