• DarkThoughts
    link
    fedilink
    111 year ago

    I doubt anyone can for as long as “AI” is synonymous with LLMs. LLMs are just inherently unreliable because of how they work.