home.social

#hallucinate — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #hallucinate, aggregated by home.social.

fetched live
  1. ⬆️ @saxnot

    >> we learned about #GIGO in both statistics and computer science.

    We never had to invent words like #hallucinate to evade accountability for problems in LLM and AI software itself. So AI hallucination is GIGO from the other end, not from bad user input.

    I chose translation software as an example. They do an admirable job in figuring out parts of speech, but not so well in figuring out figures of speech, eg Godfather makes an "offer you can't refuse."

    @avandeursen @TeflonTrout

  2. ⬆️ @saxnot

    >> we learned about #GIGO in both statistics and computer science.

    We never had to invent words like #hallucinate to evade accountability for problems in LLM and AI software itself. So AI hallucination is GIGO from the other end, not from bad user input.

    I chose translation software as an example. They do an admirable job in figuring out parts of speech, but not so well in figuring out figures of speech, eg Godfather makes an "offer you can't refuse."

    @avandeursen @TeflonTrout

  3. ⬆️ @saxnot

    >> we learned about #GIGO in both statistics and computer science.

    We never had to invent words like #hallucinate to evade accountability for problems in LLM and AI software itself. So AI hallucination is GIGO from the other end, not from bad user input.

    I chose translation software as an example. They do an admirable job in figuring out parts of speech, but not so well in figuring out figures of speech, eg Godfather makes an "offer you can't refuse."

    @avandeursen @TeflonTrout

  4. Ouch. Sullivan & Cromwell, over 900 attorneys, has a face full of egg. Letter of apology, hearing today w/ judge re: fake cases (hallucinated #citations).
    Are #attorneys so concerned w/ speed that they forget their duty to the #court? Yes.
    If a #FirstYear handed them this research (especially a woman), they'd check it. But because it comes from #TechBros, it gets a pass.
    #GenAI #hallucinate #cite #cases #Judge #bar #fine #censure #disbarment #legal #LawFirm #NYC
    news.bloomberglaw.com/business

  5. Ouch. Sullivan & Cromwell, over 900 attorneys, has a face full of egg. Letter of apology, hearing today w/ judge re: fake cases (hallucinated #citations).
    Are #attorneys so concerned w/ speed that they forget their duty to the #court? Yes.
    If a #FirstYear handed them this research (especially a woman), they'd check it. But because it comes from #TechBros, it gets a pass.
    #GenAI #hallucinate #cite #cases #Judge #bar #fine #censure #disbarment #legal #LawFirm #NYC
    news.bloomberglaw.com/business

  6. Ouch. Sullivan & Cromwell, over 900 attorneys, has a face full of egg. Letter of apology, hearing today w/ judge re: fake cases (hallucinated #citations).
    Are #attorneys so concerned w/ speed that they forget their duty to the #court? Yes.
    If a #FirstYear handed them this research (especially a woman), they'd check it. But because it comes from #TechBros, it gets a pass.
    #GenAI #hallucinate #cite #cases #Judge #bar #fine #censure #disbarment #legal #LawFirm #NYC
    news.bloomberglaw.com/business

  7. Ouch. Sullivan & Cromwell, over 900 attorneys, has a face full of egg. Letter of apology, hearing today w/ judge re: fake cases (hallucinated #citations).
    Are #attorneys so concerned w/ speed that they forget their duty to the #court? Yes.
    If a #FirstYear handed them this research (especially a woman), they'd check it. But because it comes from #TechBros, it gets a pass.
    #GenAI #hallucinate #cite #cases #Judge #bar #fine #censure #disbarment #legal #LawFirm #NYC
    news.bloomberglaw.com/business

  8. #MissKittyRaw @[email protected], while you #messianically #hallucinate your way into the grave and half the world with you, you cannot wonder why #productivity #has #cratered. Everybody sitting around wondering why they should bother doing anything since you're going to kill the world. 😵‍💫🤯😱

  9. #MissKittyRaw @[email protected], while you #messianically #hallucinate your way into the grave and half the world with you, you cannot wonder why #productivity #has #cratered. Everybody sitting around wondering why they should bother doing anything since you're going to kill the world. 😵‍💫🤯😱

  10. #MissKittyArtWalk #AI #Research Don't use #fast mode if you really want to be accurate. I had to go back for the third time on the prompt to get actual active accounts. Unbelievable. The willingness to just #hallucinate #wildly. I'm going to hide back in the #thinking mode later LOL.

  11. A Large Language Model (LLM) is a deep-learning algorithm, often using a transformer architecture, that is trained on massive amounts of text data to understand, process, and generate human-like text.

    A major shortcoming of LLMs is their tendency to "#Hallucinate" or confidently generate false or nonsensical information, along with the risk of perpetuating present in their training data.

    knowledgezone.co.in/trends/bro

  12. A Large Language Model (LLM) is a deep-learning algorithm, often using a transformer architecture, that is trained on massive amounts of text data to understand, process, and generate human-like text.

    A major shortcoming of LLMs is their tendency to "#Hallucinate" or confidently generate false or nonsensical information, along with the risk of perpetuating #Biases present in their training data.

    knowledgezone.co.in/trends/bro

  13. A Large Language Model (LLM) is a deep-learning algorithm, often using a transformer architecture, that is trained on massive amounts of text data to understand, process, and generate human-like text.

    A major shortcoming of LLMs is their tendency to "#Hallucinate" or confidently generate false or nonsensical information, along with the risk of perpetuating #Biases present in their training data.

    knowledgezone.co.in/trends/bro

  14. A Large Language Model (LLM) is a deep-learning algorithm, often using a transformer architecture, that is trained on massive amounts of text data to understand, process, and generate human-like text.

    A major shortcoming of LLMs is their tendency to "#Hallucinate" or confidently generate false or nonsensical information, along with the risk of perpetuating #Biases present in their training data.

    knowledgezone.co.in/trends/bro

  15. A Large Language Model (LLM) is a deep-learning algorithm, often using a transformer architecture, that is trained on massive amounts of text data to understand, process, and generate human-like text.

    A major shortcoming of LLMs is their tendency to "#Hallucinate" or confidently generate false or nonsensical information, along with the risk of perpetuating #Biases present in their training data.

    knowledgezone.co.in/trends/bro

  16. You know you're doomed when your operating system vendor is selling their "#AI" fetish to you with a text like this.

    »Agentic AI has powerful capabilities today—for example, it can complete many complex tasks in response to user prompts, transforming how users interact with their PCs. As these capabilities are introduced, AI models still face functional limitations in terms of how they behave and occasionally may #hallucinate and produce unexpected outputs. Additionally, agentic AI applications introduce novel #security #risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data #exfiltration or #malware installation.«

    support.microsoft.com/en-us/wi

  17. You know you're doomed when your operating system vendor is selling their "#AI" fetish to you with a text like this.

    »Agentic AI has powerful capabilities today—for example, it can complete many complex tasks in response to user prompts, transforming how users interact with their PCs. As these capabilities are introduced, AI models still face functional limitations in terms of how they behave and occasionally may #hallucinate and produce unexpected outputs. Additionally, agentic AI applications introduce novel #security #risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data #exfiltration or #malware installation.«

    support.microsoft.com/en-us/wi

  18. You know you're doomed when your operating system vendor is selling their "#AI" fetish to you with a text like this.

    »Agentic AI has powerful capabilities today—for example, it can complete many complex tasks in response to user prompts, transforming how users interact with their PCs. As these capabilities are introduced, AI models still face functional limitations in terms of how they behave and occasionally may #hallucinate and produce unexpected outputs. Additionally, agentic AI applications introduce novel #security #risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data #exfiltration or #malware installation.«

    support.microsoft.com/en-us/wi

  19. You know you're doomed when your operating system vendor is selling their "#AI" fetish to you with a text like this.

    »Agentic AI has powerful capabilities today—for example, it can complete many complex tasks in response to user prompts, transforming how users interact with their PCs. As these capabilities are introduced, AI models still face functional limitations in terms of how they behave and occasionally may #hallucinate and produce unexpected outputs. Additionally, agentic AI applications introduce novel #security #risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data #exfiltration or #malware installation.«

    support.microsoft.com/en-us/wi

  20. You know you're doomed when your operating system vendor is selling their "#AI" fetish to you with a text like this.

    »Agentic AI has powerful capabilities today—for example, it can complete many complex tasks in response to user prompts, transforming how users interact with their PCs. As these capabilities are introduced, AI models still face functional limitations in terms of how they behave and occasionally may #hallucinate and produce unexpected outputs. Additionally, agentic AI applications introduce novel #security #risks, such as cross-prompt injection (XPIA), where malicious content embedded in UI elements or documents can override agent instructions, leading to unintended actions like data #exfiltration or #malware installation.«

    support.microsoft.com/en-us/wi

  21. #OpenAI Says Models Programmed To Make Stuff Up Instead of Admitting Ignorance - Slashdot

    #AI models often produce false outputs, or "hallucinations." Now OpenAI has admitted they may result from fundamental mistakes it makes when #training its models. The Register:
    The admission came in a paper [PDF] published in early September, titled "Why Language Models #Hallucinate ,"
    #llm #hallucinations #artificialintelligence

    slashdot.org/story/25/09/17/17

  22. #OpenAI Says Models Programmed To Make Stuff Up Instead of Admitting Ignorance - Slashdot

    #AI models often produce false outputs, or "hallucinations." Now OpenAI has admitted they may result from fundamental mistakes it makes when #training its models. The Register:
    The admission came in a paper [PDF] published in early September, titled "Why Language Models #Hallucinate ,"
    #llm #hallucinations #artificialintelligence

    slashdot.org/story/25/09/17/17

  23. #OpenAI Says Models Programmed To Make Stuff Up Instead of Admitting Ignorance - Slashdot

    #AI models often produce false outputs, or "hallucinations." Now OpenAI has admitted they may result from fundamental mistakes it makes when #training its models. The Register:
    The admission came in a paper [PDF] published in early September, titled "Why Language Models #Hallucinate ,"
    #llm #hallucinations #artificialintelligence

    slashdot.org/story/25/09/17/17

  24. Says Models Programmed To Make Stuff Up Instead of Admitting Ignorance - Slashdot

    models often produce false outputs, or "hallucinations." Now OpenAI has admitted they may result from fundamental mistakes it makes when its models. The Register:
    The admission came in a paper [PDF] published in early September, titled "Why Language Models ,"

    slashdot.org/story/25/09/17/17

  25. #OpenAI Says Models Programmed To Make Stuff Up Instead of Admitting Ignorance - Slashdot

    #AI models often produce false outputs, or "hallucinations." Now OpenAI has admitted they may result from fundamental mistakes it makes when #training its models. The Register:
    The admission came in a paper [PDF] published in early September, titled "Why Language Models #Hallucinate ,"
    #llm #hallucinations #artificialintelligence

    slashdot.org/story/25/09/17/17

  26. Emphasis on #AI means lots of #scientists around the world now submit #research #papers to English-language #journals after using #AI to “write” them & to #hallucinate #reference #lists w/o consulting human #science #editors like me to fix things. If the world is lucky, some journals #retract them.

  27. Emphasis on #AI means lots of #scientists around the world now submit #research #papers to English-language #journals after using #AI to “write” them & to #hallucinate #reference #lists w/o consulting human #science #editors like me to fix things. If the world is lucky, some journals #retract them.

  28. Emphasis on #AI means lots of #scientists around the world now submit #research #papers to English-language #journals after using #AI to “write” them & to #hallucinate #reference #lists w/o consulting human #science #editors like me to fix things. If the world is lucky, some journals #retract them.

  29. Emphasis on #AI means lots of #scientists around the world now submit #research #papers to English-language #journals after using #AI to “write” them & to #hallucinate #reference #lists w/o consulting human #science #editors like me to fix things. If the world is lucky, some journals #retract them.

  30. Emphasis on #AI means lots of #scientists around the world now submit #research #papers to English-language #journals after using #AI to “write” them & to #hallucinate #reference #lists w/o consulting human #science #editors like me to fix things. If the world is lucky, some journals #retract them.

  31. How shall I phrase my system prompt to stop my #AI coding agent to #hallucinate methods and functions?

  32. How shall I phrase my system prompt to stop my #AI coding agent to #hallucinate methods and functions?

  33. How shall I phrase my system prompt to stop my #AI coding agent to #hallucinate methods and functions?

  34. How shall I phrase my system prompt to stop my #AI coding agent to #hallucinate methods and functions?

  35. How shall I phrase my system prompt to stop my #AI coding agent to #hallucinate methods and functions?