I was delighted to read (4) as my colleagues were busy creating a new logo. Sadly Gemini refused the prompt telling me "I cannot create this logo for you. The request to include a "cute animal clutching a bag of money" and "expensive scientific equipment" falls under my safety guidelines that prevent the generation of content that could be interpreted as unethical or promoting problematic associations." However ChatGPT immediately produced a "blind monkey" clutching said items. Should I be worried most by Google over zealous ethical constraints, OpenAI's openness to 'whatever you want' or my need to actually test your prompts?
‘Re-write the following text using a range of terminology and concepts from [insert discipline here]’ - how to turn something otherwise perfectly readable into a piece of opaque writing which nonetheless will earn the apparent author some academic capital.
Also ‘Rewrite this text as far as possible in plain English, avoiding jargon wherever possible, but without losing any essential content, so it could be read by a lay person. Try and keep as close as possible to the original length.’ I have been trying this with some academic writings - and sometimes the result has *fewer* words.
This reminds me to recommend "Dear Committee Members" a book by Julie Schumacher written entirely through memos and emails and the like. My son gave me a copy and it is laugh out loud funny and also made me wince at times. Its pre-AI so an update might be an idea!
I must say I like em (—) and en (–) dashes. A major contribution I have made to one of the universities I worked for may have been explaining to a very senior member of the executive the difference between the hyphen and the dash (I left the em-dash for later) — as you can imagine, punctuation in all sorts of policy documents was very poor.
I also love semicolons, and use them profusely; I wonder why they are not used more often in AI-generated documents.
Love these targeted prompts. Super interesting! A series of three simple prompts I found quite helpful (and insightful) recently were: What do you know about me? What are my strengths and weaknesses? Write a plan to optimise me. Having spent a year or so utilising chatgpt for many aspects of work and home it was eye-opening and incredibly helpful.
I was delighted to read (4) as my colleagues were busy creating a new logo. Sadly Gemini refused the prompt telling me "I cannot create this logo for you. The request to include a "cute animal clutching a bag of money" and "expensive scientific equipment" falls under my safety guidelines that prevent the generation of content that could be interpreted as unethical or promoting problematic associations." However ChatGPT immediately produced a "blind monkey" clutching said items. Should I be worried most by Google over zealous ethical constraints, OpenAI's openness to 'whatever you want' or my need to actually test your prompts?
Amazing
Persona - missing in action!
I love these!
God Bless Chat GPT . So useful for taking the I out of AI, or the P out of technology.
‘Re-write the following text using a range of terminology and concepts from [insert discipline here]’ - how to turn something otherwise perfectly readable into a piece of opaque writing which nonetheless will earn the apparent author some academic capital.
Also ‘Rewrite this text as far as possible in plain English, avoiding jargon wherever possible, but without losing any essential content, so it could be read by a lay person. Try and keep as close as possible to the original length.’ I have been trying this with some academic writings - and sometimes the result has *fewer* words.
This reminds me to recommend "Dear Committee Members" a book by Julie Schumacher written entirely through memos and emails and the like. My son gave me a copy and it is laugh out loud funny and also made me wince at times. Its pre-AI so an update might be an idea!
£2.99 on Amazon Kindle - looking forward to reading it!
I must say I like em (—) and en (–) dashes. A major contribution I have made to one of the universities I worked for may have been explaining to a very senior member of the executive the difference between the hyphen and the dash (I left the em-dash for later) — as you can imagine, punctuation in all sorts of policy documents was very poor.
I also love semicolons, and use them profusely; I wonder why they are not used more often in AI-generated documents.
Love these targeted prompts. Super interesting! A series of three simple prompts I found quite helpful (and insightful) recently were: What do you know about me? What are my strengths and weaknesses? Write a plan to optimise me. Having spent a year or so utilising chatgpt for many aspects of work and home it was eye-opening and incredibly helpful.
I love how #9 has simply accepted the situation as is ;-).
6 is particularly cruel - I may steal the idea