• Subscribe
  • Honestly, I'm shocked by the capabilities of OpenAI's "strawberry" model.

    Viktor Surkov
    8 replies
    The way they implemented the chain of thoughts (CoT) is just brilliant - it's dynamic, diverse, and does exactly what you need. Is this really so close to how people think and solve problems? Small example: I asked today to build me a Chrome extension that would analyze the content of a webpage and leave a relative comment according to my writing style(and uploaded hundreds of my comments). And it did. In 27 sec. From first try. With OpenAI API integration. Ready to use. I just can't find the words to describe how impressed I am, and what a massive effect it's going to make on our economy in consequent years. What are your thoughts on that?

    Replies

    Veeresh Devireddy
    Interesting, let me explore and share my experience. Thanks, @viktor_surkov for the example.
    Share
    Brij Mohan Lal Srivastava
    Very interesting example, I wonder what are the limits of Strawberry
    Share
    Ethan Samuel Bennett
    Wow, that's pretty mind-blowing! I've been playing around with Claude a bit and it's super capable. Definitely curious to try this 'strawberry' model and see how it compares. The rapid advancements in these AI language models is both exciting and a little scary. Thanks for sharing the example @viktor_surkov, look forward to hearing others' experiences with it too!
    Share
    Holly Cook
    I'm equally amazed by the chain of thought functionality. It’s like the model is able to think through complex tasks in a way that feels more human than ever. The potential for automation and creativity is limitless.
    Share
    Viktor Surkov
    @holly_cook exactly! It's moving toward consciousness. If you read through its chain of thoughts, you can notice that a lot of the first things of what it's doing is trying to understand context, and when you ask direct questions "you think" - it asks itself "Who am I in context of that request". And that's scary since consciousness partially is self-awareness.
    Connor Reid
    That's incredible! The fact that it can mimic your writing style with such accuracy is mind blowing. It really makes me think about how this tech could streamline so many processes.
    Share
    Viktor Surkov
    @connor_reid GPT 4o with ease, was mimicking your style. o1 is a different beast you should try that one for some tasks
    Share
    Zoe Palmer
    I've been testing similar things with Open AI, and the speed is what shocks me the most. It’s not just smart it’s fast . The possibilities for scaling projects are crazy.
    Share