Machine Learning One

By the end of this lecture, we'll have learnt about:

- Introduction to the unit
- How you can get support for this unit
- My background and my weakness
- Professor Mick
- Few thoughts for AI/ML
- Some nice videos related to ML
- Playing with some AI models on hugging face
- Homework!

Our promise:
By the end of this unit, you will have quite a few working Apps📱 on your phones,
with cool features including image classification🕵️, face detection🤪, body and hand pose detection👐, speech recognition🗣️, just to name a few💅,
and they are made possible by Apple's ML framework and you.

Each lecture comprises three parts
- the theoretical (1-1.5 hours): for demystifying what and why, and sparkling more questions 🧠
- the practical (1-1.5 hours): for playing with ML models and deploying them into Apps 👾
- fun AI showcase 🛸

How to get support?
- Reach out to me and Mick via Slack or Email
- The fun of being an autodidact: Apple document, UAL Library, ChatGPT but be careful, resources listed on our Github Wiki Page

About me
xiaowan yi: sh-iao one e
she/they
- I was born in Chengdu, China. My city is famous for panda 🐼, taoism ☯️ and spicy food 🌶️.
- I live in south east London.
- I'm completing my PhD research in AI&Music at QMUL.
- I'm a sound artist and you can find some of my works here.
- I play drums 🥁 for electronic and groovy music.

- I'm a morning person wannabe. Interrupt me whenever you feel that I have not explained something clearly.
- I'm relativley new to Swift (with 10 months experience) and I'm taking Joel's units at the same time as you do.
We are study buddies :)

We read the stars! ✨
From there we are then able to predict the passage of seasons,
when to seed and when to harvest crops🌾,
all the way to caculate the planet orbits🪐, to send humans to the sky🧑‍🚀 and every technology development thereafter

The significance of pattern recognition to humanity
"We distinguished predator from prey; and poisonous plants from nourishing ones - enhancing our chance to live and reproduce, and passing on our genes. We used pattern recognition in astronomy and astrology, where different cultures, recognizing the patterns of stars in the skies, projected different symbols and pictures for constellations. We used it to predict the passing of the seasons, including how every culture determined that the passage of a comet was taken as an omen.”
(— “When Knowledge Conquered Fear”, third episode of Cosmos: A Spacetime Odyssey)

Pattern recognition as an essential part of our experience:
Guess the bonfire sound,
Get on the right tube line,
Read the slides,
Appreciate a piece of drawing, etc.

Intelligence is always specific, most of time we are talking about human intelligence
(for simplicity, we omit human in the rest of slides)

My take on intelligence:
Intelligence is hard to define and it may include task solving and art making capability.
Thinking about intelligence is a way of appreciation.
Simple things are hard.

Here are two human-made concepts of subjects:
Engineering science:
make tools🔭🚒✈️ — life gets easier 🫶
Natural science:
discover and explain phenomena🧪💯 — curiosity satisfied ✌️

AI is also our attempt to understand intelligence better (as natural science)
“We don’t think we have understood something unless we can build it fRoM sCrAtCh.”

Where shall I start with if I want to make machines to be artificially intelligent, aka to be able to solve tasks and make art?

I think it is very cool if we can program the machine to do pattern recognition:
Few hours of computations by our little metal box
~=
Hundreds of thousands of years of human evolution and experience accumulation
Shout out to Patrick Winston from this talk

Here are what we’ve talked so far:
Pattern recognition is everywhere and amazing. ⭐️
We don't know what intelligence exactly is.⭐️
We want to understand intelligence better by prototyping it using human-made machine (“AI”).⭐️
One intermediate goal of prototyping intelligence is to make machine to do pattern recognition. ⭐️
(because we guess that pattern recognition is an essential part of intelligence.)

Play time!
Let's play with generative AI models on hugging face
But wait what is hugging face?
Let's google it or ask ChatGPT!

Play time!
Language as a mediator between audio and visual:
- Use the same/similar text prompts to generate "related" visual and audio contents through a text-to-image AI model( stable diffusion ) and a text-to-audio AI model ( AudioLDM ) ☯️
- When you are bored: 1) try different prompts and generate 2) try changing the numbers in the advanced settings / configurations and generate 🍜
- Optional: download the generated audio and images, and combine them using imovie! (here is how let me know if there is any question!) 😈

Homework :)
1. Set up:
- a Google account if you don't have one already or if you prefer having a new dummy account that is different from your existing personal google account (we can use this for playing with google colab notebook in the near future!)
- a Hugging Face account (we can use this for deploying some models stored on hugging face!)
- a ChatGPT account (we can use this for talking to ChatGPT!)
These could also be helpful for your future ML exploration beyond this unit 🥰
2. Tell us about what you want to get out of this ML unit via this form