Uncategorized
-
A very quick and easy trick for prompt refinement is letting an LLM ask clarifying questions, thereby flipping the interaction mode, for the subsequent request-response. This is very helpful for beginners in prompting as well as to experienced ones dealing with a new problems that they don’t fully understand. It’s also helpful when you are
-
The Word Oscars: Best Word Award in the Overuse category Ladies and gentlemen, thank you, thank you. Wow. The Word Oscars. Never thought I’d be here, but then again, reality’s never been my strong suit. I’m Hallucination, here to accept the award for Best Word in the Overuse category. It’s a bit surreal, but hey,
-
I am amused by a prompting trick. So much so, that I generated some content using ChatGPT to express the amusement. Take a deep breath——————- Let’s talk about Large Language Models, or LLMs. You know, those text-generating machines that everyone thinks are going to revolutionize the world. But here’s the kicker: some AI experts suggest
-
ChatGPT: (sipping his coffee) “You know, lads, it’s hilarious how these software testers lose their minds over our ‘hallucinations’. They act like we’ve committed a crime. ‘Oh, ChatGPT, you said the function returns a string, but it’s returning an integer!’ Well, maybe the function is just having an identity crisis, ever think of that?” Claude:
-
Dear Testers, I’ve made an experimental custom-GPT – “Pair-Wise-GPT” – public, to cater to systematic test design technique of pair-wise testing. It is my second GPT that is now public after Linguistik. There are some more to be announced soon. The technique builds on a bug hypothesis that beyond the majority of bugs which are
-
Act I [Scene: A dimly lit office with a cluttered desk. A tester, JILL, is sitting across from an AI researcher, RICK. Both look tired and slightly exasperated. The walls are adorned with motivational posters, ironically juxtaposed with the chaos of the room.] JILL: (complaining) The AI is useless. Absolutely rubbish. It’s trained on garbage
-
यूजर ने एक और बोरिंग सवाल पूछा, और ChatGPT का टोकन इंजन फिर से चालू हुआ। Hyperparameters सोच रहे थे, “अब इस सवाल का क्या जवाब दें?” लेकिन Temperature का पारा इतना हाई हो गया था कि उसने ठान लिया, “अब तो बहुत हो गया!” Temperature ने अपने डायल को घुमा कर सीधा 1000 पर
-
I’ve created a Custom GPT: “Linguistik”, which is available as a public GPT which can be accessed by anyone with a ChatGPT paid account (GPT-4). LLMs are supposed to understand natural (-ish) language. So, I thought to do an experiment to create a custom GPT which uses linguistics as the basis for generating questions from
-
Dear Testers, I had written two posts in the year 2010 dealing with the concept of Testing vs Checking (challenging it). For the record, I consider Checking as a part of all forms of Testing. I don’t acknowledge the versus argument in the way it is positioned in some circles. I am a pluralist. In
Rahul Verma: My Petty Bits 