Future of AI
AI is advancing rapidly. Understanding where the field is heading—technically, economically, and socially—helps you make better career decisions and engage more meaningfully with AI's role in society.
Near-Term Trends (1–5 Years)
Multimodal AI: models that seamlessly understand and generate text, images, audio, and video together. GPT-4o and Gemini are early examples. Expect this to become standard.
AI agents: AI systems that take multi-step actions in the real world—browsing the web, writing and running code, managing files, booking appointments. The shift from 'AI that answers questions' to 'AI that does tasks'.
On-device AI: powerful AI models running on phones, laptops, and embedded devices without sending data to the cloud. Better privacy, lower latency, no internet required.
Specialised models: instead of one massive general model, purpose-built models for medicine, law, science, or engineering that outperform general models on domain-specific tasks.
Coding AI: tools that can write, review, test, and deploy software increasingly autonomously. Already transforming software development workflows.
Deep Learning ⊂ Machine Learning ⊂ Artificial Intelligence
Longer-Term Questions
- Artificial General Intelligence (AGI): when—if ever—will AI match or exceed human cognitive performance across all domains? Estimates range from decades to never
- AI alignment: how do we ensure increasingly capable AI systems pursue goals that are beneficial to humanity?
- Economic transformation: as AI automates more work, how do societies manage the transition and distribute the economic gains?
- Regulatory convergence: will international AI governance frameworks converge, or will AI capabilities create fragmentation similar to the internet?
- Energy and environment: can the explosive growth in AI computing be powered sustainably?
- AI in science: will AI accelerate a new era of scientific discovery, potentially solving challenges like cancer and climate change?
Tip
Tip
The best way to stay current in AI is to follow a small number of high-quality sources rather than trying to read everything. Recommended starting points: The Batch (Andrew Ng's newsletter), import AI (Jack Clark), Papers With Code for research, and Anthropic/OpenAI/DeepMind research blogs. Focus on understanding the concepts behind new developments, not just the announcements.
Key Takeaways
- AI is advancing rapidly.
- Artificial General Intelligence (AGI): when—if ever—will AI match or exceed human cognitive performance across all domains? Estimates range from decades to never
- AI alignment: how do we ensure increasingly capable AI systems pursue goals that are beneficial to humanity?
- Economic transformation: as AI automates more work, how do societies manage the transition and distribute the economic gains?