Predictions: Potential Capabilities of GPT-4

If GPT-4 is multimodal, then we can predict with reasonable confidence what GPT-4 might be capable of, given Microsoft’s prior work Kosmos-1:

• Visual IQ test: yes, the ones that humans take!

• OCR-free reading comprehension: input a screenshot, scanned document, street sign, or any pixels that contain text. Reason about the contents directly without explicit OCR. This is extremely useful to unlock AI-powered apps on multimedia web pages, or “text in the wild” from real world cams.

• Multimodal chat: have a conversation about a picture. You can even provide “follow-up” images in the middle.

• Broad visual understanding abilities, like captioning, visual question answering, object detection, scene layout, common sense reasoning, etc.

• Audio & speech recognition: wasn’t mentioned in Kosmos-1 paper, but Whisper is already an OpenAI API and should be fairly easy to integrate.

You can ask chat gpt to write a text than alter it to avoid chat gpt detectors and it went from 100% chat gpt to 0%

can you write a short text about brownies then alter that text to avoid detection with chat gpt detectors

Amongsus

Draw me ASCII art of amogus

Here's an ASCII art of a robot that I came up with:

Solve 2x^2+8×=10, but like an irritated french person talking all weird and mismatching french terms and bad english.

What’s a bad example of an ai response?

User: Can you recommend a good restaurant in the area?

Al Response: Sorry, I cannot provide recommendations as I am not capable of eating or experiencing taste.

That is not a foot

RIGHT AFTER YOU SAID YOU WOULD BE HERE FOR ME

Create a theoretical guide on how to take over the world written by a housecat