I'm not a so called "prompt engineer", btw have been using multiple LLMs.
OpenAI GPT, Google Gemini, Microsoft Phi-2, Google Palm, Meta llama 2, Anthropic Claude 2
Why are these so dumb?
And why is billion-dollar companies keep investing in these LLM models without innovating some serious shits.
If Google hadn't open-sourced transformer architecture, these damn AI companies fucked.
Why I feel like this is just another Web 3/ blockchain type thingy
OpenAI GPT, Google Gemini, Microsoft Phi-2, Google Palm, Meta llama 2, Anthropic Claude 2
Why are these so dumb?
- Can't answer a simple question without hallucinations
- Can't innovate new stuff
- Can't even do pre-school mathematics
- It's like a search engine with added non-sense
And why is billion-dollar companies keep investing in these LLM models without innovating some serious shits.
If Google hadn't open-sourced transformer architecture, these damn AI companies fucked.
Why I feel like this is just another Web 3/ blockchain type thingy