The prompts that "should be illegal" were all weaksauce. I've got a few prompts that you could maybe say should be illegal and they're like 2 pages long and do hours of work for you in one go
ETHICS
-
Securing AI Prompts: Leak Prevention and ChatGPT Updates
By
–
Look for all the possible ways in which prompt can be leaked and try to rectify that by imposing programatic filters. A brilliant example is how @OpenAI improved the security of ChatGPT with the recent update so that even with tweaking of input you can’t get the source prompt.
-
Negative Context Issues and Topic Removal Solutions
By
–
Same issues no matter how you word it. If you bring up a topic in any negative context, you might get a positive result on the same topic. Best way around it is giving a task to explicitly remove that topic.
-
Teaching Writing in the Age of AI Assistance
By
–
I'm starting to think about this a lot lately. How do we teach the next era of would-be writers to make the best use of this tool? How do we help people learn writing in these conditions where the work can be done 80%+ of the way for them?
-
The Paradox of Choice: AI Tools and Decision-Making Anxiety
By
–
The Paradox of Choice suggests that too many options make people less happy, and more anxious So exercise caution with ChatGPT and similar AI tools. Don't put yourself into a state of overwhelm. Too many choices make decision-making harder, not easier.
-
GPT-2 Output Detector Limitations and AI Text Detection Evolution
By
–
No realmente. GPT-2 Output Detector, como bien indica su nombre, funciona para GPT-2. Fue una tecnología que OpenAI sacó cuando publicaron GPT-2 hace años, y aunque funcionaba, manipulando un poco el texto se podía engañar al detector. Ya con GPT-3 y demás IAs ni lo intentaron.
-
Embracing Our Past Selves: Identity and Technological Change
By
–
“I’ve already lost touch with a couple of people I used to be. I think we are well advised to keep on nodding terms with the people we used to be, whether we find them attractive company or not”- the brilliant Joan Didion.
-
Simulating Trump vs ordinary people: AI capability differences
By
–
Like he says ppl like Trump would be easy to simulate while I say normal people are easy to simulate
-
Bypassing AI Detection: Natural Text and Watermarking Constraints
By
–
I'm not sure about that. E.g. tricking the ZeroGPT algo means actually writing more natural text (so it'll be better, not worse), and tricking the watermarking means not being limited by the constraints of the watermarking algo.