There is a recent spawn of tools “powered by AI”, “powered by ChatGPT”, … e.g. the “Kubectl OpenAI plugin”. Such third party clients may limit, or outright sabotage, you in certain ways however: often based on the older, more primitive language model endpoints (davinci-003 in this case) unaware of breaking changes in the target subject sa... Read more 21 Mar 2023 - 1 minute read
Today’s discussion revolved around capabilities of various OpenAI language models, focusing on the differences between the Davinci model and the newer GPT models. During these conversations, we touched on the specific use cases of using these models for SEO purposes and as a personal assistant. First, it’s important to note that there is a vari... Read more 20 Mar 2023 - 1 minute read
Inspired by YouTube content discussing the viability of ChatGPT as a developer assistant (akin to GitHub Copilot), I shared my perspective on how OpenAI’s capabilities have evolved since I received my invitation in July 2022: After some update, ChatGPT-3 managed to figure out access strategies to nested JSON/XML within REST API results. Around ... Read more 19 Mar 2023 - 1 minute read
I am excited to share that I got my GPT-4 waitlist request for expedited access honored. This report is based on my first evening of experimentation with the new AI model (meaning: GPT-4 proper, not through ChatGPT Plus). When filing the waitlist request, I expressed my research interest in exploring GPT-4’s capabilities in inquiry spanning acro... Read more 18 Mar 2023 - 2 minute read
During a discussion in March 2023 a colleague asked why text-davinci-003 produced concise, predictable answers, whereas ChatGPT tended to be more verbose and occasionally digressive. The explanation is largely technical. Model family text-davinci-003 belongs to the GPT-3 series. ChatGPT, at that point in time, was powered by GPT-3.5 and had b... Read more 30 May 2025 (Updated) - 1 minute read