Reading The AI Index 2025: Is AI Your Friend, Foe, Or Co-Pilot?

The 2025 Artificial Intelligence Index Report released by the Stanford University Institute for Human-Oriented Artificial Intelligence provides a good overview of the ongoing artificial intelligence revolution. Let’s interpret it in four simple concepts: cognition (understand what is happening), appreciation (seeing benefits), acceptance (face challenges), and responsibility (find our responsibilities).

Cognition: Artificial intelligence is everywhere and is developing rapidly

We need to be keenly aware of how quickly artificial intelligence is developing and spreading. Artificial intelligence systems are constantly improving, achieving excellent results in math and complex thinking tests, and just a year ago they failed miserably in these tests. Imagine AI solving complex coding problems or graduate-level scientific problems—since 2023, performance has improved significantly in benchmarks testing these skills. In a coding test (SWE-bench), AI only solved 4% of the problems in 2023, but solved nearly 72% of the problems in 2024!

This is not just happening in the lab. Artificial intelligence is emerging in daily life. In 2023, the U.S. Food and Drug Administration approved 223 medical devices using artificial intelligence, compared with only 6 in 2015. In more and more cities, you can now call self-driving taxis from companies like Waymo (in the United States) or Baidu (in China).

In 2024, global investment in artificial intelligence reached an astonishing US$252.3 billion. From this perspective - the investment in this year alone is almost equivalent to the annual economic output of countries such as Portugal, which in 2023 will be approximately US$289 billion. It also dwarfs the budgets of major government agencies; for example, it is about ten times the entire NASA budget for fiscal 2023 ($25.4 billion). In addition, this annual investment in artificial intelligence significantly exceeds the estimated total historical cost of building an international space station over decades, which is generally believed to be between $100 billion and $150 billion.

More and more companies (78% in 2024, up from 55% in 2023) are using AI. While the United States is still creating state-of-the-art AI models, China is quickly catching up, and their best models are now almost as good as the top American models. Good news for users: Artificial intelligence is becoming cheaper and easier to obtain. Powerful AI systems are getting cheaper – the cost of certain tasks has dropped more than 280 times between late 2022 and late 2024! Governments are also paying attention to this matter, investing billions of dollars in artificial intelligence development and starting to formulate rules for its use.

Appreciation: The advantages of artificial intelligence – helping us and society

It’s easy to focus on futuristic hype, but artificial intelligence has brought about practical benefits. That’s what appreciation is about – see how AI can help us, especially when combined with human skills (hybrid intelligence) and used in social good (prosocial artificial intelligence).

One major advantage is productivity. Research shows that artificial intelligence tools can help people get more work done at work, often by helping inexperienced employees effectively level the playing field. Artificial intelligence can be considered as an intelligent assistant that handles daily tasks, allowing humans to handle more complex problem-solving or creative work.

In addition to personal interests, artificial intelligence is expected to solve major social problems. Prosocial AI aims to leverage this technology to realize the best potential of people and the planet – think of using AI to monitor climate change, manage resources like water more effectively, improve health care services in remote communities, or create personalized learning tools for students. Science and medicine are making great progress. Artificial intelligence plays a role in the discovery worthy of the Nobel Prize. Tools like AlphaFold3 are helping scientists learn about diseases and develop new drugs faster than ever before. Artificial intelligence systems perform better than doctors on certain specific tasks, such as finding cancer in scans. Tools that listen during a doctor’s visit (AI Stenographer) are helping to reduce the paperwork burden of doctors.

The public’s perception of artificial intelligence presents a hybrid picture, highlighting the possible gap between current perception and future outlook. According to the HAI Index Report 2025, using data from Ipsos in 2024, a slightly larger majority (55%) of people around the world believe that current AI products and services bring more benefits than disadvantages, indicating a cautious global optimism. However, when examining what specific regions expect for the future, a very different view emerges; a recent Pew Research Center report (April 2025) found that only 17% of American adults believe that AI will have a positive impact on the United States in the next 20 years, and more than half express more concern than excitement about the growing role of AI. This difference emphasizes how perceptions about AI vary by region and whether people consider current benefits or future implications.

(For more information on the potential of prosocial artificial intelligence, please check out the white paper released after the 2024 United Nations Science Summit).

Acceptance: Facing problems and traps

Appreciating the benefits also means we have to accept the shortcomings and risks. Building trust in artificial intelligence is a major challenge. Many people are worried about how companies use their personal data, and concerns about unfairness or bias in AI are real. Even AI systems designed to remain neutral may absorb and repeat harmful racial or gender stereotypes found in their training data.

Disinformation is another concern. Artificial intelligence makes it easier to create fake images, videos or audio (deep forgery) that can be used to spread lies (especially during elections) or harass others. We are also seeing a growing number of reports that AI is causing problems, such as false allegations of facial recognition or chatbots giving harmful advice. Even AI-learning data has become a problem, as more and more websites block access, which may limit the productivity of AI in the future.

Despite the progress, AI is not perfect. It remains difficult to perform complex reasoning that requires common sense or planning multiple steps ahead of time. Training the most powerful AI models requires a lot of energy, resulting in carbon emissions. While AI has helped many workers, the concerns about jobs being replaced by automation are understandable. Accepting AI means being honest with these complex issues.

Responsibility: Take responsibility for the future of artificial intelligence

Finally, responsibility refers to who is responsible for guiding the development and use of artificial intelligence. Efforts to ensure that AI is developed responsibly are growing, but there is still work to be done. We are starting to see better ways to test the security and fairness of AI, but that is not the standard practice for all important AI developers.

Governments are becoming more active, investing in artificial intelligence and setting rules. Countries have begun to cooperate more on AI security , establishing international organizations and agreements with a focus on making AI trustworthy and in line with human rights.

However, companies often discuss responsible AI more than implementing powerful safeguards. (You may remember that more than 1,000 technology leaders, researchers and others signed a moratorium on the development of artificial intelligence that they failed to comply with)

We need more transparency about how AI systems are built and trained. In this regard, everyone plays a role—developers, businesses, governments, and citizens—each of us. Education is the key. We need to make sure people understand AI, but there are big gaps. For example, in the United States, while 81% of computer science teachers believe that artificial intelligence should be taught in schools, less than half (including only 46% of high school computer science teachers) think they have the ability to teach it. Bridging this gap is crucial to building a future where everyone can navigate and shape the AI ​​world. Meanwhile, it must be remembered that billions of people have never been exposed to ChatGTP—not because they don’t want to, but because they have difficulty accessing clean water, nutritious food or the internet.

Your role: 4 practical steps to use natural intelligence to control artificial intelligence

So, what can you do? It's not just the responsibility of a technician or a politician. Each of us should play a role—this requires natural intelligence:

Cognition: Be curious about artificial intelligence. Learn how it works and where it appears in your life. Focus on reliable news sources (not just hype) to see its impact.

Appreciate: Think about how artificial intelligence can help you or your community. Can AI tools make your work easier or more creative? Looking for examples of artificial intelligence being used for positive social change.

Accept: Be a critical thinker. Don't just accept what AI tells you. Question its output, beware of potential biases, and understand the risks from privacy to false information.

Responsibility: Use your voice. Support companies and policies that prioritize ethical and responsible AI. Think about the way you use artificial intelligence – are you using it responsibly? Discuss with others about the importance of guiding AI for the benefit of everyone.

The 2025 Artificial Intelligence Index Report is a good starting point for understanding the ever-changing pattern of artificial intelligence. It is probably one of the most powerful technologies ever created by humans. By staying alert, appreciating benefits, accepting challenges and taking responsibility for ourselves and others, we can help shape an AI future that benefits all of us.

The above is the detailed content of Reading The AI Index 2025: Is AI Your Friend, Foe, Or Co-Pilot?. For more information, please follow other related articles on the PHP Chinese website!