On Feb 11, 2025, I asked a large-language model trained by Google – Gemini Advanced 1.5 Pro with Deep Research – what skills, knowledge, and attitudes children need to operate successfully as adults in a future world where life is integrated with artificial intelligence. The response is provided here below. The only change I made to the response was that I removed a section on using AI in education.  

What is the value of this response?  

1. Using AI. This response highlights the current use of an easily accessible AI model to summarize and compile information in a way that has not been formerly possible. Within a few minutes, I was able to get a sense of results possible from scouring the internet, and a starting point for deeper understanding and further research by looking into the references.
2. Focusing attention on future skills (and knowledge and attitudes). There are many lists available to direct parents to what may be most important for their children to learn. This response from Gemini helps to consolidate some of those ideas. Of course, this is not the final word on the subject, but it is a helpful bunch of words.
3. Considering Volo. It is helpful to evaluate practices by considering from different angles. Some of the strengths of Volo can be seen in this summary, which is to be expected because we considered research and the current state of knowledge in education and society while developing Volo. For example, the McKinsey report on future skills is one that helped to inform Volo’s development. As we have been moving forward with Volo, it has been exciting to see how new research and thinking supports the principles and practices at Volo. For example, the article by Yiu, Kosoy, and Gopnik (2024) recently came through my info feed; it reinforces our choices at Volo. Of course, as new information and insights develop, and as our learning community develops and expresses its needs and possibilities, Volo will continue to adapt to keep pace.

Details

My prompt:

summarize recent research on the skills, knowledge, and attitudes children need to operate successfully as adults in a future world where life is integrated with artificial intelligence. Indicate which skills, knowledge, attitudes should best be learned at which age. Indicate also research-supported suggetions for the most effective pedagogy to help young people learn what they need to know for success as adults. Define success. Include articles that cite this one: “Transmission Versus Truth, Imitation Versus Innovation: What Children Can Do That Large Language and Language-and-Vision Models Cannot (Yet)” Include all research articles. Include full source citations. Indicate if you have access to the full text or a summary or simply a secondary reference. If you cite a source from a summary, indicate how reliable the summary is and cite the full source. If citing as a secondary reference, provide the full source reference.

Gemini Advanced Response:

Preparing Children for a Future With AI: A Research Synthesis

This report summarizes recent research on the skills, knowledge, and attitudes children need to thrive in a future characterized by artificial intelligence (AI) integration. It analyzes research articles citing “Transmission Versus Truth, Imitation Versus Innovation: What Children Can Do That Large Language and Language-and-Vision Models Cannot (Yet)” by Yiu, Kosoy, and Gopnik (2024) to identify common themes and areas of agreement or disagreement across studies. Based on the analysis, the report proposes a comprehensive framework of skills, knowledge, and attitudes, and pedagogical approaches for equipping children for success in an AI-driven world.

Key Findings from Yiu, Kosoy, and Gopnik (2024)

Yiu, Kosoy, and Gopnik (2024) explore the fundamental differences between human cognition and the capabilities of large language and language-and-vision models (LLMs). Their research highlights the unique capacity of children for innovation, a critical skill that current AI systems lack. While LLMs excel at imitating and transmitting existing information, they struggle to generate novel solutions or discover new knowledge through exploration and experimentation. This distinction has significant implications for how we prepare children for a future where AI will play an increasingly prominent role.

One key finding is that causal representations are formed through truth-seeking epistemic processes, which involve actively exploring the world and testing hypotheses¹. In contrast, LLMs primarily rely on statistical learning from massive datasets, which limits their ability to understand causal relationships and generate truly innovative solutions.

The study involved 42 children (ages 3 to 7) and 30 adults who were presented with descriptions of everyday objects. In the first part, participants were asked to identify which objects would “go best” together. In the second part, they were asked to innovate on the expected use of everyday objects to solve problems. When given the same text descriptions, LLMs performed similarly to humans on the imitation task, but were far less accurate on the innovation task, with accuracy rates ranging from 8% to 75%². Children were able to discover how a new machine worked just by experimenting and exploring, but the language models could not.

The authors argue that AI’s reliance on statistically predicting linguistic patterns is insufficient for discovering new information about the world. They suggest that taking inspiration from children’s curious and intrinsically motivated approach to learning could help researchers design new AI systems that are better prepared to explore the real world¹.

Furthermore, Yiu, Kosoy, and Gopnik (2024) draw an analogy between LLMs and earlier cultural technologies like writing, print, libraries, the internet, and even language itself¹. These technologies have revolutionized the way humans access and transmit information, but they do not replace the need for human cognition and innovation. LLMs, like these earlier technologies, can be powerful tools for cultural transmission and knowledge sharing, but they are not a substitute for human intelligence.

Overimitation and its Downsides

The study also discusses the concept of “overimitation,” which refers to the tendency of children to imitate actions even when they seem unnecessary or irrelevant¹. While overimitation can be beneficial for cultural learning and social cohesion, it can also hinder innovation by leading children to blindly follow established norms and procedures. In the context of an AI-driven world, it is important to encourage children to question assumptions, think critically, and develop their own unique solutions.

Interpolative vs. Extrapolative Generalization

Yiu, Kosoy, and Gopnik (2024) distinguish between two types of generalization: interpolative and extrapolative¹. Interpolative generalization involves applying existing knowledge and skills to new situations within a familiar domain. This is what LLMs excel at, as they can identify patterns and make predictions based on the vast amounts of data they have been trained on. Extrapolative generalization, on the other hand, involves going beyond existing knowledge and generating truly novel ideas and solutions. This is where human cognition, particularly in children, surpasses current AI systems.

AI and Inverse Problems

The authors also highlight the potential of AI systems to solve “inverse problems,” which involve inferring the underlying causes of observed events¹. This capability could have significant implications for various fields, such as scientific discovery, medical diagnosis, and engineering design. However, it is important to note that AI systems still require human guidance and interpretation to ensure that their solutions are meaningful and relevant to real-world problems.

Analysis of Citing Articles

This section synthesizes the findings from various articles that cite Yiu, Kosoy, and Gopnik (2024), identifying common themes and areas of disagreement. These articles provide further insights into the skills, knowledge, and attitudes needed for success in a future with AI⁴.

Common Themes

A recurring theme across the articles is the need for children to develop strong critical thinking and problem-solving skills⁴. This includes the ability to analyze information, evaluate different perspectives, and identify creative solutions. In an AI-driven world, where information is readily available, the ability to discern credible sources and think critically becomes even more crucial.

Another common theme is the importance of digital literacy and data fluency⁴. Children need to understand how AI systems work, how to interact with them effectively, and how to interpret the data they generate. This includes developing skills in data analysis, information evaluation, and online safety.

Creativity and innovation are also highlighted as essential skills⁴. While AI can automate repetitive tasks, it struggles with tasks that require originality and imagination. Encouraging children to explore their creative potential and think outside the box will be crucial for their future success.

A report from consulting firm McKinsey’s research arm identified 56 foundational skills that will help people thrive in the future of work⁸. Eleven of these skills are related to digital fluency and citizenship, software use and development, and understanding digital systems. The other 44 skills have nothing to do with digital knowledge or capabilities, highlighting the importance of a broad range of cognitive, interpersonal, and self-leadership skills.

In addition to the skills emphasized in the McKinsey report, it is important to recognize the value of art, languages, and spatial awareness in an AI-driven world⁸. These areas foster creativity, communication, and problem-solving abilities that are essential for navigating complex situations and adapting to new challenges.

Areas of Disagreement

While there is general agreement on the importance of cognitive and digital skills, there are some disagreements regarding the potential risks of AI in education. Some articles express concerns about the digital divide and the potential for AI to exacerbate existing inequalities⁶. They argue that AI tools may be implemented unevenly, favoring students in affluent schools and further disadvantaging those in low-income schools.

Another concern is the potential for AI to reduce access to learning opportunities that incorporate cognitive science principles⁶. Some researchers worry that AI tools may be used to promote passive forms of learning, such as drilling and rote memorization, rather than active, inquiry-based learning.

There are also concerns about the potential for AI misuse in education, such as plagiarism and cheating⁷. As AI tools become more sophisticated, it will be important to develop strategies to prevent academic dishonesty and ensure that students are using AI responsibly.

One specific concern is the potential for AI to perpetuate human biases and inaccuracies⁹. Since AI systems are trained on data created by humans, they can inherit and amplify existing biases, leading to unfair or discriminatory outcomes. It is crucial to be aware of this potential and to develop strategies to mitigate bias in AI systems and ensure that they are used ethically and responsibly.

Teachers are becoming increasingly aware of the characteristics of AI-generated text, such as a lack of detail, made-up facts, and excessive repetition⁹. This awareness can help educators identify potential misuse of AI and guide students towards responsible use of these tools.

Defining Success in an AI-Driven World

Success in an AI-driven world goes beyond traditional measures of academic achievement. It encompasses a broader set of skills, knowledge, and attitudes that enable individuals to thrive in a rapidly changing environment. This includes:

• Adaptability and resilience: The ability to adapt to new technologies and challenges, and to bounce back from setbacks. This is essential because AI is constantly evolving, and individuals need to be able to learn and adapt quickly to keep pace with these changes.
• Critical thinking and problem-solving: The ability to analyze information, evaluate evidence, and make sound judgments. This is crucial for navigating the abundance of information available in an AI-driven world and for making informed decisions.
• Creativity and innovation: The ability to generate new ideas and solutions, and to think outside the box. This is essential for solving complex problems and for developing new applications of AI technology.
• Digital literacy and data fluency: The ability to understand and use technology effectively, and to interpret and analyze data. This is fundamental for interacting with AI systems and for understanding the implications of AI-generated information.
• Collaboration and communication: The ability to work effectively with others, including AI systems. This is crucial for collaborating on projects, sharing knowledge, and building relationships in a world where AI is increasingly integrated into various aspects of life.
• Social-emotional intelligence: The ability to understand and manage one’s own emotions, and to empathize with others. This is essential for building strong relationships, navigating social situations, and understanding the human impact of AI technologies.
• Ethical awareness: The ability to consider the ethical implications of AI and to use technology responsibly. This is crucial for ensuring that AI is used for good and that its potential benefits are realized while mitigating its risks.

Skills, Knowledge, and Attitudes by Age Group

While all the skills mentioned above are important, certain skills may be more relevant or easier to develop at different ages¹⁰. Here’s a suggested framework for developing skills, knowledge, and attitudes by age group:

It is important to note that this is a general framework, and the specific skills and pedagogical approaches may need to be adapted based on the individual needs and interests of each child.

The Importance of Play

Play is essential for children’s development, especially in early childhood¹¹. Through play, children learn to explore their environment, experiment with new ideas, and develop social-emotional skills. In an AI-driven world, play can also help children develop the creativity, imagination, and problem-solving skills that will be essential for their future success.

Types of AI and Their Applications

AI encompasses a wide range of technologies with various applications¹². Some common types of AI include:

• Machine learning: This involves training algorithms to learn from data and improve their performance over time. Machine learning is used in various applications, such as image recognition, natural language processing, and personalized recommendations.
• Computer vision: This involves enabling computers to “see” and interpret images and videos. Computer vision is used in applications such as facial recognition, object detection, and self-driving cars.
• Natural language processing: This involves enabling computers to understand and process human language. Natural language processing is used in applications such as chatbots, language translation, and sentiment analysis.
• Robotics: This involves combining AI with robots to create intelligent machines that can perform tasks autonomously. Robotics is used in various applications, such as manufacturing, healthcare, and exploration.

Understanding the different types of AI and their applications can help children develop a broader perspective on the potential of AI and its impact on various aspects of life.

Conclusion

Preparing children for a future with AI requires a multifaceted approach that goes beyond simply teaching them how to use technology. It involves fostering a love of learning, encouraging curiosity and exploration, and developing a range of cognitive, interpersonal, and digital skills. By equipping children with the necessary tools and mindset, we can empower them to navigate the complexities of an AI-driven world and contribute to a future where technology serves humanity.

The research synthesized in this report highlights the importance of fostering children’s unique capacity for innovation, a skill that current AI systems lack. While AI can be a powerful tool for learning and knowledge transmission, it is crucial to ensure that it is used in a way that complements, not replaces, human intelligence and creativity. By promoting active learning, ethical awareness, and a holistic approach to skill development, educators can help children thrive in an AI-driven world and become responsible and innovative contributors to society.

Source Citations

Works cited

1. Transmission Versus Truth, Imitation Versus Innovation: What …, accessed February 11, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC11373165/
2. Artificial Intelligence Systems Excel at Imitation, but Not Innovation …, accessed February 11, 2025, https://www.psychologicalscience.org/news/2023-december-ai-systems-imitation.html
3. Artificial intelligence systems excel at imitation, but not innovation …, accessed February 11, 2025, https://www.sciencedaily.com/releases/2023/12/231212131041.htm
4. The Essential Skills for Children in the Age of AI – The Center For Child Behavioral Health, accessed February 11, 2025, https://www.childbehavioralhealth.com/post/the-essential-skills-for-children-in-the-age-of-ai
5. How your kids can use the power of AI without compromising learning, accessed February 11, 2025, https://www.nordangliaeducation.com/wps-florida/news/2024/09/27/how-your-kids-can-use-the-power-of-ai-without-compromising-learning
6. Using Learning Science To Analyze the Risks and Benefits of AI in K-12 Education, accessed February 11, 2025, https://www.americanprogress.org/article/using-learning-science-to-analyze-the-risks-and-benefits-of-ai-in-k-12-education/
7. 7 Examples of AI Misuse in Education – Inspera, accessed February 11, 2025, https://www.inspera.com/ai/examples-of-ai-misuse-in-education/
8. 219: The skills your child will need in the age of AI | Your Parenting Mojo, accessed February 11, 2025, https://yourparentingmojo.com/captivate-podcast/ai/
9. AI in the Classroom – Helping or Hurting? – Washington FAMILY Magazine, accessed February 11, 2025, https://www.washingtonfamily.com/ai-in-classrooms/
10. Americans and Lifetime Learning in the Knowledge Age | The Pew Charitable Trusts, accessed February 11, 2025, https://www.pewtrusts.org/en/trend/archive/spring-2020/americans-and-lifetime-learning-in-the-knowledge-age
11. The Ultimate Guide to AI in Education: Benefits, Challenges, & Real-World Uses, accessed February 11, 2025, https://www.powerschool.com/blog/ai-in-education/
12. The Thrilling World of AI for Kids 2025: Igniting Brilliance – JetLearn, accessed February 11, 2025, https://www.jetlearn.com/blog/power-of-ai-for-kids