ChatGPT 101

Sdílet
Vložit
  • čas přidán 11. 09. 2024
  • It is an incredible computer program that can chat with you just like a person would. It's like having a super-smart friend who knows a lot about everything! This program has been trained using tons of information from books, articles, and the internet, so it has a wide range of knowledge, but what truly is ChatGPT? Halıcıoğlu Data Science Institute Assistant Professor Justin Eldridge breaks down the large language model chatbot and helps us understand the breakthroughs and implications of this new artificial intelligence. Recorded on 04/19/2023. [6/2023] [Show ID: 38930]
    More from: Data Science Channel
    (www.uctv.tv/da...)
    #chatgpt #ai #machinelearning #artificialintelligence #computerscience
    Explore More Science & Technology on UCTV
    (www.uctv.tv/sc...)
    Science and technology continue to change our lives. University of California scientists are tackling the important questions like climate change, evolution, oceanography, neuroscience and the potential of stem cells.
    UCTV is the broadcast and online media platform of the University of California, featuring programming from its ten campuses, three national labs and affiliated research institutions. UCTV explores a broad spectrum of subjects for a general audience, including science, health and medicine, public affairs, humanities, arts and music, business, education, and agriculture. Launched in January 2000, UCTV embraces the core missions of the University of California -- teaching, research, and public service - by providing quality, in-depth television far beyond the campus borders to inquisitive viewers around the world.
    (www.uctv.tv)

Komentáře • 14

  • @uctv
    @uctv  Před rokem

    Check out "Dark Persuasion - The History of Brainwashing from Pavlov to Social Media" here: czcams.com/video/bWl8FXhdEio/video.html

  • @PCMenten
    @PCMenten Před rokem +1

    Cathy Wood differs about LLMs, saying that training that used to cost millions now can be done for thousands and that trend shall continue.

  • @metaphysicalArtist
    @metaphysicalArtist Před rokem +2

    ChatGPT 3.5 free today: Please keep in mind that these assessments are approximate and subjective, as they are based on the data I have been trained on rather than individual language proficiency tests. Here's a rough estimate of my proficiency in the languages:
    English: Very high proficiency
    Spanish: High proficiency
    French: High proficiency
    German: High proficiency
    Italian: High proficiency
    Portuguese: High proficiency
    Dutch: Moderate to high proficiency
    Russian: Moderate to high proficiency
    Chinese (Simplified and Traditional): Moderate to high proficiency
    Japanese: Moderate to high proficiency
    Korean: Moderate to high proficiency
    Arabic: Moderate to high proficiency
    Turkish: Moderate to high proficiency
    Swedish: Moderate proficiency
    Polish: Moderate proficiency
    Danish: Moderate proficiency
    Norwegian: Moderate proficiency
    Finnish: Moderate proficiency
    Greek: Moderate proficiency
    Hebrew: Moderate proficiency
    Hindi: Moderate proficiency
    Bengali: Moderate proficiency
    Indonesian: Moderate proficiency
    Thai: Moderate proficiency
    Vietnamese: Moderate proficiency
    Czech: Moderate proficiency
    Hungarian: Moderate proficiency
    Romanian: Moderate proficiency
    Ukrainian: Moderate proficiency
    Catalan: Moderate proficiency
    Filipino (Tagalog): Moderate proficiency
    Malay: Moderate proficiency
    Persian (Farsi): Moderate proficiency
    Slovak: Low to moderate proficiency
    Slovenian: Low to moderate proficiency
    Croatian: Low to moderate proficiency
    Serbian: Low to moderate proficiency
    Bulgarian: Low to moderate proficiency
    Lithuanian: Low to moderate proficiency
    Latvian: Low to moderate proficiency
    Estonian: Low to moderate proficiency
    Icelandic: Low to moderate proficiency
    Swahili: Low proficiency
    Tamil: Low proficiency
    Telugu: Low proficiency
    Malayalam: Low proficiency
    Kannada: Low proficiency
    Marathi: Low proficiency
    Gujarati: Low proficiency
    Punjabi: Low proficiency
    Welsh: Low proficiency
    Irish (Gaelic): Low proficiency
    Scottish Gaelic: Low proficiency
    Basque: Low proficiency
    Galician: Low proficiency
    Albanian: Low proficiency
    Maltese: Low proficiency
    Yiddish: Low proficiency
    Armenian: Low proficiency
    Georgian: Low proficiency
    Macedonian: Low proficiency
    Bosnian: Low proficiency
    Kurdish: Low proficiency
    Amharic: Low proficiency
    Somali: Low proficiency
    Hausa: Low proficiency
    Zulu: Low proficiency
    Xhosa: Low proficiency
    Sesotho (Southern Sotho): Low proficiency
    Swazi: Low proficiency
    Tswana: Low proficiency
    Afrikaans: Low proficiency
    Xitsonga: Low proficiency
    Oromo: Low proficiency
    Lingala: Low proficiency
    Luganda: Low proficiency
    Malagasy: Low proficiency
    Uzbek: Low proficiency
    Kazakh: Low proficiency
    Pashto: Low proficiency
    Nepali: Low proficiency
    Sinhala: Low proficiency
    Bengali (Bangla): Low proficiency
    Gujarati: Low proficiency
    Kannada: Low proficiency
    Telugu: Low proficiency
    Odia: Low proficiency
    Assamese: Low proficiency
    Maithili: Low proficiency
    Sindhi: Low proficiency
    Burmese: Low proficiency
    Khmer: Low proficiency
    Lao: Low proficiency
    Tibetan: Low proficiency
    Dzongkha: Low proficiency
    Maldivian (Dhivehi): Low proficiency
    Sundanese: Low proficiency
    Javanese: Low proficiency
    Tagalog: Low proficiency
    Hawaiian: Low proficiency

  • @gerardtrigo380
    @gerardtrigo380 Před rokem +5

    There is the problem that ChatGPT lies. There are a number of cases where students have asked ChatGPT to write a paper using references. They will get a well written paper with totally bogus references.

    • @JoeyVol
      @JoeyVol Před rokem +1

      You still have to use your own personal intelligence and craftiness as to HOW you use it, like with most tools. Although they do claim GPT version 4 corrects this issue by a huge margin, in the range removing 95%+ of outright falsehoods that gpt 3 suffers from - as you rightly state.
      The insane part is they’re already testing GPT 4 code meaning it’s live on their internal servers already and it’s only been a few months since the last release version launched 😮
      If said student uses it to create a paper and then simply fact checks it with the assistance of Google search, that route is far “easier” than having to manually study the material and then conceptualize a paper on any said topic or subject.

    • @beaulah_califa9867
      @beaulah_califa9867 Před rokem

      so it's like a human being. lies and all. let me know when ai beings a holocaust or genocide or lets me people starve to death and sleep outside. oh, humans got that covered.

    • @gerardtrigo380
      @gerardtrigo380 Před rokem

      @@JoeyVol I think using ChatGPT to write your papers is a form of cheating. Some method of checking for this is needed.

    • @ThomasBaxterSoutar
      @ThomasBaxterSoutar Před 7 měsíci +1

      Dont humans around you lie too? How do you detect their lies? Probably not using AI? ;) x

  • @icherishcrochetandknit309

    I belong to the artistic world of the web and I have seen renditions by Chat that were nowhere near interpreting a crochet pattern, a dance routine, etc... and to date, tjhere are only knittiing machines. MAN cannot achieve producing and manufacturing a crocheting machine. Will Chat be able to design one?

  • @bhowmikubkv
    @bhowmikubkv Před rokem

    Very informative

  • @ThomasBaxterSoutar
    @ThomasBaxterSoutar Před 7 měsíci

    Left to Right...Terabytes...seconds (time)...teacher ...creativity is maybe only human quantum (on our planet anyway?)

  • @AvinashTavares
    @AvinashTavares Před rokem

    i tried the last test. It said "The sentence "This is a real test of artificial intelligence." has 8 words."