AI, Explained: Why It’s Different This Time | Tech News Briefing | WSJ
Here's a summary:
Artificial intelligence or AI refers to technology that can learn, reason, plan, perceive, and solve problems. Machine learning is a type of AI where systems learn without being explicitly programmed. They learn from large amounts of data.
Generative AI tools like ChatGPT have learned from huge amounts of text data and can generate responses based on what they've learned. They don't have human-level knowledge but can process language and make associations.
Neural networks are a type of machine learning modeled after the human brain. They have nodes that process information at different levels, allowing them to perceive and learn complex patterns.
AI is already widely used in apps, search engines, navigation, delivery routing, and more. New large language models are generating a lot of interest because of their human-like language abilities, though they only repeat what they've learned from data.
Risks of AI include inaccuracy, bias, and systems failing in unexpected ways. More data, better training, and oversight can help address these risks. But the risks of AI radically advancing out of human control are still mostly in science fiction.
There are few laws governing AI in the U.S., though some cities and states have limited government use of facial recognition. Europe has passed some initial AI regulations. Broader governance and guidelines are still developing.
This series "Artificially Minded" will explore what's new with AI and what it means for the future. The first episode focused on basics and current events. Upcoming episodes will address questions from listeners.
RECOMMENDED BOOK:
Age of Invisible Machines: A Practical Guide to Creating a Hyperautomated Ecosystem of Intelligent Digital Workers (English Edition)