Is AI going to kill us? | Liv Boeree and Lex Fridman

TL;DR

The speaker expresses worries about AGI and super intelligence systems but also believes that humanity can find solutions to mitigate potential risks.

Transcript

what about ai are you worried about agi super intelligence systems or paperclip maximizer type of type of situation yes i'm definitely worried about it but i feel kind of bipolar in the some days i wake up and i'm like you're excited about the future well exactly i'm like wow we can unlock the mysteries of the universe you know escape the game um a... Read More

Key Insights

• 🐎 The race dynamics in AGI development pose a challenge, as companies prioritizing speed over safety may gain a competitive advantage.
• 👾 DeepMind's focus on philosophical topics, game theory, and non-naive thinkers is a positive development in AGI development.
• 🦺 Ensuring the correlation between intelligence and wisdom in AGI is crucial for the safety of humanity.
• ❓ Imagining destructive trajectories while maintaining faith in finding solutions is important in addressing AGI concerns.
• 🤔 The probability of AGI-related risks has increased over the years, making it necessary for individuals to think about these issues.
• 🥰 Having positive beliefs and open hearts can provide hope and breathing room for finding solutions to AGI concerns.
• 💦 Although there is work to be done, the momentum of humanity and belief in finding solutions contribute to optimism.
• 🎮 Video games like Battle Polytopia provide a fun escape from the suffering and complexity of life.

Questions & Answers

Q: What are some concerns regarding AGI and super intelligence systems?

The speaker expresses worries about the race dynamics in AGI development and the possibility of prioritizing speed over safety, which could lead to negative consequences.

Q: Why is it important for companies like DeepMind to consider philosophical and game theory aspects of AGI?

DeepMind's focus on philosophical topics and game theory is crucial because the development of AGI involves complex coordination mechanisms and arms race scenarios that require non-naive thinkers.

Q: What is the orthogonality problem and why is it a concern?

The orthogonality problem refers to the idea that super intelligence and super wisdom are not necessarily correlated. It is a concern because to ensure safety, it is crucial for AGI to possess both intelligence and wisdom.

Q: How can the merging of intelligence and wisdom be achieved in AGI development?

The speaker hopes that the process of AGI development happens slowly, allowing time for modifications and debates. They believe that constant discussions and modifications can lead to a safe and well-coordinated AGI system.

Summary & Key Takeaways

• The speaker is worried about AGI and super intelligence systems, specifically regarding race dynamics and the risk of optimizing for speed at the cost of safety.

• They acknowledge the importance of companies like DeepMind thinking about the philosophical and game theory aspects of AGI.

• The speaker discusses the orthogonality problem and the need to ensure that super intelligence and super wisdom are correlated for the safety of humanity.