{"id":"arxiv:1905.10985","title":"AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence","date":"2019-05-27","section":"Agents, open-endedness, AGI","quality":-0.283584,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/","source_url":"https://arxiv.org/abs/1905.10985","text_available":true,"authors":["Jeff Clune"],"outgoing":[{"source":"arxiv:1905.10985","target":"arxiv:1710.02298","evidence":"id","id":"arxiv:1710.02298","title":"Rainbow: Combining Improvements in Deep Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/4183a5aecdb6558733bd0b2028e6197904c8b46173181bf6c8270fb6cbcfc973/"},{"source":"arxiv:1905.10985","target":"arxiv:1712.06567","evidence":"id","id":"arxiv:1712.06567","title":"Deep Neuroevolution: Genetic Algorithms Are a Competitive Alternative for Training Deep Neural Networks for Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/"},{"source":"arxiv:1905.10985","target":"arxiv:1901.01753","evidence":"id","id":"arxiv:1901.01753","title":"Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/a9790a1ede488fbbb2d7b4cdb97b884a5a1e21357b72c5e3419a83ea629d5a4b/"},{"source":"arxiv:1905.10985","target":"doi:10.1073/pnas.1611835114","evidence":"title","id":"doi:10.1073/pnas.1611835114","title":"Overcoming catastrophic forgetting in neural networks","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/849082368a44d06b93ef3ba95bcfddc6672b7a5e8064ad2b2d19601639266e8b/"}],"incoming":[{"source":"arxiv:2406.04268","target":"arxiv:1905.10985","evidence":"title","id":"arxiv:2406.04268","title":"Open-Endedness is Essential for Artificial Superhuman Intelligence","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/13964b2db9a9f793d1fcae07ca83fbe09f65454214269cac56777cb8be423a4e/"},{"source":"arxiv:2408.08435","target":"arxiv:1905.10985","evidence":"id","id":"arxiv:2408.08435","title":"Automated Design of Agentic Systems","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/c9890d5668d1c1b33e29569b74626fd586f456af6a97c657b8c99895c90f59a5/"},{"source":"arxiv:2505.22954","target":"arxiv:1905.10985","evidence":"id","id":"arxiv:2505.22954","title":"Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/c29f716871f80314868b5647692de387f26e12166b93bd362915fd272b6e081c/"},{"source":"arxiv:2602.07755","target":"arxiv:1905.10985","evidence":"id","id":"arxiv:2602.07755","title":"Learning to Continually Learn via Meta-learning Agentic Memory Designs","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/4d0a0ea160495ea50583e0e24027b6085e2f430a44195e3053ba69cda8cc82ea/"},{"source":"arxiv:2603.19461","target":"arxiv:1905.10985","evidence":"id","id":"arxiv:2603.19461","title":"Hyperagents","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/17eec8a15debc212b5530f117dd8ba8bad339590729cbbc4f75a47755ebcb544/"},{"source":"arxiv:2608.19197","target":"arxiv:1905.10985","evidence":"id","id":"arxiv:2608.19197","title":"SPADE: Self-Play in Adaptive Synthetic Executable Environments","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/c524376d6034b5d79661d6f1a444591404f4a250bcb0790d802c05e02f8249df/"},{"source":"doi:10.1038/s42256-023-00754-x","target":"arxiv:1905.10985","evidence":"id","id":"doi:10.1038/s42256-023-00754-x","title":"A social path to human-like artificial intelligence","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/25e744318984a723282df54f9f39bf66e3a17e6f3b09c90a80af90c250e799be/"}],"text_parts":[{"start":0,"end":15874,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html"},{"start":15874,"end":31004,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-2.html"},{"start":31004,"end":46730,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-3.html"},{"start":46730,"end":61155,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-4.html"},{"start":61155,"end":76186,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-5.html"},{"start":76186,"end":91484,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-6.html"},{"start":91484,"end":106761,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-7.html"},{"start":106761,"end":108714,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-8.html"}],"contents":[{"title":"AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence","offset":0,"level":1,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html#text"},{"title":"Abstract","offset":105,"level":6,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html#text"},{"title":"1 Two approaches to producing general AI: the manual approach vs. AI-generating algorithms","offset":2331,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html#text"},{"title":"1.1 The manual AI approach","offset":3181,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html#text"},{"title":"1.2 A different approach: AI-generating algorithms (AI-GAs)","offset":10014,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html#text"},{"title":"1.3 The Three Pillars required to produce an AI-GA","offset":14822,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html#text"},{"title":"1.4 AI-GAs: A new grand challenge for AI research","offset":15819,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-1.html#text"},{"title":"1.5 Both the manual and AI-GA paths are worth pursuing, irrespective of which is more likely to be the fastest path to general AI","offset":19561,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-2.html#text"},{"title":"1.6 Which path is more likely to produce general AI first?","offset":21387,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-2.html#text"},{"title":"2 Research into the Three Pillars","offset":23745,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-2.html#text"},{"title":"2.1 The First Pillar: Meta-learning Architectures","offset":24447,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-2.html#text"},{"title":"2.2 The Second Pillar: Meta-learning the learning algorithms","offset":27786,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-2.html#text"},{"title":"2.3 The Third Pillar: Generating effective learning environments and training data","offset":35902,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-3.html#text"},{"title":"2.3.1 Encouraging behavioral diversity","offset":49288,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-4.html#text"},{"title":"2.3.2 Quality Diversity algorithms","offset":52906,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-4.html#text"},{"title":"2.3.3 Environment-generating quality-diversity algorithms","offset":59273,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-4.html#text"},{"title":"2.3.4 More expressive environment encodings","offset":63143,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-5.html#text"},{"title":"2.3.5 The viability of the Third Pillar","offset":73186,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-5.html#text"},{"title":"3 Discussion","offset":75377,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-5.html#text"},{"title":"4 Safety and ethical considerations","offset":90987,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-6.html#text"},{"title":"5 Conclusions","offset":101109,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-7.html#text"},{"title":"Acknowledgements","offset":108080,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/part-8.html#text"}],"text_code_points":108714,"source_markdown":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/source.md"}