The Stanford AI Index Report '24, Advocating for AI in Education, and implementing AI programs.
Credit : Stanford University

The Stanford AI Index Report '24, Advocating for AI in Education, and implementing AI programs.


Before you read:

This piece is first of my Technology, Innovation, Policy, & Society (T.I.P.S) Review biweekly series where I try to make sense of the current trends at the intersection of technology and policy, using my decade of progressive experience in public policy, innovation programs, and federal project management.

This article has six sections, starting with an introduction followed by key insights of the report. Opinions on potential implication of AI practice on education, policy, and society are briefly shared and backed by evidence, data, and insights from the report and some of my personal experiences while working on AI curriculum for school students & National Strategy for AI in India. The article concludes with the role of AI education policy advocacy and thought leadership within the US in this AI renaissance.

The purpose of this article is purely educational, to co-learn and hold space for dialogues. The writer does not intend to infringe upon any copyrights. Views expressed are personal.


1. Navigating the AI Landscape with the 2024 AI Index Report

On 30th May, 2024, I attended Code.org's TeachAI Education X Industry learning series organized by Aleksandar Lazovski ??, Arianne Lapidus, Pam Vimonmas Vachatimanont, and Pat Yongpradit.

During this event, Nestor Maslej from Stanford Institute for Human-Centered Artificial Intelligence (HAI) summarized the report with some top takeaways:

Credit: Stanford University

AI is a transformative force reshaping industries, influencing public policy, and fundamentally altering societal structures globally. The 2024 AI Index Report by Stanford University provides a comprehensive analysis of the current state of AI, highlighting significant trends, achievements, and challenges.

This article intents to distill key insights from the report, examine their implications for education, policy, and society, and reflect on the Indian context alongside my personal experiences with AI initiatives in India. The core value is - Technology is to make society better and the aim is to identify advantages, shortcomings, and opportunities for techno-crats of the nation.


2. Key Insights:

Technical Progress

The number of newly released foundation models more than doubled in 2023 compared to 2022, with a marked shift towards open-source models. Approximately 65.7% of these models were open-source, reflecting a significant increase from previous years. Despite this, closed-source models continue to outperform open-source ones on various benchmarks, with a median performance advantage of 24.2%.

The technical progress in AI is highlighted by the achievements of state-of-the-art models like GPT-4, Gemini, and Claude 3. These models are impressively multimodal, capable of generating fluent text in multiple languages, processing audio, and even explaining visual content like memes. For instance, GPT-4 achieved a 0.97 mean win rate score on the comprehensive Holistic Evaluation of Language Models (HELM) benchmark, which includes the Massive Multitask Language Understanding (MMLU) among other evaluations.

Industry Trends

The report underscores the dominance of industry players in AI development. Companies like Google and OpenAI are at the forefront, with Google’s Gemini Ultra and OpenAI’s GPT-4 costing $191 million and $78 million, respectively, to train. This highlights the escalating costs associated with developing large-scale AI models. While global private investment in AI decreased for the second consecutive year, investment in generative AI surged, attracting $25.2 billion in 2023 alone. This surge is often attributed to the "ChatGPT effect," which has sparked widespread interest and investment in generative AI technologies.

Geopolitical Dynamics

The United States remains the leader in developing notable machine learning models, producing 61 models in 2023. This leadership extends to AI private investment, where the U.S. saw $67.2 billion in investments in 2023, vastly outpacing other countries. The report also underscores the growing importance of AI in geopolitical strategies and economic competitiveness. China and the UK follow the U.S. in AI model development and investment, but there is a considerable gap between these countries and the U.S.

For example, India has seen an amazing rate of growth, but the system needs more support and resources.


3. India's current position

In this section, we are going to explore India's current position with the tension between two pairs of contrasting observations or competing ideas:

  1. The disconnect between increasing AI skills in India's workforce and the slow growth in public and private investments.
  2. The growing skepticism towards AI in India, yet the unprecedented growth of awareness and use of OpenAI 's ChatGPT.



Over the past decade, India has made significant strides in the AI landscape, positioning itself as a key player on the global stage. The country has seen a substantial increase in AI-related research publications, with Indian institutions contributing to numerous groundbreaking studies. Action-oriented policy and strategic implementation have positioned Indians as leaders in AI technology. Two examples at the intersection of policy and education are shared in Chapter 4 of this article.

India commands the workforce with the highest reported relative AI skill penetration rates, as per the Stanford AI Index Report '24.

Credit: Stanford University

Several major economies registered substantial increases in their AI talent pools. The countries showing the greatest increases are India (263%), Cyprus (229%), and Denmark (213%). India has been actively participating in international AI forums, underscoring its commitment to shaping global AI policies and standards

India's growing investment in AI startups, supported by a robust ecosystem of tech incubators and accelerators, which is crucial for nurturing homegrown AI talent and innovations. However, the infrastructure Government of India (GoI) provides is not enough. For e.g, the total budget of Atal Innovation Mission Official is Rs. 2000 crores (approx. US$250M) and the government has earmarked a budget of approx US$1.25 billion. India, like the United States, needs private investments in AI.

When aggregating private AI investments since 2013, the United States leads with $335.2 billion invested, followed by China with $103.7 billion, and the United Kingdom at $22.3 billion. India is barely hitting $10 billion.

Credits: Stanford University

Organisations like nasscom, INDIAai, etc. are trying their best to create consortium and groups and work closely with relevant government departments to inform policy design as per rapidly evolving industry trends. However, the number of people who understand both technology and policy are quite too few.

Skepticism v/s Use

India is also one of the very few countries where we have seen an interesting trend of decreasing trust on AI. Fewer people think that products and services using AI have more benefits than drawbacks.

Credits: Stanford University

One potential reason for this phenomenon might be the excessive use of AI in political campaigns and deepfakes. Over the past few months, India was undergoing its election cycle. Numerous deep fake videos and AI-generated WhatsApp messages were used by politicians from all parties, making AI potentially more harmful.

This has not stopped common Indian citizens from educating themselves about AI or using ChatGPT. Access to data and the internet in India is arguably the cheapest globally. Access to affordable smartphones makes information dissemination more equitable.

As per the Global Public Opinion on Artificial Intelligence (GPO-AI) report by the Schwartz Reisman Institute for Technology and Society (SRI) in collaboration with the Policy, Elections and Representation Lab (PEARL) at the Munk School of Global Affairs and Public Policy at the University of Toronto, . In October and November 2023, researchers from SRI and PEARL conducted a 21-country survey examining global attitudes toward AI. Figure 9.1.9 explores the extent of global public awareness of ChatGPT. Among global respondents, 63% claim awareness of ChatGPT. Countries with the highest awareness rates include India (82%), Kenya (81%), Indonesia (76%), and Pakistan (76%). Poland reported the lowest awareness, at 43%. Figure 9.1.10 highlights how frequently respondents who report being familiar with ChatGPT use the tool. Globally, 17% of users utilize it daily, 36% weekly, and 16% monthly. India (36%), Pakistan (28%), and Kenya (27%) report the highest levels of daily usage.

Credits: Stanford University
Credits: Stanford University

4. Opinion: Implications for Education, Policy, and Society

Education

AI is revolutionizing education by enhancing learning tools, personalizing educational experiences, and automating administrative tasks. Advanced AI models like GPT-4 are being used to develop intelligent tutoring systems, create interactive learning environments, and provide real-time feedback to students. These technologies can cater to individual learning styles and paces, thereby improving educational outcomes. For example, AI-powered platforms can analyze students' strengths and weaknesses and provide tailored resources to address specific learning needs.

Policy

Governments worldwide are increasingly focused on AI policy and governance. The 2024 AI Index Report highlights a surge in AI-related regulations, with policymakers aiming to balance innovation with ethical considerations. Key areas of focus include privacy, transparency, fairness, and the prevention of misuse. The introduction of AI-specific regulations reflects a recognition of AI's profound impact on society and the need to mitigate associated risks. For instance, the report notes that U.S. regulators passed more AI-related regulations in 2023 than ever before, addressing issues such as data privacy, algorithmic transparency, and ethical AI deployment.

Society

AI's integration into various aspects of daily life has led to significant societal impacts. Public opinion reflects growing concerns about AI’s influence on jobs, privacy, and ethical issues. Younger generations, in particular, are apprehensive about AI’s potential to disrupt the job market and societal norms. According to the report, 66% of Gen Z respondents believe AI will significantly affect their current jobs, compared to 46% of baby boomer respondents. Additionally, there is a growing awareness of the ethical implications of AI, such as bias in algorithms and the use of AI in surveillance.


5. Personal Experiences at the intersection of AI Education Policy implementation, program ideation, and project management

In 2015, NITI Aayog came into existence followed by an influx of technology-led public policy interventions, one of them notably is India's National Strategy for Artificial Intelligence in 2018. #GenderEquity and #Education became one of the key priority areas for AI penetration in India.

National Strategy for Artificial Intelligence

The Artificial Intelligence Education Strategy in India

In January 2019, nasscom foundation and AIM formalized a partnership to bring AI curriculum to schools in India, by leveraging the IT industry of the nation. A stellar example of not just policy making, but also program implementation deserves a chapter of its own.

A key objective of the AI Base and Step Up modules is to prepare the young minds of India to harness AI's potential, which is projected to significantly boost the country's GDP by 1.3% annually through the use of machine learning and AI technologies. By equipping students with AI skills from a young age, we aim to cultivate a generation of innovators who can contribute to solving both national and global issues.

The steps were:

  1. Organize a task force of Industry experts
  2. Develop AI curriculum with Industry and academia
  3. Create regulatory and policy guidelines for regional education departments
  4. Launch the Base-AI module
  5. COVID-19 hits us. We go digital with the "TinkerfromHome" YouTube Series - Every Sunday during India's COVID-19 lockdown, we cover one chapter of the Base Module. Here is the YouTube playlist link.
  6. Conduct online capacity-building workshops for educators with an incentive mechanism.
  7. Build AI competency and awareness by introducing an AI competition for students.
  8. Invite local politicians, education administrators, and other key government officials to nudge policy towards school AI adoption.
  9. Launch the AI Step-Up Module
  10. Create precedence and early success for Central Board of Secondary Education to collaborate with 英特尔 and IBM to develop AI curriculum for Schools in India.
  11. Support the adoption of AI curriculum in Indian schools by formally introducing AI as a subject in India.
  12. Democratize AI Education by handing over a toolkit to local governments.

Launch of AI Base Module | Feb 27, 2020 | India

6. AI Education Advocacy in the United States.

In the United States, Code.org 's TeachAI has introduced a robust TeachAI Toolkit. The toolkit does a great job of leveraging the mindset and attitudes of its stakeholders towards AI advocacy.

Credits : Code.org

However, teachers and local administration need help when the rubber hits the road. What will they need to implement these policies effectively? AI’s impact on education, policy, and society is profound, with opportunities and challenges ahead. The world will look at the United States ecosystem for best practices so they don't reinvent the wheel.

Focusing on developing a responsible AI education framework is crucial, ensuring ethical considerations are at the forefront. Governments and organizations must work together to create policies that promote innovation while safeguarding against potential risks. Investing in AI research and education, fostering public-private partnerships, and ensuring transparency and accountability in AI deployment are essential steps toward a sustainable AI future.

The Education X Industry Learning Series is a great start. I am excited to learn more with this community of educators and industry leaders.


Closing Remarks

Every state, county, district, city, and school system differs. There is no silver bullet. To succeed, we need to have a robust and exhaustive set of experiments to try to implement AI within our educational systems. If you are interested in building a program for your beneficiaries, these five priorities could help -

  • Fairness
  • Robustness
  • Privacy
  • Explainability
  • Governance

For the reader, I would love to ponder and ideate what AI Education implementation would be for your students, educators, institution, and ecosystem. I am happy to share my scars so that others are not wounded.

#MakersGonnaMake

#HappyTinkering



Ishu Bansal

Optimizing logistics and transportation with a passion for excellence | Building Ecosystem for Logistics Industry | Analytics-driven Logistics

5 个月

Innovative How can we effectively bridge the gap between policy and technology to create meaningful societal impact?

回复

要查看或添加评论,请登录

社区洞察

其他会员也浏览了