AI's Global Impact: Progress, Productivity, & Growth

Plunging Costs and Lowered Barriers

The remarkable decrease in the cost of utilizing AI models marks a significant turning point. The price of querying an AI model comparable to GPT-3.5 experienced a dramatic drop from $20 per million tokens in late 2022 to a mere $0.07 by late 2024. This represents a price reduction of over 99%, and its implications extend far beyond a simple technological achievement. It acts as a crucial gateway to greater accessibility. This affordability empowers innovators and entrepreneurs located in regions with limited resources to leverage potent tools that were previously accessible only to the world’s largest corporations. They can now apply these tools to address localized challenges in a variety of essential domains, including:

• Healthcare: AI can be a transformative force in healthcare, assisting with diagnosis, treatment planning, and drug discovery. This is particularly impactful in underserved communities where access to specialized medical expertise is often limited. By leveraging AI, these communities can experience improved healthcare outcomes and a higher quality of life.
• Agriculture: AI-powered tools are revolutionizing agricultural practices. They can optimize farming techniques, predict crop yields with greater accuracy, and manage resources more efficiently. These capabilities are especially important for enhancing food security in regions where food production is vulnerable to climate change and other environmental factors. By increasing efficiency and reducing waste, AI contributes to a more sustainable and resilient agricultural sector.
• Education: AI has the potential to personalize learning experiences, tailoring educational content to the individual needs and learning styles of each student. It can provide tutoring support, offering students customized guidance and assistance, and automate administrative tasks, freeing up teachers to focus on instruction and student interaction. This can make education more accessible and effective for all students, regardless of their background or location.
• Public Service: AI can play a critical role in enhancing government services, improving infrastructure management, and assisting with disaster response. By leveraging AI, communities can become safer and more resilient, capable of effectively responding to emergencies and providing essential services to their citizens.

This democratization of AI technology empowers individuals and organizations to tackle critical issues and generate positive change within their communities. The potential for innovation is substantial, and the possibilities are limited only by our imagination and our willingness to collaborate and share knowledge.

Narrowing the Performance Gap

The gap in performance between open-weight and proprietary closed-weight models has significantly diminished. As of 2024, open-weight models are now rivaling their commercial counterparts, which is fostering increased competition and innovation across the entire AI landscape. This convergence in performance creates a more level playing field, providing researchers and developers who have limited resources with access to cutting-edge AI capabilities that were previously out of reach.

Furthermore, the performance difference among the top frontier models has also shrunk. Smaller models are now capable of achieving results that were once considered exclusive to massive-scale systems. A notable example is Microsoft’s Phi-3-mini, which delivers performance comparable to models that are 142 times larger. This brings powerful AI within reach of environments that have constrained resources, opening up exciting possibilities for deployment in resource-limited settings, such as:

• Edge Computing: Smaller AI models can be efficiently deployed on edge devices, enabling real-time processing and analysis of data without relying on constant cloud connectivity. This is particularly advantageous in remote locations or situations where network access is unreliable.
• Mobile Applications: AI-powered features can be seamlessly integrated into mobile apps, providing users with personalized experiences and intelligent assistance directly on their smartphones and tablets. This enhances the usability and functionality of mobile devices, making them even more valuable tools for everyday life.
• Embedded Systems: AI models can be embedded within devices such as sensors and robots, enabling them to perform complex tasks autonomously and adapt to changing conditions in real-time. This opens up a wide range of possibilities for automation and intelligent control in various industries.

The ability to run sophisticated AI models on smaller, more efficient hardware platforms democratizes access to AI and unlocks new applications in a wide range of industries, making it a powerful tool for innovation and problem-solving across the globe.

Remaining Obstacles: Reasoning and Data

Despite the remarkable progress that has been made in the field of AI, certain challenges persist and need to be addressed. AI systems still struggle with higher-order reasoning tasks, such as arithmetic and strategic planning – capabilities that are essential in domains where reliability and accuracy are of paramount importance. While AI can excel at tasks like pattern recognition and data analysis, it often falls short when it comes to complex problem-solving and decision-making that require critical thinking and contextual understanding.

For example, AI-powered systems may encounter difficulties in the following areas:

• Understanding Nuanced Language: AI models may misinterpret sarcasm, irony, or cultural references, leading to inaccurate or inappropriate responses. This highlights the importance of developing AI models that are more sensitive to the complexities of human communication.
• Applying Common Sense Reasoning: AI systems may lack the ability to make logical inferences or draw conclusions based on real-world knowledge. This limits their ability to understand context and make sound judgments in complex situations.
• Dealing with Ambiguity: AI models may struggle to handle situations where information is incomplete or contradictory, leading to uncertainty and errors. Developing AI models that can effectively manage ambiguity is crucial for ensuring their reliability and robustness.

Continued research and responsible application are crucial to overcoming these limitations and ensuring that AI systems are used safely and ethically. We must prioritize the development of AI models that are robust, reliable, and aligned with human values. This requires a commitment to ongoing research and development, as well as a strong ethical framework for guiding the development and deployment of AI technologies.

Another emerging concern is the rapid reduction of publicly available data used to train AI models. As websites increasingly restrict data scraping, model performance and generalizability may suffer – especially in contexts where labeled datasets are already limited. This trend may necessitate new learning approaches tailored to data-constrained environments. The availability of high-quality training data is crucial for developing effective AI models, and the increasing restrictions on data access pose a significant challenge to the AI research community.

To address this challenge, researchers are actively exploring alternative approaches to data collection and model training, such as:

• Synthetic Data Generation: Creating artificial datasets that mimic the characteristics of real-world data, providing a valuable alternative when real-world data is scarce or unavailable.
• Federated Learning: Training AI models on decentralized data sources without the need to share the raw data, protecting privacy and enabling collaboration across diverse datasets.
• Transfer Learning: Leveraging knowledge gained from training on one dataset to improve performance on another dataset, enabling the development of more robust and adaptable AI models.

By developing innovative solutions to the data scarcity problem, we can ensure that AI remains accessible and beneficial to all, regardless of data availability. This requires a collaborative effort from researchers, policymakers, and industry stakeholders to develop and promote responsible data practices that support the development of AI while protecting privacy and promoting ethical considerations.

Real-World Impact on Productivity and Workforce

One of the most promising aspects of AI is its demonstrable impact on human productivity. Last year’s AI Index was among the first to highlight research showing that AI meaningfully improves productivity. This year, follow-up studies have confirmed and expanded those findings – especially in real-world workplace environments. These studies provide compelling evidence that AI is not just a theoretical concept but a practical tool that can enhance human capabilities and drive economic growth.

One such study tracked over 5,000 customer support agents using a generative AI assistant. The tool increased productivity by 15%, with the most significant improvements seen among less experienced workers and skilled trade workers, who also boosted the quality of their work. This finding suggests that AI can help bridge the skills gap and empower individuals with limited experience to perform at a higher level, contributing to a more inclusive and equitable workforce.

The benefits of AI assistance extended beyond productivity gains. The study also found that:

• AI Helped Employees Learn on the Job: By providing real-time guidance and feedback, AI assisted employees in developing new skills and improving their performance, fostering a culture of continuous learning and development.
• AI Improved English Fluency Among International Agents: By providing access to language translation tools and personalized language learning resources, AI helped international agents communicate more effectively with customers, promoting better customer service and enhancing global collaboration.
• AI Enhanced the Work Environment: Customers were more polite and less likely to escalate issues when AI was involved, creating a more positive and collaborative work environment, reducing stress and improving employee satisfaction.

These findings highlight the potential of AI to not only improve productivity but also to enhance the overall employee experience, creating a more engaging and rewarding work environment for all.

Complementing these findings, Microsoft’s internal research initiative on AI and productivity compiled results from over a dozen workplace studies, including the largest known randomized controlled trial of generative AI integration. Tools like Microsoft Copilot are already enabling workers to complete tasks more efficiently across roles and industries. The research underscores that the impact of AI is greatest when tools are adopted and integrated strategically – and that the potential will only grow as organizations recalibrate workflows to take full advantage of these new capabilities. The key to unlocking the full potential of AI lies in thoughtful planning, careful implementation, and a commitment to continuous improvement, ensuring that AI is used effectively and ethically to benefit both individuals and organizations.

Expanding Access to Computer Science Education

As AI becomes more deeply integrated into daily life, computer science education is more critical than ever before. It’s encouraging that two-thirds of countries now offer or plan to offer K–12 CS education, a figure that has doubled since 2019. This progress reflects a growing recognition of the importance of computer science education in preparing students for the future workforce, equipping them with the skills and knowledge they need to succeed in the age of AI.

African and Latin American countries have made some of the most significant strides in expanding access to computer science education. These regions have recognized the potential of computer science education to drive economic development and empower their citizens, creating opportunities for innovation and entrepreneurship. However, the benefits of this progress are not yet universal – many students across Africa still lack access to computer science education due to basic infrastructure gaps, including lack of electricity in schools. Closing this digital divide is essential to preparing the next generation to not only use AI, but to shape it, ensuring that everyone has the opportunity to participate in the AI revolution.

To ensure that all students have access to quality computer science education, we must address the following challenges:

• Infrastructure Development: Investing in basic infrastructure, such as electricity and internet connectivity, in schools and communities, providing the foundation for effective computer science education.
• Teacher Training: Providing teachers with the training and resources they need to effectively teach computer science concepts, ensuring that they are equipped to inspire and empower the next generation of computer scientists.
• Curriculum Development: Developing engaging and relevant computer science curricula that meet the needs of diverse learners, making computer science accessible and exciting for all students.
• Equity and Inclusion: Ensuring that all students, regardless of their background or location, have equal opportunities to participate in computer science education, creating a more diverse and inclusive computer science community.

By addressing these challenges, we can create a more inclusive and equitable computer science education system that prepares all students to thrive in the age of AI, empowering them to become innovators, problem-solvers, and leaders in the digital world.

Our Shared Responsibility

We stand at a significant inflection point – one that calls for thoughtful action as much as innovation. The rapid progress in AI brings enormous potential to improve productivity, solve real-world challenges, and drive economic growth. But realizing that potential requires continued investment in robust infrastructure, high-quality education, and responsible deployment of AI technologies. We must embrace a holistic approach that considers the ethical, social, and economic implications of AI, ensuring that it is used for the benefit of all humanity.

To make the most of this moment, we need to support workers with learning new skills and tools to apply AI effectively in their jobs. Nations and businesses that invest in AI skilling will foster innovation and open doors to more people to build meaningful careers that contribute to a stronger economy. This requires a collaborative effort between governments, businesses, and educational institutions to create training programs and resources that equip workers with the skills they need to succeed in the age of AI, preparing them for the jobs of the future.

The goal is clear: to turn technical breakthroughs into practical impact at scale. By working together, we can harness the power of AI to create a more prosperous, equitable, and sustainable future for all. This requires a long-term commitment to research, development, and deployment of AI technologies that are aligned with human values and promote the common good, ensuring that AI is used to create a better world for all.