Finding Meaning in AI Generated Content

Eugene Based Group

Meetups last Thursdays

Except Nov and Dec

Meetup Organizer

Senior Engineer

Antonio Ortega Jr

Thank you!!

https://idxbroker.com/

twenty Ideas

https://twentyideas.com/

The District

https://thedistrictcowork.com/

https://www.slicepizzeriaandbar.com/

https://www.fallingskybrewing.com/

This is a non-technical technical talk

New AI Services 2022

GitHub Copilot

Chat GPT

Stable Diffusion

Text Buzz:

Chat GPT

Image Buzz:

Stable Diffusion

DALE - 2

Midjourney

The Buzz

Can AI Replace Humans ...?

Machines can and have before

A Timeless Question

shorturl.at/cqCIT

History of Lace 1865

Earlier in 2022

Actual Job Lost to Machines

What is AI selling?

With this technology becoming more accessible it raises bigger questions way beyond the "will it replace people?" question.

curl https://api.openai.com/v1/images/generations \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -d '{
  "prompt": "Eugene Web Devs is a good meetup",
  "n": 2,
  "size": "1024x1024"
}'

Easy for Automation

Write a description of Chat GPT

ChatGPT is a conversational language model developed by OpenAI. It uses a neural network architecture known as a transformer to generate human-like text based on a given prompt. The model is pre-trained on a large dataset of text and can be fine-tuned for specific tasks such as question answering and dialogue generation. ChatGPT is capable of understanding context and generating coherent and fluent responses, making it a useful tool for natural language processing and human-computer interaction.

Data is prepared for machines

Explain the meaning of

"You can’t judge a book by its two birds with one stone"

Attaching parameters from a set of 175 billion

Chat with this Artificial Intelligence

journalistic texts, fiction and short stories, how to, computer code, scientific articles, even general chat with this Artificial Intelligence.

Generate text like:

For 💰💰💰

How Machines Learn

Data Set

Lots and lots of data samples.

1. Train

Feed the machine data to find patterns in the samples.

2. Refine

Iterate and refine to get good results.

3.

(in a nutshell)

Numbers Represent Words and Images

Data Sample

Machines Are Good at Patterns

Steps in a Neural Network Process

Sample Data pieces

Predict and iterate

Magic Math

Attempt to Classify Input and/or Generate New Output

Patterns are everywhere

and

machines are good at finding them

AI Image Creation Combines:

Natural Language Model

Image Generator Model

Image Validation Model

In Practice?

Local Artist: Alysa Brown

This is a dazzling 11x14 piece displaying a colorful portraiture of a wolf. It's a profile view of the wolf looking off into the distance. The fur is made of gentle brush strokes consisting of layers of yellow, pink, purple, orange, and brown hues. The soft edges of the fur blend nicely into the background. Light grey tones surround the wolf, which really makes the wolf pop. Blue paint drips fall from the top, along with some purple streaks which really add to the cohesiveness of the painting.

/alybrowny