Comparison of Sorting Algorithms animations
Binary Search dance - Flamenco, 3 mins
Linear Search dance - Flamenco, 7 mins
Bubble Sort dance - Hungarian, 5 mins
Insert Sort dance - Romanian, 4 mins
Quick Sort dance - Hungarian, 5 mins
Select Sort dance - Gypsy (do we still use this term? unsure), 5 mins
Merge Sort dance - German, 4 mins
Heap Sort dance - Hungarian, 10 mins
4 Queens Problem dance - Ballet - 6 mins - this demonstrates a backtracking search algorithm with the N Queens problem - how to create a program that looks for N queens on a chess board that do not exist on the same row, column, or diagonal.
Blown to Bits - You Life, Liberty, and Happiness After the Digital Explosion
Available for download under Creative Commons license
from the Marshall Memo
In this article in School Library Journal, Virginia educators Idamae Craddock and Kristen Wilson suggest six ways to improve results when tapping into ChatGPT and other large language models:
Be specific. Include the information, tone, voice, audience, length, and type of output you’re seeking; long prompts will work just as well as a short one.
Provide background information. If you’re asking for a lesson plan, tell how many students you’re teaching, their ages, the languages they speak, and the preceding curriculum unit.
Mind your manners. If your prompt uses crude language, the LLM will respond in kind; if you use polite language, you’ll train it to respond in kind.
State your constraints. You can request a 20-minute lesson plan, a 750-word article, a list of 10 historical events.
Fine-tune. If you don’t get what you’re looking for the first time, re-word your prompt.
Push back. “If the output is incorrect, too long, too short, the wrong tone, wrong information, or wrong structure,” say Craddock and Wilson, “you need to reply with that feedback. Treat the prompt as a conversation – that’s what it is.”
“Six Prompt Tips” by Idamae Craddock and Kristen Wilson in School Library Journal, November 2023 (Vol. 69, #11, p. 16)