FAQ

Classes

Projects

About

Blog

Store

Support Us

Newsletter

Email

IG

School

for

Poetic

Computation

Apply Now

Handmade Datasets: Strategies for working critically with small data and Artificial Intelligence

Teachers
Aarati Akkapeddi, Isabella Haid
Date
Section 1: June 14, 2026 to August 16, 2026
Section 2: June 17, 2026 to August 19, 2026

(10 classes)
Time
Section 1: Sundays, 6:30–9pm ET Section 2: Wednesdays, 6:30–9pm ET
Location
Online (Zoom)
Cost
$1200 Or pay $600, $300, or $0 with scholarship
Deadline
Applications open until April 27, 2026

Apply Now

Description

What if datasets were small enough to care for? This class traces the data pipeline behind major AI systems and proposes an alternative: handmade datasets built with intention, consent, and scale. Students of all levels will train custom models using human-scale data and a fraction of the energy resources of large AI models .

Images courtesy of teachers.

Too long; didn’t read

  • Level: Beginner–Intermediate
  • Tools: Google Colab, GANs, LoRA, RAG, RAVE
  • Focus: Small-scale, ethical machine learning + data collection
  • Prereqs: Basic computer literacy; curiosity about AI + data ethics
  • Project: A handmade dataset + trained custom model

Disclaimer

While this class works primarily with Google Colab and Google Drive, there are alternatives to Google when it comes to working with ML that will be shared as resources to explore in the future.

Full Description

How does data become AI—and who does the labor? Handmade Datasets unpacks the pipeline behind large AI systems like Stable Diffusion by tracing the journey from web scraping to training to generation.

We'll examine each step: how websites get crawled and archived, how images and their alt-text descriptions become training material, how click-workers and automated systems curate data, and how all of this shapes what a model generates. We'll then propose an alternative: "handmade datasets," or human-scale, personally assembled datasets, small enough that creators can handle each datapoint.

Drawing on examples from artists like Anna Ridler and Stephanie Dinkins as well as events like the Dataset Farmers Market, we'll explore how slowness can be a form of subversion, creating space for consent, stewardship, and intentionality.

Students will learn practical techniques for training models with limited data, including data augmentation, GANs, LoRA, RAVE, microgpt and RAG. Through hands-on projects and critical discussion, students will develop both the technical capability to work with small-scale ML and a more nuanced understanding of data collection and the labor behind AI systems. By the end of the class, participants will have created their own handmade dataset and trained a custom model.

Course of Study

  • Week 1: What is a Handmade Dataset?
  • Week 2: Dataset Prep and Augmentation
  • Week 3: Image Models pix2pix
  • Week 4: Stable Diffusion Models and Fine-tuning
  • Week 5: LoRA
  • Week 7: Intro to Audio Models
  • Week 8: Audio Models Continued
  • Week 9: Guest Lecture (TBD) and Final Project Intro
  • Week 10: Final Project Presentations

Expectations

Time & Workload

3–4 hours a week outside class on weekly assignments building toward a final project

Technical Experience

No prior machine learning or advanced coding experience required. Students should have basic computer literacy and a willingness to experiment.

Materials
  • Google account with ≥5GB storage
  • Students will need to purchase compute units from Google Colab (~$30–45 recommended).

Is this class for me?

This class may be for you if you:

  • Think the term “artificial intelligence” is loaded and assumes a lot about what intelligence actually is
  • Value human labor and thought
  • Take joy in messy, circuitous, and slower routes over the fastest way from A to B

This class may NOT be for you if you:

  • Want to take this class so that you can create an AI startup company
  • Want to become a mid (journey) artist
  • Think your colleagues at work should just get used to using ChatGPT because AI is here to stay

Meet the Teachers

teacher

Aarati Akkapeddi

Aarati Akkapeddi is a cross-disciplinary artist, coder, and educator based in Lenapehoking (Brooklyn, NY). They often use personal and institutional archival materials, combining computational and analog techniques like machine learning & printmaking to create artwork that investigates overlooked relationships and histories. Their creative work has been supported by institutions such as The Photographers' Gallery, ETOPIA Center for Art & Technology, and LES Printshop. They work at The Experimental Humanities Collaborative Network, creating digital spaces and tools.

they/them · website · instagram

teacher

Isabella Haid

Isa Haid is a researcher, writer, and artist. She is fascinated by the intertwined histories of labor, technology, and infrastructure. Seeking new network topologies to reclaim our technological imagination, Isa likes to experiment with the in-between of physical and digital media. She is a co-founder of MissVideo4u, a USB-based video distribution network. Isa works at the Museum of the Moving Image in Queens, NY and is affiliated with the Border Tech Lab at UT Austin.

How do I apply?

Apply Now

Applications open until Applications closed on April 27, 2026.

You can expect to hear back from us about the status of your application on May 11, 2026. Please email us at admissions@sfpc.study with any questions you have.

How much does it cost to attend?

For 10 classes, it costs $1200 + processing fees, for a one-time payment. We also offer payment plans. Participants can schedule monthly payments of the same amount. First and last payments must be made before the start and end of class. *Processing fees apply for each payment.

SFPC processes all payments via Withfriends and Stripe. Please email admissions@sfpc.study if these payment options don't work for you.

Applicant FAQ

For more information about what we look for in applicants, scholarships, and other frequently asked questions, please visit our applicant FAQ.

Interested in more learning opportunities at the School for Poetic Computation? Join our newsletter to stay up to date on future sessions and events, and follow us on Instagram and Twitter. Support our programming through scholarships. Get in touch over email.