In his paper titled "AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence," Jeff Clune discusses the creation of general artificial intelligence (AI), which refers to AI that is as smart or even smarter than humans. The current dominant approach in the machine learning community is the "manual AI approach," where researchers attempt to discover each piece required for intelligence with the assumption that some future group will combine them into a complex thinking machine. However, Clune proposes another path that may be more successful at producing general AI based on the trend in machine learning that hand-designed solutions are eventually replaced by more effective learned solutions. This new approach involves creating an AI-generating algorithm (AI-GA) that automatically learns how to produce general AI. Three pillars are essential for this approach: meta-learning architectures, meta-learning the learning algorithms themselves, and generating effective learning environments. Clune argues that either approach could produce general AI first and both are scientifically worthwhile irrespective of which is faster. However, because both approaches are promising and yet the ML community is currently committed to the manual approach, he suggests increasing research investment in the AI-GA approach. To encourage such research, he describes promising work in each of the three pillars and discusses safety and ethical considerations specific to AI-GAs. Clune believes pursuing AI-GAs should be considered a new grand challenge of computer science research because it may be the fastest path to general AI and it is scientifically interesting to understand how a simple algorithm can produce general intelligence similar to how Darwinian evolution produced human intelligence on Earth. Overall, Clune's paper provides an exciting alternative path towards achieving general artificial intelligence through automated learning algorithms rather than relying solely on manual design.
Jeff Clune wrote a paper about making smart robots. Right now, people have to teach robots how to do things one by one. But Jeff has an idea to make a robot that can learn on its own! He wants to make a special computer program called an AI-GA that can create smart robots automatically. To do this, he needs three important things: special computer designs, teaching the AI-GA how to learn better, and making good learning environments for the robots. This is a big challenge for scientists because it could help us make really smart robots faster and also help us understand how humans became so smart.
Definitions
- Artificial Intelligence (AI): when computers or machines can do things that normally need human intelligence like thinking and problem-solving
- Machine Learning: when computers or machines can learn from data without being specifically programmed
- Algorithm: a set of instructions given to a computer or machine to solve a problem or complete a task
- Meta-learning: when computers or machines learn how to learn better
AI-GAs: AI-Generating Algorithms, an Alternate Paradigm for Producing General Artificial Intelligence
The field of artificial intelligence (AI) has been making tremendous advances in recent years, with many researchers believing that general AI – AI that is as smart or even smarter than humans – may be achievable in the near future. In his paper titled "AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence," Jeff Clune discusses a potential new approach to creating general AI and why it should be considered a grand challenge of computer science research.
The Manual Approach to Creating General AI
Currently, the dominant approach in the machine learning community is the "manual AI approach," where researchers attempt to discover each piece required for intelligence with the assumption that some future group will combine them into a complex thinking machine. This manual approach involves designing systems from scratch and requires significant effort from both humans and computers.
An Alternative Path Towards General AI
Clune proposes another path that may be more successful at producing general AI based on the trend in machine learning that hand-designed solutions are eventually replaced by more effective learned solutions. This new approach involves creating an Artificial Intelligence Generating Algorithm (AI-GA) which automatically learns how to produce general AI without requiring human intervention. Three pillars are essential for this approach: meta-learning architectures, meta-learning the learning algorithms themselves, and generating effective learning environments.
Promising Work in Each Pillar
Clune describes promising work being done in each of these three pillars which could lead to breakthroughs towards achieving general artificial intelligence through automated learning algorithms rather than relying solely on manual design. He also discusses safety and ethical considerations specific to using such algorithms since they would likely have far reaching implications if successful.
A New Grand Challenge of Computer Science Research?
Clune believes pursuing this alternative path should be considered a new grand challenge of computer science research because it may be the fastest path to general AI and it is scientifically interesting to understand how a simple algorithm can produce general intelligence similar to how Darwinian evolution produced human intelligence on Earth. Overall, Clune's paper provides an exciting alternative path towards achieving general artificial intelligence through automated learning algorithms rather than relying solely on manual design.
Perhaps the most ambitious scientific quest in human history is the creation of general artificial intelligence, which roughly means AI that is as smart or smarter than humans. The dominant approach in the machine learning community is to attempt to discover each of the pieces required for intelligence, with the implicit assumption that some future group will complete the Herculean task of figuring out how to combine all of those pieces into a complex thinking machine. I call this the "manual AI approach". This paper describes another exciting path that ultimately may be more successful at producing general AI. It is based on the clear trend in machine learning that hand-designed solutions eventually are replaced by more effective, learned solutions. The idea is to create an AI-generating algorithm (AI-GA), which automatically learns how to produce general AI. Three Pillars are essential for the approach: (1) meta-learning architectures, (2) meta-learning the learning algorithms themselves, and (3) generating effective learning environments. I argue that either approach could produce general AI first, and both are scientifically worthwhile irrespective of which is the fastest path. Because both are promising, yet the ML community is currently committed to the manual approach, I argue that our community should increase its research investment in the AI-GA approach. To encourage such research, I describe promising work in each of the Three Pillars. I also discuss AI-GA-specific safety and ethical considerations. Because it it may be the fastest path to general AI and because it is inherently scientifically interesting to understand the conditions in which a simple algorithm can produce general AI (as happened on Earth where Darwinian evolution produced human intelligence), I argue that the pursuit of AI-GAs should be considered a new grand challenge of computer science research.
