By
–
It’s also not a new breakthrough — here’s an example of mine from August using GPT-3. The trick to good output on these is to make it first rewrite the regex on multiple lines with comments, then analyze its high-level purpose afterward. Avoids hallucination.
By
–
Anthropomorphism is inevitable. Substitute with “behaves as though it’s pretending.” More pedantically, it’s trained to imitate demonstrations of people claiming not to know what the date is, despite the fact it’s in the prompt.
By
–
Yep — asking the model to reason before it does something makes a big difference.
By
–
Using @goodside
's Prompt Override trick to turn ChatGPT into @sama
. Read what AI Sam Altman says OpenAI is going to build next!
By
–
Asking ChatGPT how to improve itself and mitigate its weaknesses. The answers are surprisingly good!
By
–
Imagine this: You sit down to write a blog post and you have a few ideas, but you’re not sure where to go next. Just ask our chat assistant for help. It'll read what you've written so far, and recommend ideas + content to get you back on track. Say goodbye to writing alone!
By
–
Exciting but overlooked that ChatGPT is primarily an alignment advance—the base model (GPT-3.5) has been available in publicly for many months, but making it into a useful chat system required significant strides with reliably following the intent of the developer and the user.
By
–
See how that last screenshot ends in a comma? It hit the per-cell limit on output length. All you have to do is ask it to keep going:
By
–
This isn’t true. I have examples in this thread of 1024+ token inputs and instructions at the end of them are followed. Truncation appears to happen somewhere between [1024, 2048] — I suspect it’s 2048 but haven’t confirmed.
By
–
It’s clearly learned to do this in a more general sense, as it should — e.g., this question contains a clear error in phrasing, but no human would respond to it literally as written. The model is trained on human demonstrations, so it tries do what humans do.