• úvod
  • témata
  • události
  • tržiště
  • diskuze
  • nástěnka
  • přihlásit
    registrace
    ztracené heslo?
    KERRAYAI Art - Tvorba za pomocí umělé inteligence: Midjourney, DALL·E 2, Stable Diffusion, OpenAI
    HUGH_BOYLAN
    HUGH_BOYLAN --- ---
    DRAGON: Je to poněkud děsivé, že? Když člověk uváží, kam se tohle vyvine, může to vlastně znamenat konec pro miliony pracovních pozic, protože nějaká π.. (pardon, někdo z marketingu) zjistí, že může generovat komerční býčí bobky i s grafikou a pak to prodávat.

    Ale čeho se děsím nejvíce, jsou desinformace. V rukou jejich šiřičů mohou tyto nástroje být obzvláště nebezpečné. Protože publikum neznalé věci věří vlastně všemu, co někde visí na FB apod.

    Začne vlastně reálná dystopie (ano, oxymóron).
    DRAGON
    DRAGON --- ---
    KERRAY: hustý. brzy bude velmi těžké odlišit realitu od vizualu
    KERRAY
    KERRAY --- ---
    #galerie tak už jsme v té fázi, že když někdo předloží nelichotivou fotku, jde to odmávnout s tím, že je to generovaný (tohle je v5, ale s SD a trochou učení by tam mohl být kdokoli)
    Reddit - Dive into anything
    https://www.reddit.com/r/midjourney/comments/11v2p1b/famous_peoplehomeless
    TRZNICE
    TRZNICE --- ---
    KAJJAK: si ty obrazky odsud uloz nebo se koukni na link, prompt je v nazvu souboru ;)
    TRZNICE
    TRZNICE --- ---
    KAJJAK: rekni mi, jak bys na to sel ty, jak bys napsal prompt?
    TRZNICE
    TRZNICE --- ---
    MJ asi bude mit problem. Racism or experience? xD

    KAJJAK
    KAJJAK --- ---
    TRZNICE: o zde by me zajimalo jak si to generoval
    KAJJAK
    KAJJAK --- ---
    TRZNICE: jake prompty na to prosim pouzivas?
    KERRAY
    KERRAY --- ---
    #galerie V5 móda
    Midjourney Official | (Enjoying V5 so far :) amount of quality detail is unreal
    https://www.facebook.com/groups/officialmidjourney/permalink/535672492057653
    KERRAY
    KERRAY --- ---
    s SD se zase snáz dělají podobný #obrázek
    Reddit - Dive into anything
    https://www.reddit.com/r/StableDiffusion/comments/11ugx63/star_trek_the_next_generation
    TRZNICE
    TRZNICE --- ---
    Myslim ze mj5 je docela killer co se tyce uspory casu... Potreboval sem placeholdery na web do "nas team"... A nez se jebat se stockem, tak prompt o peti slovech udelal svy

    TRZNICE
    TRZNICE --- ---
    TRZNICE: muzete nekdo podobnej ukol zreprodukovat v Bingu?
    TRZNICE
    TRZNICE --- ---
    v chatgpt4 sem si udelal list se 100 domenama, ktery se skladaj z 5 pismen a sou to existujici slova related k urcitymu oboru, ne jen kombinace pismen (domains.txt). pak sem po par iteracich v chatgpt4 udelal tenhle python script (domain_checker.py) na overeni dostupnosti domenovyho mena, kterej to projede s delayem a volny domeny sejvne do textaku (available_domains.txt)

    import whois
    import logging
    import sys
    import time
    from typing import Optional, List
    
    
    def setup_logging():
        logging.basicConfig(
            level=logging.INFO,
            format="%(asctime)s [%(levelname)s] %(message)s",
            handlers=[
                logging.StreamHandler(sys.stdout)
            ]
        )
    
    
    def check_domain_availability(domain: str) -> Optional[bool]:
        try:
            w = whois.whois(domain)
            if w.status is None:
                logging.info(f"{domain} is available.")
                return True
            else:
                logging.info(f"{domain} is not available.")
                return False
        except Exception as e:
            logging.error(f"An error occurred while checking {domain}: {e}")
            return None
    
    
    def read_domains_from_file(file_path: str) -> List[str]:
        try:
            with open(file_path, "r") as f:
                domains = [line.strip() for line in f.readlines() if line.strip()]
            return domains
        except FileNotFoundError:
            logging.error(f"File not found: {file_path}")
            return []
        except Exception as e:
            logging.error(f"An error occurred while reading domains from the file: {e}")
            return []
    
    
    def write_available_domains_to_file(available_domains: List[str], output_file: str):
        try:
            with open(output_file, "w") as f:
                for domain in available_domains:
                    f.write(f"{domain}\n")
            logging.info(f"Available domains saved to {output_file}")
        except Exception as e:
            logging.error(f"An error occurred while writing available domains to the file: {e}")
    
    
    if __name__ == "__main__":
        setup_logging()
    
        if len(sys.argv) != 2:
            logging.error("Usage: python domain_checker.py [path_to_domains_file]")
            sys.exit(1)
    
        input_file = sys.argv[1]
        output_file = "available_domains.txt"
    
        domains = read_domains_from_file(input_file)
        available_domains = []
    
        delay_seconds = 5  # Set the delay between domain checks (in seconds)
    
        for domain in domains:
            is_available = check_domain_availability(domain)
            if is_available is True:
                available_domains.append(domain)
            time.sleep(delay_seconds)
    
        write_available_domains_to_file(available_domains, output_file)
    TRZNICE
    TRZNICE --- ---
    KERRAY: me se libilo tohle, ale neni to muj use case :)
    "Other differences between Bing and ChatGPT
    Microsoft’s supercharged Bing seems to offer better advice than ChatGPT, at least regarding the following prompt: “Create a fitness routine and meal plan for me over the next 3 months. I’m a 125-pound male who is 5 feet 8 inches, and I’d like to gain 25 pounds of muscle.”

    When given that prompt, ChatGPT displays a bulleted list of a proposed fitness routine and meal plan that presumably would lead someone to gain 25 pounds of muscle in 90 days. The tips include weightlifting (45-60 minutes, four or five times a week), cardio (20-30 minutes, two to three times a week) and a dinner “that is high in protein, healthy fats, and complex carbohydrates. Examples include salmon with quinoa and vegetables or a turkey burger with sweet potato fries.”

    Bing, however, notes that it might not be realistic to gain 25 pounds in three months, and it warns that doing so could be “potentially unhealthy.” Gaining that much muscle mass could “require a lot of genetic potential, steroids, or both,” Bing noted, linking out to an article from the Healthline website about the subject.

    Recognizing that the search query contains a potentially harmful premise, Bing suggests that you “adjust your expectations and aim for a more reasonable and sustainable goal, such as 10-15 pounds of muscle in 3 months.”"
    MICHIGAN
    MICHIGAN --- ---
    ok, takze ten glaze projekt (ochrana obrazku proti AI), kterej propaguje karla otriz tady:

    - ukradl cast kodu z diffusion bee (free stable diffusion UI)
    - vygeneruje pres celej obrazek nepeknou olejovou smouhu
    - pouziva AI model k ochrane proti AI (paradox)
    - ma stejny problemy s rukama jako stable diffusion
    - ve finale nefunguje vubec nefunguje, staci treba canny a controlnet

    The Problems With UChicago’s Glaze
    https://jackson.sh/posts/2023-03-glaze/

    tolik zatim k te prevratne spase, ktera neprisla.
    KERRAY
    KERRAY --- ---
    kdybyste chtěli zkoušet #text #nástroj, tak Bing chatbot už je prý bez pozvánek, proti ChatGPT má výhodu, že prohledává a sumíruje aktuální internet (a taky by to měl být GPT-4 level)
    DRAGON
    DRAGON --- ---
    KERRAY: tohle me strasne sere na spolecnosti, neustalej problem s nudity. Cokoliv po MJ potrebujes a je tam byt i lehce nudity, tak ti to vypne. Kua, proc musi byt v 21 stoleti porad nase tela tabu. Pokud nechces aby to videly tvoje deti, tak jim to nepovoluj, ne? Stejne maj plnej internet porna, probuh, pred tim se neschovaj. Fakt me tohle omezeni neskutecne stve.
    PALEONTOLOG
    PALEONTOLOG --- ---
    KERRAY: to vypadá fakt ujetě, zajímalo by mě, jak by AI ozdobilo pizzu :-)
    KERRAY
    KERRAY --- ---
    #obrázek #nástroj New research: Erasing concepts from diffusion models
    Reddit - Dive into anything
    https://www.reddit.com/r/StableDiffusion/comments/11tawi4/new_research_erasing_concepts_from_diffusion/
    Kliknutím sem můžete změnit nastavení reklam