Open-Endedness is Essential for Artificial Superhuman Intelligence
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909218529869824 |
|---|---|
| author | Hughes, Edward Dennis, Michael Parker-Holder, Jack Behbahani, Feryal Mavalankar, Aditi Shi, Yuge Schaul, Tom Rocktaschel, Tim |
| author_facet | Hughes, Edward Dennis, Michael Parker-Holder, Jack Behbahani, Feryal Mavalankar, Aditi Shi, Yuge Schaul, Tom Rocktaschel, Tim |
| contents | In recent years there has been a tremendous surge in the general capabilities of AI systems, mainly fuelled by training foundation models on internetscale data. Nevertheless, the creation of openended, ever self-improving AI remains elusive. In this position paper, we argue that the ingredients are now in place to achieve openendedness in AI systems with respect to a human observer. Furthermore, we claim that such open-endedness is an essential property of any artificial superhuman intelligence (ASI). We begin by providing a concrete formal definition of open-endedness through the lens of novelty and learnability. We then illustrate a path towards ASI via open-ended systems built on top of foundation models, capable of making novel, humanrelevant discoveries. We conclude by examining the safety implications of generally-capable openended AI. We expect that open-ended foundation models will prove to be an increasingly fertile and safety-critical area of research in the near future. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_04268 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Open-Endedness is Essential for Artificial Superhuman Intelligence Hughes, Edward Dennis, Michael Parker-Holder, Jack Behbahani, Feryal Mavalankar, Aditi Shi, Yuge Schaul, Tom Rocktaschel, Tim Machine Learning Artificial Intelligence In recent years there has been a tremendous surge in the general capabilities of AI systems, mainly fuelled by training foundation models on internetscale data. Nevertheless, the creation of openended, ever self-improving AI remains elusive. In this position paper, we argue that the ingredients are now in place to achieve openendedness in AI systems with respect to a human observer. Furthermore, we claim that such open-endedness is an essential property of any artificial superhuman intelligence (ASI). We begin by providing a concrete formal definition of open-endedness through the lens of novelty and learnability. We then illustrate a path towards ASI via open-ended systems built on top of foundation models, capable of making novel, humanrelevant discoveries. We conclude by examining the safety implications of generally-capable openended AI. We expect that open-ended foundation models will prove to be an increasingly fertile and safety-critical area of research in the near future. |
| title | Open-Endedness is Essential for Artificial Superhuman Intelligence |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2406.04268 |