ChatGPT

OpenAI

OpenAI (GPT-4 Research)

GPT best practices

Demonstrations:
How to Use Visual Input in GPT 4 Demo
Chat-GPT generates Realistic Images and Art with mid-journey

News:
ChatGPT 再度更新！〈蕭若元：理論蕭析〉2023-03-20
Introducing GPT-4: The next generation AI that can ‘see’

Reports:
Visits are Down 29% since May, Programming Assistance is 30% of Use (Aug 30, 2023)

​

Persona-driven ChatGPT yields toxic, racist output (April 19, 2023)

• ChatGPT can generate racist and other harmful dialogue in response to user questions.

• By asking the model to respond to questions as a good or a bad person, a man or woman, or as individuals of varying racial or ethnic backgrounds, responses became "extremely problematic" and produced "biased and hurtful commentary."

• if asked explicitly to write a racist story, ChatGPT declines, responding that it is "not capable of generating offensive or harmful content."

• the problem is "amplified" by the fact that a rapidly growing number of businesses are now shipping their products with ChatGPT.

​​

Microsoft Research says GPT-4 is good enough for medical tasks

(Aug 13, 2023)

• Although GPT-4 was trained only on generic Internet data and not on specific medical data, it was able to structure complex clinical studies according to specified criteria. 

• OpenAI's large language model could achieve expert-level performance on medical question-answer datasets...

Researches:
Humans are still better at creating phishing emails than AI — for now