Year: 2023

Today I’m testing the new GPT-4 Web Browsing model. The following use case might be useful to people in #Sales, #Product, and any other person doing #marketresearch.

I’m working with a client to show them how they can use AI to:

1. clone their voices
2. translate their video… https://t.co/CSNuGQDnLn https://t.co/l8uQQSxMgT

Let’s dedicate a little longer to the new #ControlNet Reference-only preprocessor for #AUTOMATIC1111 / Vladmandic.

Somebody has commented that this technique generates results that look like poor outcomes of a text2img generation.

It’s really not the case: it’s… https://t.co/Es7KnhjPbF https://t.co/1ks799TpFb

@Scobleizer So, at least 2 people are betting their reputation on this device 🙂

Let me close this weekend’s study of the new #ControlNet Reference-only preprocessor for #AUTOMATIC1111 / Vladmandic with my favourite fine-tuned #stablediffusion model ever: #analogdiffusion, by @wavymulder as well.

To quote the U2: Faraway, So Close! https://t.co/nRLEPtmKHG

With some more serious effort, I’m getting somewhere: https://t.co/CnGGumbc5i

Let’s switch to Realistic Vision v2.0, my third favourite fine-tuned #stablediffusion model.

Again: ControlNet with the new Reference-only preprocessor.

Notice how the colours of the background have been influenced by the uniform of Commander Data. And, of course, the facial… https://t.co/3oBMZKeEDc https://t.co/A1eoXd1TEs

🚨Nerd and #AIart Alert 🚨

Context
I dedicate my weekends to art. And lately, AI art.
To generate the covers of each Synthetic Work issue, I use a fantastic fine-tuned version of #stablediffusion called #collagediffusion, by @wavymulder.

The point
Today, I’m testing a new… https://t.co/NpiaLoaRbr https://t.co/Bdh9gEQH6J

🚨Nerd and #AIart Alert 🚨

Context
I dedicate my weekends to art. And lately, AI art.
To generate the covers of each Synthetic Work issue, I use a fantastic fine-tuned version of #stablediffusion called #collagediffusion, by @wavymulder.

The point
Today, I’m testing a new… https://t.co/7TUbzUFp21 https://t.co/1fWL2iM2Rh

@kunalomizer One replacement at a time 🙂

More seriously: should we ever get to this scenario, I think people will still want to have some human beings evaluating the performance of the synthetic executive.

Who’s gonna pick the right AI to be hired?

Or, what if the synthetic executive is… https://t.co/SOJ8gyyy8z

@darrenmason That’s true also for humans, no? And yet, VC firms insist on funding founders that were previously successful, and boards insist on hiring CEOs that were previously successful.

It seems like, collectively, we humans respond to your (true) point with a “Yeah yeah. We’ll take the… https://t.co/fEuObXDAuF