3 ms·

> the year 7 comp sci classes they teach in our local high school have what on their curriculum? Yep, that's right, you guessed it: AI. Because that's apparentl

by xg15 1y ago

> the year 7 comp sci classes they teach in our local high school have what on their curriculum? Yep, that's right, you guessed it: AI. Because that's apparently the absolute basic CS that every student should start with these days. I think, if you went back to the origin of the term "AI" and tried to teach an introduction to the very fundamentals, this could actually be a fun and inspiring class - one that might not even need a lot of computer knowledge. There are a number of board games with "self-playing" antagonists that are governed through clever sets of game rules. There is also the historical predecessor of computer science, cybernetics, that dealt with self-governing analogous control systems, like thermostats. Finally, there are the classical pathfinding algorithms (Depth-First/Breadth-First, Dijkstra, A*) which I still think are some of the most "bang for the buck" algorithms in terms of "intelligent-looking" behavior vs simplicity of the algorithm. All that stuff could be engaging for high school students in the author's "hands-on" way. All that of course if the "AI" class is really about giving a broad introduction to the field, and not just "we have to put ChatGPT into the curriculum somehow". > After all, flashing screen surely release more endorphins than non-interactive physical exhibits The irony is that this might not even be true. In the article, the author observed that the physical exhibits were much more interesting to the kids than the screens.

bee_rider 1y ago
Another funny option could be to have the AI class be a linear algebra class.
- moi2388 1y ago
  Nothing funnier than tricking teens into liking math
  - fxtentacle 1y ago
    Show them Red Faction (the video game) and then explain that all the destruction is calculated using binary space partitioning and that all they need for making similar games is to accurately calculate the intersection of planes, lines and points. Add a few linear forces and numerical integration and there you have your Trojan horse for getting kids hooked.
    - xg15 1y ago
      ...and then hit them with the "Attention is all you need" paper!
    - moi2388 1y ago
      Oh, I absolutely love that you referenced that game. What a wonderful experience that was as a kid
      - bee_rider 1y ago
        Was that the one where you could blow up almost every wall (or at least the main type of terrain, which seemed to be some sort of dirt or rock). No idea if the game was actually fun to play “competitively.” But as a tech demo it rocked.
        moi2388 1y ago
        It was. The campaign was good fun
    - jldugger 1y ago
      > Show them Red Faction (the video game) Modern example: Donkey Kong Bananza[1] [1]: https://www.youtube.com/watch?v=mIddsPkdX9U https://www.youtube.com/watch?v=mIddsPkdX9U
      - fxtentacle 1y ago
        I'm 100% sure that they never intended this for my age group ... but that trailer looks like it'll be so much fun. Thank you for sharing :)
        jldugger 1y ago
        IDK your age group, but there's tons of nostalgic references and easter eggs. The game appears to be based on the Super Mario Odyssey engine, and the way Nintendo generally approaches the "all ages" problem is to make the plot easy, with tons of optional challenges and a set of hard levels in the "post-game" content. But I'm not very far in yet, so I can't promise anything on that front.
jldugger 1y ago
Well, the first paper on AI was published in 1943[1], but unfortunately requires some knowledge of brain science that was ultimately incorrect to understand, and includes some mathematical notation that is relatively obscure for high school. And this, Turing, and Shannon predate the Macy conferences / Ratio Club by a few years, so I'm not sure I'd call cybernetics a predecessor so much as a defunct offshoot. Though where math stops and CS starts varies depending on the scholar you talk to. [1]: https://jontalle.web.engr.illinois.edu/uploads/410-NS.F22/McCulloch-Pitts-1943-neural-networks-ocr.pdf https://jontalle.web.engr.illinois.edu/uploads/410-NS.F22/Mc...

by xg15 1y ago

bee_rider 1y ago

Another funny option could be to have the AI class be a linear algebra class.

moi2388 1y ago
Nothing funnier than tricking teens into liking math
- fxtentacle 1y ago
  Show them Red Faction (the video game) and then explain that all the destruction is calculated using binary space partitioning and that all they need for making similar games is to accurately calculate the intersection of planes, lines and points. Add a few linear forces and numerical integration and there you have your Trojan horse for getting kids hooked.
  - xg15 1y ago
    ...and then hit them with the "Attention is all you need" paper!
  - moi2388 1y ago
    Oh, I absolutely love that you referenced that game. What a wonderful experience that was as a kid
    - bee_rider 1y ago
      Was that the one where you could blow up almost every wall (or at least the main type of terrain, which seemed to be some sort of dirt or rock). No idea if the game was actually fun to play “competitively.” But as a tech demo it rocked.
      - moi2388 1y ago
        It was. The campaign was good fun
  - jldugger 1y ago
    > Show them Red Faction (the video game) Modern example: Donkey Kong Bananza[1] [1]: https://www.youtube.com/watch?v=mIddsPkdX9U https://www.youtube.com/watch?v=mIddsPkdX9U
    - fxtentacle 1y ago
      I'm 100% sure that they never intended this for my age group ... but that trailer looks like it'll be so much fun. Thank you for sharing :)
      - jldugger 1y ago
        IDK your age group, but there's tons of nostalgic references and easter eggs. The game appears to be based on the Super Mario Odyssey engine, and the way Nintendo generally approaches the "all ages" problem is to make the plot easy, with tons of optional challenges and a set of hard levels in the "post-game" content. But I'm not very far in yet, so I can't promise anything on that front.

jldugger 1y ago

Well, the first paper on AI was published in 1943[1], but unfortunately requires some knowledge of brain science that was ultimately incorrect to understand, and includes some mathematical notation that is relatively obscure for high school. And this, Turing, and Shannon predate the Macy conferences / Ratio Club by a few years, so I'm not sure I'd call cybernetics a predecessor so much as a defunct offshoot. Though where math stops and CS starts varies depending on the scholar you talk to. [1]: https://jontalle.web.engr.illinois.edu/uploads/410-NS.F22/McCulloch-Pitts-1943-neural-networks-ocr.pdf https://jontalle.web.engr.illinois.edu/uploads/410-NS.F22/Mc...