Impact of Artificial Intelligence
Question 4 of 6: Which big forces are reshaping society? · The whole story
In short: AI is best seen as a general-purpose technology, like electricity or the internet: not one product but a set of abilities spreading into almost everything. It already recognises patterns, handles language, predicts, writes code and helps with creative work, and hundreds of millions of people use it every week. It is raising productivity, but raises hard questions about who gains, about trust in information, surveillance, inequality and health. Thinking clearly about AI means judging what it actually does, asking who benefits and who pays, separating the near term from the long term, and focusing on how we govern it.
The Question
Section titled “The Question”Enhanced Automation looks at what machines do to work. This article steps back and asks: what is AI, what can it do, and what does it change for society?
A General-Purpose Technology
Section titled “A General-Purpose Technology”Like electricity or the internet, AI is not a single product or application. It is a cluster of abilities being built into tools, systems and processes worldwide, and its effects will be felt in almost every area of life.
What AI Can Do Now
Section titled “What AI Can Do Now”- Recognise patterns: objects in images, unusual data, fraud.
- Handle language: summarise, translate, write, answer questions.
- Predict: demand, risk, what you might want to watch or buy.
- Write software: write, explain and fix code.
- Help with creative work: images, music and design variations.
These abilities are already in tools used by hundreds of millions of people, often without them noticing. ChatGPT alone reported about 800 million weekly users in late 2025.
The Economic Impact
Section titled “The Economic Impact”AI is raising productivity in health care, finance, logistics and more. Much of this comes through automation: AI pushes it beyond routine physical tasks into thinking work, such as analysis, writing, diagnosis and planning. That raises questions about who gets the gains:
- If AI makes capital (tools and software) much more productive compared to human work, who captures the value?
- How do workers keep their position when their skills become less rare?
- What new kinds of valuable work appear?
The Social Impact
Section titled “The Social Impact”- Information and trust: AI makes convincing text, images, audio and video cheap to produce, a real risk for misinformation and for trust in what we see and read (see Democracy and Reliable Information).
- Surveillance and privacy: AI makes watching people far easier and cheaper, for governments, companies and individuals, with deep consequences for privacy and political freedom.
- Inequality: access to powerful AI is uneven. Those who use it well gain an advantage; those who can’t risk falling further behind.
- Health: AI is speeding up drug discovery and improving diagnosis, while raising questions about health data and who controls these systems.
Thinking Clearly About AI
Section titled “Thinking Clearly About AI”AI attracts both too much hype and too much fear. A more grounded approach:
- Judge capabilities concretely: what can a specific system actually do, reliably, at scale?
- Ask who benefits and who pays: technology is never neutral in who it helps.
- Separate near term from long term: today’s AI is genuinely transformative and genuinely limited; long-term scenarios are far more uncertain (see Big Risks).
- Focus on governance: the most important questions are social and political, not technical: how we choose to use AI, and how we limit it.
Examples
Section titled “Examples”- Protein folding: in 2020, DeepMind’s AlphaFold predicted the 3D shapes of proteins with near-experimental accuracy, solving a 50-year-old problem in biology. Its creators shared the 2024 Nobel Prize in Chemistry.
- The EU AI Act: in force since August 2024, it is the first broad law on AI. It bans some uses, such as social scoring by governments, and sets stricter rules for high-risk uses like hiring, credit and policing.
As an Individual
Section titled “As an Individual”- Learn to use AI tools at work and at home: the productivity gains are real.
- Check what they produce: they make confident mistakes.
- Follow the policy debates: these decisions will shape the world you live in.
The Human Side
Section titled “The Human Side”AI can feel exciting, threatening or both. It can save time and open new possibilities, but it can also make people feel replaceable or unsure what is real. Staying curious, keeping your own judgment, and keeping real human contact at the centre help keep AI a tool rather than a replacement. See Managing fear in the Personal Guide.
Open Questions
Section titled “Open Questions”- Who should own and control the most powerful AI systems?
- How can societies keep trust in information when anything can be faked?
- Should some decisions, such as in justice or warfare, always be made by people?
Where Next
Section titled “Where Next”- Previous: Enhanced Automation
- Next: Life After Automation
- Related: Big Risks, Democracy and Reliable Information
Sources
Section titled “Sources”- Stanford Institute for Human-Centered AI, AI Index Report 2025
- OpenAI, DevDay announcement of weekly ChatGPT users (October 2025)
- Erik Brynjolfsson, Danielle Li and Lindsey Raymond, “Generative AI at Work”, Quarterly Journal of Economics (2025)
- Jumper and others, “Highly accurate protein structure prediction with AlphaFold”, Nature (2021), and the Nobel Prize in Chemistry 2024
- European Union, Artificial Intelligence Act (Regulation 2024/1689)
- International AI Safety Report (2025), chaired by Yoshua Bengio