Uncategorized
-
This lie could be because of hype, a self created illusion, a desire of how LLMs should be rather than how they are, not keeping up with how the world moves –> the reasons could be any. Whatever the reason might be, “One can use natural language to communicate with LLMs” is a very misleading
-
A key aspect of prompting an LLM is to make up your mind. For a given problem, what is going to be your role in combination with the LLM. This is where you take control of the conversation in your hands E.g. Who is going to generate the ideas? You? LLM? Both? Who’s defining the
-
As my experiments with ChatGPT went deeper, the prompts were getting more and more complex. Sometimes, all such complexities have simple solutions, which we know about, which we have used elsewhere but we are not able to recognise them at the moment. The solution to re-structuring the prompts was really simple – Use Markdown. I
-
LLMs could have been worse 😁 . It’s great, that they aren’t that bad as of today. They are in a shape where they can be directed with careful prompting. This is not to be confused with over-positioning and hype. But we also need to do better than fear and negation. PS: If you are
-
When exploring LLMs becomes a headache, make it fun. About the screenshot: Although more understandable to my Hindi-bhaashi and Hindi-knowing friends, you can pick up the idea and find your way of having fun with it. PS: Technically, this is the Root Prompt Pattern. Non-technically? Well, just enjoy maadi.
-
Asking an LLM to directly solve a math problem is problematic, given the nature of tech. Good. Now, let’s move from Exploring to Test -> Exploring to Make User Of – when it comes to a technology. Let’s see if we can use the prompt structure from my previous posts to solve this math problem.
-
Because LLM’s are not meant for precise calculations. They are subsequent token predictors based on language patterns. The problem discussed here can be solved using LLM by constructing a Prompt fit for the purpose. You:Goal is to do basic arithmetic. To do this calculation: 1. Ask me what is the arithmetic expression to be evaluated.2.
-
Twenty years is not a long enough time. It’s too short to write the kind of things I am going to say here. However, twenty years of working in the field of testing is long enough to understand and realise how little I know. It’s a long enough time to do experiments which are only
-
Generative AIs have gained a lot of attention and popularity, pretty quickly. This is only going to head in a forward direction in terms of their reach and the interest they generate. Applications of AI have become more and more approachable for otherwise AI-ignorant people like myself. Having access in the form of prompting to
-
Dear Testers, Let us translate and project. That’s what we do when dealing with something new or completely unknown or scarcely known. The context is usage of LLMs in testing. I’m going to project based on my knowledge of automation framework design. A test automation framework typically is a layered work. E.g. let’s say one
Rahul Verma: My Petty Bits 