CSC-105: Thinking With Machines

Prompt engineering and AI fluency.

👋

About me

  • Designer, developer, teacher
  • 30+ working in design and development
  • Cofounder/CTO of Pathwright (my day job)
  • Adjunct professor for 18+ years (Web Design, UI/UX Design, Intro to Computer Science)
  • Ongoing research in AI & Learning
  • I have a twin (watch out!)

The PathMX research paper

Everything is changing

(again)

We're in a weird timeline

  • We can talk to our computers now?!
  • AI is taking your job?
  • AI is evil/terrible
  • AI is exciting/interesting

Will AI replace programming?

When things are changing fast, its time look at history

Time is a flat circle

An America Online installation CD

The America Online login screen

A GeoCities homepage

A GeoCities homepage for a cat

The Macromedia Flash logo

The Portland Studios website

The PARG game website

The NoiseTrade music-sharing website

Steve Jobs introducing the iPhone

A sad SWF file icon

Steve Jobs beside an Adobe Flash gravestone

Twins starting something new

Liminal Spaces

backrooms

Sam Altman kills your job?

here we go again

Here we go again.

Coding vs. Programming

  • Coding is the syntax/typing of code
  • Programming is the act of solving a problem with code
  • Programming is the hard part with or without AI

My Opinion

  • AI likely will replace most coding.
  • AI will not replace programming, it will make programming much more powerful.

"The hot new programming language is "English"."

"Prompt Engineering"

Using words to shape the output of a machine

Malleable Software

Malleable Curriculum

What to expect

  • Change
  • Lots of new tools
  • Personalized learning (malleable curriculum)
  • Hands-on work (Thursday)

Principles for AI Use

These are the principles we strive to follow when learning and building with our thinking machines.

1. Own your understanding.

Do not outsource learning to another human or agent. Unless you understand it and can articulate it clearly, it will not materially impact your life or future work.

The work of understanding cannot be done on your behalf. It's similar to learning to play an instrument: you can't hire someone to learn piano for you.

2. Machines are tools.

AI, LLMs, and other machines are not conscious or sentient; they are best thought of as "calculators for everything." They are useful utilities, not conscious entities.

Do not anthropomorphize AI!

3. Put humans first.

Our work serves real human needs. We are not building tools for tools' sake.

4. Use your own voice.

Don't let AI write, design, or think for you. Use your own words and voice. Writing is thinking. It's hard to care about things you didn't put your own voice into. We want to build things that we care about because that's how our best work happens.

5. Know when to say enough.

AI will build and build and build ... you right into a corner. It's way too easy to slop up a project and almost impossible to de-slop it later without starting over. Focus on things that you actually need and will use.

Class Website

👉 csc105.path.app

Questions?