20 things to do: различия между версиями

Материал из Поле цифровой дидактики
 
(не показано 8 промежуточных версий этого же участника)
Строка 1: Строка 1:
{{Book
{{Book
|Description=The concept of "Twenty Things to Do with a Computer," originally presented by Seymour Papert and Cynthia Solomon in 1971, highlights various engaging activities that can be accomplished through programming.
|Description=The concept of "Twenty Things to Do with a Computer," originally presented by Seymour Papert and Cynthia Solomon in 1971, highlights various engaging activities that can be accomplished through programming.
# Make a Turtle
# Program the Turtle to Draw a Man
# Turtle Biology
# Make a Display Turtle
# Play Spacewar
# Differential Geometry
# Draw Spirals
# Have a Heart (and learn to DEBUG)
# Growf lowers
# Make a Movie
# Make A Music Box and Program A Tune
# Play with Semi-Random Musical Effects and then Try Serious  Composing
# Computerize an Erector Set Crane and Build a Tower of Blocks
# Make a Super Light Show
# Write Concrete Poetry
# Try C.A.I. (Computer-Assisted Instruction) and Psychology
# Physics in the Finger-Tips
#  Explain Yourself
# Puppets
# Recursion Line
|Field_of_knowledge=Информатика, Педагогика
|Field_of_knowledge=Информатика, Педагогика
|launch year=1971
|launch year=1971
|Inventor=Papert, Solomon
|Inventor=Papert; Solomon
|Environment=Logo
|Environment=Logo
}}
}}
== 1971 ==
# Make a Turtle
# Program the Turtle to Draw a Man
# Turtle Biology
# Make a Display Turtle
# Play Spacewar
# Differential Geometry
# Draw Spirals
# Have a Heart (and learn to DEBUG)
# Growf lowers
# Make a Movie
# Make A Music Box and Program A Tune
# Play with Semi-Random Musical Effects and then Try Serious  Composing
# Computerize an Erector Set Crane and Build a Tower of Blocks
# Make a Super Light Show
# Write Concrete Poetry
# Try C.A.I. (Computer-Assisted Instruction) and Psychology
# Physics in the Finger-Tips
#  Explain Yourself
# Puppets
# Recursion Line
== 2024 ==
== 2024 ==


Строка 11: Строка 53:
# '''Face Filter''' -  Young people interact with AI/ML powered face filters when using social media (SnapChat, TikTok) with their friends. Creating filters can involve them in learning about AI/ML in two ways: exploring how to recognize facial features and body parts and using generative models to create images to transform what is captured or superposed over the camera feed.
# '''Face Filter''' -  Young people interact with AI/ML powered face filters when using social media (SnapChat, TikTok) with their friends. Creating filters can involve them in learning about AI/ML in two ways: exploring how to recognize facial features and body parts and using generative models to create images to transform what is captured or superposed over the camera feed.
# '''Weird Recipes''' -  Tseng and colleagues (2023) show how youth can use classifiers to build an application that provides recipe recommendations. Using libraries such as nanoGPT, could enable learners to play with language models, modify training data sets, and change parameters to create recipes.
# '''Weird Recipes''' -  Tseng and colleagues (2023) show how youth can use classifiers to build an application that provides recipe recommendations. Using libraries such as nanoGPT, could enable learners to play with language models, modify training data sets, and change parameters to create recipes.
# 4 Dance Game Dance  
# '''Dance Game Dance''' in AI/ML education has been growing over the past few years with projects that have learners partner to dance with AI agents (Long et al., 2020) and have learners create choreographies and animations (Castro et al., 2022; Jordan et al., 2021. Building on these efforts, children could design their own dance games similar to Just Dance, where they train models to recognize players’ moves (sequences of poses) using either pose recognition or data from wearable sensors and compare these to predetermined choreographies or where based on a set of basic moves the system generates possible choreographies that match music rhythms.
in AI/ML education has been growing over the past few years with projects that have learners partner to dance with AI agents (Long et al., 2020) and have learners create choreographies and animations (Castro et al., 2022; Jordan et al., 2021. Building on these efforts, children could design their own dance games similar to Just Dance, where they train models to recognize players’ moves (sequences of poses) using either pose recognition or data from wearable sensors and compare these to predetermined choreographies or where based on a set of basic moves the system generates possible choreographies that match music rhythms.
# '''Poetry Generator''' Papert and Solomon (#15, 1971) explored how creating sentence generators may support learners to better understand language structures and computing. With current off-the-shelf AI/ML libraries, learners can generate poetry while exploring more sophisticated natural language processing techniques. From simple Markov chains to training models using nanoGPT, youth can create synthetic text including poetry, short stories, and music lyrics.
# 5 Poetry Generator  
# '''Sports Training App''' - Similarly, to #4, there’s a lot of promise for young people to learn about AI/ML while engaging with sports (Jordan et al., 2021; Kumar & Worsley, 2023; Zimmermann-Niefield; 2019).
Papert and Solomon (#15, 1971) explored how creating sentence generators may support learners to better understand language structures and computing. With current off-the-shelf AI/ML libraries, learners can generate poetry while exploring more sophisticated natural language processing techniques. From simple Markov chains to training models using nanoGPT, youth can create synthetic text including poetry, short stories, and music lyrics.
# '''Music Generator''' Papert & Solomon (# 11, 1971) suggested creating a music box with programmed tunes. Later in #12, they present a project that creates random songs. Today, children could build data sets of tunes and use them to train a model that generates music, these “compositions” could follow the traditions of aleatoric and minimalist music or even explore noise music. Here, like in #2 learners could engage in discussions about creativity in humans and machines and copyrigh
# 6 Sports Training App
# '''Artificial Creatures and Systems for Natural Habitats'''  Building on advances in embodied cognition, Eisenberg and colleagues propose that children could “create free-standing simple creatures whose job is to be placed in a terrarium or aquarium alongside living creatures and to interact with their (biological) environment” (Eisenberg et al. 2017 p.3) to reflect on the embodied nature of their own thinking. Such projects would involve building AI/ML models for the systems to recognize environmental behaviors and react to them.
Similarly, to #4, there’s a lot of promise for young people to learn about AI/ML while engaging with sports (Jordan et al., 2021; Kumar & Worsley, 2023; Zimmermann-Niefield; 2019).
# '''Game with a Game Player Making games''' has always been a rich context for engaging with computing in constructionist ways. With AI/ML children cannot only create games to play with each other (#5 Papert & Solomon, 1971), but also train models that when incorporated into the games can enable them and their peers to play against the machine. Examples include creating, training, and then using neural networks to play TicTacToe as Kahn and Winter (2021) suggested in half-baked AI projects or by designing and playing a Rock-Paper-Scissors game with an application that recognizes speech and hand gestures (Kahn & Winter, 2018)
# 7 Music Generator  
# '''Explain Yourself'''  In #18, Solomon and Papert (1971) invite learners to explain themselves asking “How good of a model could you make of a person?.” A similar thing could be done with AI/ML, inviting learners to think about how the qualities of synthetic text differ from those of text written by humans, how good of a co-learner an AI/ML system can be, or how good of a player an AI agent can be. The explain yourself activity can serve as a space for learners to think about the differences between human and machine learning, to explore the limitations of both humans and machines and to think about their own thinking.
Papert & Solomon (# 11, 1971) suggested creating a music box with programmed tunes. Later in #12, they present a project that creates random songs. Today, children could build data sets of tunes and use them to train a model that generates music, these “compositions” could follow the traditions of aleatoric and minimalist music or even explore noise music. Here, like in #2 learners could engage in discussions about creativity in humans and machines and copyrigh
# '''Drawing Generator Learners''' - could create drawing generators that make images in comic or anime styles. Here they could use existing public image large datasets that they can modify and or personalize. While creating an image generator (using off-the-shelf libraries) learners could also reflect on the environmental impact of generating synthetic images (see Luccioni et al., 2023) and what are the ethics of this in the context of climate change. Similarly, this activity could also involve learners in discusions related to how synthetic images may contribute to misinformation.
# 8 Artificial Creatures and Systems for Natural Habitats  
# '''Adaptable Interactive Stories'''  At Fablearn/Constructionism 2023, Rita Freudenberg suggested that children could “create interactive stories [or games], where the storyline not just depends on a deliberate action by the player, but the presented story options change based on previous decisions”. Rita explained learners would have to decide on what factors would influence the development of the narrative and how transparent these are to users.
Building on advances in embodied cognition, Eisenberg and colleagues propose that children could “create free-standing simple creatures whose job is to be placed in a terrarium or aquarium alongside living creatures and to interact with their (biological) environment” (Eisenberg et al. 2017 p.3) to reflect on the embodied nature of their own thinking. Such projects would involve building AI/ML models for the systems to recognize environmental behaviors and react to them.
# '''Artificial Tutors for Peers'''  Similar to #9, learners could create a system that interacts with junior students. Fablearn/Constructionism 2023 participant Daniel Agostini argued that this could support learners to "learn by teaching" an AI/ML system.
# 9 Game with a Game Player Making games has always been a rich context for engaging with computing in constructionist ways. With AI/ML children cannot only create games to play with each other (#5 Papert & Solomon, 1971), but also train models that when incorporated into the games can enable them and their peers to play against the machine. Examples include creating, training, and then using neural networks to play TicTacToe as Kahn and Winter (2021) suggested in half-baked AI projects or by designing and playing a Rock-Paper-Scissors game with an application that recognizes speech and hand gestures (Kahn & Winter, 2018)
#1 '''Modeling Climate and Carbon Emissions'''  Students could create models using publicly available climate data to recognize patterns and make predictions. This suggestion by an anonymous participant at Fablearn/Constructionism 2023, highlights how such an activity can engage learners in thinking about AI/ML in a highly relevant area of concern that is socio-technical and involves thinking about complexity.
# 10 Explain Yourself  
# '''Artificial Co-Learners''' - Eisenberg and colleagues also propose having learners create artificial co-learners. This, they describe, would involve having youth “design models of students (rather than teachers) that would accompany children in learning new material and articulate (perhaps simplistic, but concrete) strategies for thinking about the material.
In #18, Solomon and Papert (1971) invite learners to explain themselves asking “How good of a model could you make of a person?.” A similar thing could be done with AI/ML, inviting learners to think about how the qualities of synthetic text differ from those of text written by humans, how good of a co-learner an AI/ML system can be, or how good of a player an AI agent can be. The explain yourself activity can serve as a space for learners to think about the differences between human and machine learning, to explore the limitations of both humans and machines and to think about their own thinking.
# '''Role-Playing Games''' Students could create games based on text inputs. We build on this idea of Fablearn/Constructionism 2023 participant Ken Kahn by proposing that learners could create roleplaying games, for instance building datasets of role-playing game scripts (such as Dungeon and Dragons campaigns), training models (see #1) and then generating new games.
# 11 Drawing Generator Learners could create drawing generators that make images in comic or anime styles. Here they could use existing public image large datasets that they can modify and or personalize. While creating an image generator (using off-the-shelf libraries) learners could also reflect on the environmental impact of generating synthetic images (see Luccioni et al., 2023) and what are the ethics of this in the context of climate change. Similarly, this activity could also involve learners in discusions related to how synthetic images may contribute to misinformation.
# '''Personal Assistants''' During Fablearn/Constructionism 2023 Nicolás Acosta and Glenn Boustead suggested that learners could build personal assistants. Acosta and Boustead explained this could involve creating a voice assistant to control music, lights, and engage people in wellness activities (e.g., meditation, reflection).
# 12 Adaptable Interactive Stories  
# Tamagotchi Learners could create virtual pets that can speak and listen, imagining for example that someone might create a “tamagotchi with sass.”
At Fablearn/Constructionism 2023, Rita Freudenberg suggested that children could “create interactive stories [or games], where the storyline not just depends on a deliberate action by the player, but the presented story options change based on previous decisions”. Rita explained learners would have to decide on what factors would influence the development of the narrative and how transparent these are to users.
# '''Workout App''' During Fablearn/Cons
# 13 Artificial Tutors for Peers  
# '''Create a constructionist activities generator'''  - Create an application that uses AI/ML to generate more constructionist activities.
Similar to #9, learners could create a system that interacts with junior students. Fablearn/Constructionism 2023 participant Daniel Agostini argued that this could support learners to "learn by teaching" an AI/ML system.
#14 Modeling Climate and Carbon Emissions  
Students could create models using publicly available climate data to recognize patterns and make predictions. This suggestion by an anonymous participant at Fablearn/Constructionism 2023, highlights how such an activity can engage learners in thinking about AI/ML in a highly relevant area of concern that is socio-technical and involves thinking about complexity.
# 15 Artificial Co-Learners  
Eisenberg and colleagues also propose having learners create artificial co-learners. This, they describe, would involve having youth “design models of students (rather than teachers) that would accompany children in learning new material and articulate (perhaps simplistic, but concrete) strategies for thinking about the material.
#16 Role-Playing Games  
Students could create games based on text inputs. We build on this idea of Fablearn/Constructionism 2023 participant Ken Kahn by proposing that learners could create roleplaying games, for instance building datasets of role-playing game scripts (such as Dungeon and Dragons campaigns), training models (see #1) and then generating new games.
# 17 Personal Assistants  
During Fablearn/Constructionism 2023 Nicolás Acosta and Glenn Boustead suggested that learners could build personal assistants. Acosta and Boustead explained this could involve creating a voice assistant to control music, lights, and engage people in wellness activities (e.g., meditation, reflection).
#18 Tamagotchi Learners could create virtual pets that can speak and listen, imagining for example that someone might create a “tamagotchi with sass.”
# 19 Workout App During Fablearn/Cons
#20 Create a constructionist activities generator  
Create an application that uses AI/ML to generate more constructionist activities.]

Текущая версия на 08:16, 24 сентября 2024



Описание книги The concept of "Twenty Things to Do with a Computer," originally presented by Seymour Papert and Cynthia Solomon in 1971, highlights various engaging activities that can be accomplished through programming.
  1. Make a Turtle
  2. Program the Turtle to Draw a Man
  3. Turtle Biology
  4. Make a Display Turtle
  5. Play Spacewar
  6. Differential Geometry
  7. Draw Spirals
  8. Have a Heart (and learn to DEBUG)
  9. Growf lowers
  10. Make a Movie
  11. Make A Music Box and Program A Tune
  12. Play with Semi-Random Musical Effects and then Try Serious Composing
  13. Computerize an Erector Set Crane and Build a Tower of Blocks
  14. Make a Super Light Show
  15. Write Concrete Poetry
  16. Try C.A.I. (Computer-Assisted Instruction) and Psychology
  17. Physics in the Finger-Tips
  18. Explain Yourself
  19. Puppets
  20. Recursion Line
Область знаний Информатика, Педагогика
Год издания 1971
Веб-сайт где можно прочитать книгу или статью
Видео запись
Авторы Papert, Solomon
Среды и средства, на которые повлияла книга Logo

1971

  1. Make a Turtle
  2. Program the Turtle to Draw a Man
  3. Turtle Biology
  4. Make a Display Turtle
  5. Play Spacewar
  6. Differential Geometry
  7. Draw Spirals
  8. Have a Heart (and learn to DEBUG)
  9. Growf lowers
  10. Make a Movie
  11. Make A Music Box and Program A Tune
  12. Play with Semi-Random Musical Effects and then Try Serious Composing
  13. Computerize an Erector Set Crane and Build a Tower of Blocks
  14. Make a Super Light Show
  15. Write Concrete Poetry
  16. Try C.A.I. (Computer-Assisted Instruction) and Psychology
  17. Physics in the Finger-Tips
  18. Explain Yourself
  19. Puppets
  20. Recursion Line

2024

  1. Smart Puppets - With AI/ML we can extend this idea to create more interactive puppets that can recognize words and speech or movements. Older children and youth could have puppet recognize speech or words (from using simple classifiers to recognize keywords to more complex voice recognition libraries) and respond to them accordingly (with preset responses—à la Eliza or using synthetic text and text to speech libraries).
  2. Face Filter - Young people interact with AI/ML powered face filters when using social media (SnapChat, TikTok) with their friends. Creating filters can involve them in learning about AI/ML in two ways: exploring how to recognize facial features and body parts and using generative models to create images to transform what is captured or superposed over the camera feed.
  3. Weird Recipes - Tseng and colleagues (2023) show how youth can use classifiers to build an application that provides recipe recommendations. Using libraries such as nanoGPT, could enable learners to play with language models, modify training data sets, and change parameters to create recipes.
  4. Dance Game Dance in AI/ML education has been growing over the past few years with projects that have learners partner to dance with AI agents (Long et al., 2020) and have learners create choreographies and animations (Castro et al., 2022; Jordan et al., 2021. Building on these efforts, children could design their own dance games similar to Just Dance, where they train models to recognize players’ moves (sequences of poses) using either pose recognition or data from wearable sensors and compare these to predetermined choreographies or where based on a set of basic moves the system generates possible choreographies that match music rhythms.
  5. Poetry Generator Papert and Solomon (#15, 1971) explored how creating sentence generators may support learners to better understand language structures and computing. With current off-the-shelf AI/ML libraries, learners can generate poetry while exploring more sophisticated natural language processing techniques. From simple Markov chains to training models using nanoGPT, youth can create synthetic text including poetry, short stories, and music lyrics.
  6. Sports Training App - Similarly, to #4, there’s a lot of promise for young people to learn about AI/ML while engaging with sports (Jordan et al., 2021; Kumar & Worsley, 2023; Zimmermann-Niefield; 2019).
  7. Music Generator Papert & Solomon (# 11, 1971) suggested creating a music box with programmed tunes. Later in #12, they present a project that creates random songs. Today, children could build data sets of tunes and use them to train a model that generates music, these “compositions” could follow the traditions of aleatoric and minimalist music or even explore noise music. Here, like in #2 learners could engage in discussions about creativity in humans and machines and copyrigh
  8. Artificial Creatures and Systems for Natural Habitats Building on advances in embodied cognition, Eisenberg and colleagues propose that children could “create free-standing simple creatures whose job is to be placed in a terrarium or aquarium alongside living creatures and to interact with their (biological) environment” (Eisenberg et al. 2017 p.3) to reflect on the embodied nature of their own thinking. Such projects would involve building AI/ML models for the systems to recognize environmental behaviors and react to them.
  9. Game with a Game Player Making games has always been a rich context for engaging with computing in constructionist ways. With AI/ML children cannot only create games to play with each other (#5 Papert & Solomon, 1971), but also train models that when incorporated into the games can enable them and their peers to play against the machine. Examples include creating, training, and then using neural networks to play TicTacToe as Kahn and Winter (2021) suggested in half-baked AI projects or by designing and playing a Rock-Paper-Scissors game with an application that recognizes speech and hand gestures (Kahn & Winter, 2018)
  10. Explain Yourself In #18, Solomon and Papert (1971) invite learners to explain themselves asking “How good of a model could you make of a person?.” A similar thing could be done with AI/ML, inviting learners to think about how the qualities of synthetic text differ from those of text written by humans, how good of a co-learner an AI/ML system can be, or how good of a player an AI agent can be. The explain yourself activity can serve as a space for learners to think about the differences between human and machine learning, to explore the limitations of both humans and machines and to think about their own thinking.
  11. Drawing Generator Learners - could create drawing generators that make images in comic or anime styles. Here they could use existing public image large datasets that they can modify and or personalize. While creating an image generator (using off-the-shelf libraries) learners could also reflect on the environmental impact of generating synthetic images (see Luccioni et al., 2023) and what are the ethics of this in the context of climate change. Similarly, this activity could also involve learners in discusions related to how synthetic images may contribute to misinformation.
  12. Adaptable Interactive Stories At Fablearn/Constructionism 2023, Rita Freudenberg suggested that children could “create interactive stories [or games], where the storyline not just depends on a deliberate action by the player, but the presented story options change based on previous decisions”. Rita explained learners would have to decide on what factors would influence the development of the narrative and how transparent these are to users.
  13. Artificial Tutors for Peers Similar to #9, learners could create a system that interacts with junior students. Fablearn/Constructionism 2023 participant Daniel Agostini argued that this could support learners to "learn by teaching" an AI/ML system.
  14. 1 Modeling Climate and Carbon Emissions Students could create models using publicly available climate data to recognize patterns and make predictions. This suggestion by an anonymous participant at Fablearn/Constructionism 2023, highlights how such an activity can engage learners in thinking about AI/ML in a highly relevant area of concern that is socio-technical and involves thinking about complexity.
  15. Artificial Co-Learners - Eisenberg and colleagues also propose having learners create artificial co-learners. This, they describe, would involve having youth “design models of students (rather than teachers) that would accompany children in learning new material and articulate (perhaps simplistic, but concrete) strategies for thinking about the material.
  16. Role-Playing Games Students could create games based on text inputs. We build on this idea of Fablearn/Constructionism 2023 participant Ken Kahn by proposing that learners could create roleplaying games, for instance building datasets of role-playing game scripts (such as Dungeon and Dragons campaigns), training models (see #1) and then generating new games.
  17. Personal Assistants During Fablearn/Constructionism 2023 Nicolás Acosta and Glenn Boustead suggested that learners could build personal assistants. Acosta and Boustead explained this could involve creating a voice assistant to control music, lights, and engage people in wellness activities (e.g., meditation, reflection).
  18. Tamagotchi Learners could create virtual pets that can speak and listen, imagining for example that someone might create a “tamagotchi with sass.”
  19. Workout App During Fablearn/Cons
  20. Create a constructionist activities generator - Create an application that uses AI/ML to generate more constructionist activities.