Key Takeaways
- Artificial General Intelligence (AGI) is a term that has been around for over two decades, but its definition and potential impact are still debated among experts.
- AGI is expected to revolutionize various fields, including science, medicine, and education, by providing a powerful tool for accelerating discovery and problem-solving.
- The development of AGI is a gradual process, and its emergence will likely be subtle, reflecting the continuum of intelligence and evolving from current AI systems.
- Experts emphasize the need for a common language to discuss AGI and its potential impacts, as well as the importance of preparing for its arrival by addressing risks, complexities, and adaptation challenges.
- AGI is expected to raise profound questions about what it means to be human and how we will grapple with the opportunities and challenges it presents.
Introduction to AGI
The concept of Artificial General Intelligence (AGI) has been around for over two decades, yet it remains a topic of debate among experts. According to Demis Hassabis, co-founder and CEO of Google DeepMind, "I’ve always been most passionate about building AI to help us answer fundamental questions about the universe and solve critical societal problems in medicine, materials, and energy." AGI is expected to revolutionize various fields by providing a powerful tool for accelerating discovery and problem-solving. However, questions that were once theoretical are now becoming more practical, and experts are working to provide answers. As Shane Legg, Chief AGI Scientist at Google DeepMind, notes, "We urgently need a common language so we can speak cogently about this technology and understand its potential impacts."
Defining AGI
Defining AGI is a crucial step in understanding its potential impact. According to Legg, "We’ve long defined AGI as a system exhibiting all the cognitive capabilities that humans have." However, he also emphasizes that intelligence is a spectrum with different levels of capability, and two in particular warrant being called out: minimal AGI and full AGI. Minimal AGI should be able to perform cognitive tasks that people can typically do, such as solving simple logic puzzles or learning a game after a few minutes of instruction. Full AGI, on the other hand, would be a system capable of the full extent of what human intelligence can achieve, from paradigm-shifting scientific theories to elegant inventions. As Legg notes, "Full AGI would be a system capable of the full extent of what human intelligence can achieve, from paradigm-shifting scientific theories on par with Einstein’s special relativity to elegant inventions like the game of Go to masterworks of art."
Current State of AGI
Currently, AI systems have some impressive capabilities, but they are "jagged intelligences" in that they lack consistency across the board. According to Hassabis, "AI systems today have some impressive capabilities, but they are ‘jagged intelligences’ in that they lack consistency across the board." They can fail at relatively trivial tasks and don’t continually learn. They also aren’t capable of true creativity or originality, which is an important benchmark for AGI. As Hassabis notes, "Can a system propose a novel and meaningful scientific hypothesis, not just prove an existing one? Can it invent a game as elegant as Go, not just master it? That is much more difficult." To address this, researchers are working on building "world models," or models capable of understanding how the physical world works. According to Hassabis, "We are building what we call ‘world models,’ or models capable of understanding how the physical world works."
Future of AGI
The emergence of AGI is expected to be a gradual process, rather than a sudden, dramatic moment. As Legg notes, "What might surprise many people is that AGI won’t likely arrive in a single, dramatic moment. Our imagination is shaped by portrayals of artificial intelligence as being suddenly ‘switched on.’" Instead, it will be a subtle emergence reflecting the continuum of intelligence and evolving from current AI systems. According to James Manyika, Senior Vice President of Research, Labs, Technology & Society at Google, "How are we going to grapple with the opportunities, the complexities, the adaptation challenges, and what it will mean to be human in the age of AI?" Manyika emphasizes the need to prepare for the arrival of AGI by addressing risks, complexities, and adaptation challenges. As he notes, "We should think about the exciting possibilities that this is going to enable us to do. With more capable and more general systems, we should be able to tackle some of humanity’s greatest challenges and opportunities."
Preparing for AGI
Preparing for the arrival of AGI requires a multifaceted approach. According to Manyika, "We should prepare to fully capitalize on that when we get there." This includes thinking about the exciting possibilities that AGI will enable, as well as the risks and complexities associated with it. Manyika notes, "How do we govern such technology responsibly? How do we make sure it’s safe and aligned with human values and preferences?" Additionally, researchers are working to identify the technical and scientific breakthroughs needed to build more capable and useful systems. As Manyika asks, "What additional research breakthroughs are going to be needed to build more capable and more useful systems?" Ultimately, the development of AGI raises profound questions about what it means to be human and how we will grapple with the opportunities and challenges it presents. As Manyika notes, "How are we going to grapple with the opportunities, the complexities, the adaptation challenges, and what it will mean to be human in the age of AI?"
https://www.theatlantic.com/sponsored/google/snapshot-of-current-thinking/4048/
