an (artificial) MachIne SafeTy Expert for Y‘all
Disclaimer: Mistey is a research tool, not an Oracle. Use good judgement in your interactions and always confirm safety-critical information with a primary source. MAKESafe takes no responsibility for what she says. If in doubt – use the chat box on the bottom right of your screen to talk to a real person at MAKESafe.
FAQs about Mistey
Mistey is an artificial intelligence. More specifically, she’s an advanced language model that uses Open AI’s Chat GPT engine.
Ask her questions, just like you would a person. She will remember the context of your conversation and the best answers come from repeated back-and-forth clarifications.
Mistey’s general training has been supplemented with information from OSHA and from MAKESafe’s internal experts but in short, no – don’t trust her. Mistey is an early generation language model and (like most people) she makes lots of mistakes, including accidentally fabricating complete lies every once in a while. However, she can still be a useful tool for research. Just make sure to check primary sources for any safety critical advice. Or better yet – use MAKESafe’s chat box (lower right of the screen) to talk to a real person from MAKESafe Tools.
Safety is hard and it can be overwhelming to sift through mountains of safety standards. Mistey is our attempt to make it easy to ask machine safety-related questions. She’s used best as a first step in your learning and can often open the door to topics and solutions you wouldn’t have thought of or found in your traditional research and googling. Just make sure to confirm things with a primary source of information since she makes mistakes! And to be clear, we didn’t “make” Mistey. We simply took Open AI’s language model and taught it a few things about machine safety.
If you want to start again from scratch, empty your reply field and the ‘send’ button will turn into a ‘clear’ button.