//Gallery

A dumping ground for some of the things I’ve been playing with from late 2022 to today.

Prompt: Be Creative

Playing around with Claude 3.5 Artefacts and Projects. Very cool. LinkedIn post and comparison with ChatGPT GPTs here.

Asking ChatGPT to “Be Creative”, with HTML and Javascript as its canvas. LinkedIn post here.

Conversational UX

Playing with a voice conversational interface. This tool helps to build a guided LLM enabled conversation. It guides the LLM in how to respond to requests from a user, whilst moving the conversation towards a particular outcome.

Streaming avatars

Playing with HeyGen’s streaming avatars API – coded real time conversational avatars, trained on your own content.

ComfyUI

Installed ComfyUI over the weekend and have been starting to play with custom workflows. Start off simple enough, with a very steep learning curve to attain more complex workflows (e.g. AnimateDiff).

Midjourney Character Consistency

Small experiment using Midjourney’s CREF and CW parameters and write up and findings here.

GPT that makes phone calls

I wrote a short post about this cool experiment which connects a GPT to Bland.AI to make phone calls on your behalf. It uses my voice too.

Stable Diffusion video

I’ve been using RunwayML Gen-2 for a while now to create text/image-to-video, but Stability’s announcement of a video model is exciting, which I’ve been playing with (doesn’t seem to handle faces very well … yet!).

Workchopping 😉

Been having fun running generative AI workshops for a mix of clients, in person and remotely too. 200+ generative AI use cases designed, over 10 in development.

Use cases

Parts from a good Inform session brining use cases to life for a client.

Midjourney 5.2 zoom out

Playing with Midjourney new ‘zoom out’ feature. Pretty cool, like an extension of generative fill. How many applications does this have!!

Writing snake with ChatGPT

Chat GPT writes exceptionally good code … but the process requires some work. Read more here.

Surf Life Saving demo

Snippets from a short engagement with Surf Life Saving.

Building an app with ChatGPT

Yep, coding iOS / Android apps with ChatGPT is pretty simple stuff.

Visual Studio Code

Getting my hands dirty with some hand-on playing with OpenAI API.

Stable Diffusion – Pose

Mucking around with Pose based prompts.

Stable Diffusion – Scribble

Mucking around with Scribble based prompts (Prompt: “aboriginal art style of Bungle Bungles”).

Time Under Tension assets

Workshop assets.

Whimsical

AI assisted mind-mapping tool whimsical.com (all branches AI generated)

Soundraw.io

Easy and quick interface to create AI generated music – no more licensing issues!

Midjourney

A random selection from the Midjourney community – so powerful.

Dalle-2

Prompt: “Photograph of a man hiking through the Australian bush up on a mountain with a backpack on a hot sunny day near a rock water pool.”

Stable Diffusion

Prompt: “A Triumph Speedtriple style sports motorcycle racing on a road through the Australian bush at sunset.”