What even is AI?
- Students are asked to define AI in their own terms and play a game of “AI or not?”
- Defining AI vs. automation [Optional: tour of Molecular Maker Lab]
- Exploration of how artificial intelligence comes to be – how do computers think? Make decisions? Is it really better than doing things on our own or sticking with automation?
- Choose your own adventure: [k-nearest neighbor analysis] or/and [neural network] activity.
- Exploration of AI search engines such as DALL·E
- Exploration of MMLI research: MMLI version = Text2Mol (Use of AI to generate molecules based on text prompt).
Parallels to MMLI Research:
While all MMLI thrusts explore and utilize the tenets of artificial intelligence, this lesson explores some of the work performed in Thrust 1. Students learn how computers think and learn, and how that relates to everyday problems. Computers can make decisions that we can make, but when it comes to hundreds and thousands of options to filter through, computers play a huge role in doing that background work for us. Students learn two ways computers learn: KNN analysis and neural networks. Students are introduced to MMLI research called “Text to Molecule” which functions similar to many AI image searches but with the specific function of producing molecules based on a given prompt. This research is being performed by PhD student Carl Edwards in the lab of Dr. Heng-Ji. Input from Carl and his lab’s research largely helped shape this curriculum.
Resources
-
Text2Mol
