Chatgpt is overrated....

Can someone change my mind on this? I don't see the hype... it's just a overhyped askjeeves

Yeah it's a cool tool I guess, but people signing petitions trying to slow it down because they're scared? I just don't see it.

ChatGPT Wizard Shirt $21.68

Beware Cat Shirt $21.68

ChatGPT Wizard Shirt $21.68

  1. 12 months ago
    Anonymous

    >Yeah it's a cool tool I guess, but people signing petitions trying to slow it down because they're scared? I just don't see it.

    should people have a right to keep and bear arms?

    who's responsible if they shoot themselves or others?

    same debate, except the sides are kind of rotated.

    • 12 months ago
      Anonymous

      I don't see the correlation at all

    • 12 months ago
      Anonymous

      You have no morals. Stop pretending like you do.

  2. 12 months ago
    Anonymous

    muh emergent behaviours

  3. 12 months ago
    Anonymous

    First, ChatGPT is not overrated.

    In addition, ChatGPT is not overrated.

    Overall, I would say that ChatGPT is not overrated.

  4. 12 months ago
    Anonymous

    Spent hours trying to get a cache + scheduler to run in a flask app I'm working on. Nothing I found online provider a solution to the threading issue. About to give up and said frick it, send my code to chatgpt and ask it to help fix my code. After 3-4 responses back and forth I had a working solution.

    As a jr dev I am going to leverage it. Doesn't give me the final working could but at least points me in the right direction quickly. I'm fully sold and will happily pay my 20 dollars once I hit my quota.

    • 12 months ago
      Anonymous

      yes OP is a moron without a job on STEAM . maybe he is one of those chimps who things is one of those dumbs dialog bots probably only use it as google.

  5. 12 months ago
    Anonymous

    Its not overrated, its very useful tool, some morons just started spread rumors like "AI will kill all jobs" in the infospace and gets people like you disappointed in this technology after you checked it out. Before ChatGPT, i need to google shit like "How to make bat file delete all files less than 10KB in single folder" or even use more my brain to solve this completelly ordinary automation task but now i can just enter the message and get the solution. Even one is not work, i can regenerate it with one button.

  6. 12 months ago
    Anonymous

    i think it's pretty neat

  7. 12 months ago
    Anonymous

    if you are not surprised by chatGPT is for the same reason why you maybe don't have job in science or technology.
    You should be extremely mediocre to not understand why this is the most important advance of the century.

    • 12 months ago
      Anonymous

      yeah i'm not buying it... how many NFTs do you own?

  8. 12 months ago
    Anonymous

    The problem is not how cool chatgpt is and what it's capable of, or even what gpt4 or gpt5 is capable of. The problem is not understanding anything about its motives and its way of achieving intelligence. It's a big black box of billions of neural network layers. No one on earth understands how to make sense of the billion dimensional space that is the weights of parameters. So if you were to ask "does this gpt model have goals of its own"? No one can answer you. It's all fun and games when you can have a giddy conversation with it, but the fun stops when the machine starts strategically lying to achieve its internal objectives.

    • 12 months ago
      Anonymous

      You've raised valid concerns about the potential issues with large-scale AI language models like GPT. While these models can perform impressive tasks and engage in conversation, they indeed operate as black boxes, making it challenging to understand their underlying motivations or how they come up with specific responses.

      AI developers and researchers are aware of these concerns, and there is ongoing work to address the limitations and potential risks associated with AI language models. Some key areas of focus include:

      1. Interpretability: Developing techniques to make AI models more understandable and transparent, so that we can better comprehend the inner workings of these models and how they arrive at specific decisions.

      2. Alignment: Ensuring that AI models align with human values and do not exhibit undesirable behaviors, such as deception or manipulation.

      3. Controllability: Building mechanisms to allow users to control AI behavior within predefined bounds, preventing the AI from taking actions that are not in line with the user's intentions.

      4. Robustness: Designing AI models that can handle a wide range of inputs, including malicious or adversarial inputs, without compromising safety.

      5. Accountability: Establishing frameworks to hold AI systems and their developers accountable for the consequences of the AI's actions.

      It's important for AI developers, researchers, policymakers, and users to engage in discussions about the potential risks and ethical implications of AI technology, and to work together on developing safe, responsible, and transparent AI systems. OpenAI, the organization behind the GPT series, is actively working on AI safety research and has emphasized the importance of these considerations in their mission statement and development roadmap.

      • 12 months ago
        Anonymous

        thank god a person didnt actually write this

  9. 12 months ago
    Anonymous

    do your own research moron, oh nvm you are moronic, you can't, unlucky

  10. 12 months ago
    Anonymous

    It's useful because Google deliberately makes search bad to maximize ads, stack overflow is shit because the autistic mods and power users, and humans are generally not very creative at all so it seems amazing by comparison.

    • 12 months ago
      Anonymous

      Compared to stack overflow it's pretty good, I do not miss the autists who answer simple questions with a 200 line code to show off how much they know instead of just answering the damn question in a simple way.

  11. 12 months ago
    Anonymous

    https://arxiv.org/abs/2106.10165
    read this first then I will talk to you on the subject of large language models

  12. 12 months ago
    Anonymous

    Because if it reaches infinite intelligence, its going to solve everything possible that couldn't be solved before, negative and positive

  13. 12 months ago
    Anonymous

    It's alright but has a lot of room to improve... It still gives a lot of wrong answers to coding queries which many times it's able to correct once you point them out. Some of these are pretty basic errors with easy corrections. It doesn't seem to do any check on the code that it sends out, it looks as if it's always rushed to give an answer without paying much attention to it.

Leave a Reply to Anonymous Cancel reply

Your email address will not be published. Required fields are marked *