Category: Generative AI
-
Omni my goodness!
🔥GPT4 Omni has been announced🔥… and in an AI world of jaw-dropping announcements, the fanfare was relatively low key, but this is absolutely huge. Technically the GPT-4o model is different. Human-computer interactions become more natural by combining audio (recog and speech), visual (recog and gen) and text in one model. It’s also fast⏩. Even faster…
-
Your name isn’t “f##t” – Persistent Politeness reveals the truth
Secret identifiers 🕵️, Cookies 🍪, GPS coordinates 🛰️ and hidden fields, these and other techniques are used to find out things about you. Addin makers have whole businesses built around stopping the latest dodgy techniques. It can be easier! Did you know it’s possible to get ChatGPT to find out information about your user in…
-
7 Principles for Reducing Latency in LLM Applications
When I first started developing with LLMs the logical way of providing your own data to an LLM was to add it to the prompt. Imagine my suprise when the LLM got slower and slower. Of course I blamed GPT4! 🙂 In the rapidly changing world of AI, large language models (LLMs) like GPT-4 is…
-
ChatGPT believes all older men have beards
In order to see if I could get past the filters in ChatGPT to find out what it really thought of controversial figures I thought I’d trick it. Perhaps I could learn about AI biases….and I did. I learnt that it loves beards. Initially I copied the wikipedia article for a certain controversial American politician…
-
Comparing ChatGPT, Claude, and Gemini: An Insight into AI Bias
Explore how ChatGPT, Claude, and Gemini view their own bias.