Analýza současného stavu prací v oblasti obecné umělé inteligence (AGI)

This undergraduate thesis examines the topic of Artificial General Intelligence. Its goals are describing the role of AGI in the broader field of AI, analyzing current research in the fields of AI and neuroscience, as well as communicating expert opinions about AGI’s feasibility and future development. The thesis also explores the relationship between AGI and the concept of technological singularity, including contemporary views of its likelihood. Whilst describing AGI, this thesis also presents theories behind the nature of AGI. The importance of AGI in the field of artificial intelligence is steadily growing, becoming a major focus of research and investment in the private sector. The current trend in AGI development is deep learning, as well as reinforced learning with human or AI feedback. Previous assumptions about the feasibility of large language models as a path towards AGI are getting reconsidered, with OpenAI’s models taking the spotlight this year. Brain emulation is conversely trailing behind, with its most recent breakthrough being highly detailed imaging of a mouse brain. Major figures in the AI sphere display notable optimism about the realization of AGI within the foreseeable future, with key milestones expected by 2045 at the latest. This optimism, however, is tempered with concerns over the potential negative consequences of AGI for human civilization. Most notable figures in the field of AI urge the importance of aligning AI development with human ethics to mitigate risks and ensure beneficial outcomes.