
Sign up to save your podcasts
Or


Read the full transcript here.
Has society reached ‘peak progress’? Can we sustain the level of economic growth that technology has enabled over the last century? Have researchers plucked the last of science's "low-hanging fruit?" Why did early science innovators have outsized impact per capita? As fields mature, why does per-researcher output fall? Can a swarm of AI systems materially accelerate research? What does exponential growth hide about the risk of collapse? Will specialized AI outcompete human polymaths? Is quality of life still improving - and how confident are we in those measures? Is it too late to steer away from the attention economy? Can our control over intelligent systems scale as we develop their power? Will AI ever be capable of truly understanding human values? And if we reach that point, will it choose to align itself?
Holden Karnofsky is a Member of Technical Staff at Anthropic, where he focuses on the design of the company's Responsible Scaling Policy and other aspects of preparing for the possibility of highly advanced AI systems in the future. Prior to his work with Anthropic, Holden led several high-impact organizations as the co-founder and co-executive director of charity evaluator GiveWell, and one of three Managing Directors of grantmaking organization Open Philanthropy. You can read more about ideas that matter to Holden at his blog Cold Takes.
Further reading:
Staff
Music
Affiliates
By Spencer Greenberg4.8
133133 ratings
Read the full transcript here.
Has society reached ‘peak progress’? Can we sustain the level of economic growth that technology has enabled over the last century? Have researchers plucked the last of science's "low-hanging fruit?" Why did early science innovators have outsized impact per capita? As fields mature, why does per-researcher output fall? Can a swarm of AI systems materially accelerate research? What does exponential growth hide about the risk of collapse? Will specialized AI outcompete human polymaths? Is quality of life still improving - and how confident are we in those measures? Is it too late to steer away from the attention economy? Can our control over intelligent systems scale as we develop their power? Will AI ever be capable of truly understanding human values? And if we reach that point, will it choose to align itself?
Holden Karnofsky is a Member of Technical Staff at Anthropic, where he focuses on the design of the company's Responsible Scaling Policy and other aspects of preparing for the possibility of highly advanced AI systems in the future. Prior to his work with Anthropic, Holden led several high-impact organizations as the co-founder and co-executive director of charity evaluator GiveWell, and one of three Managing Directors of grantmaking organization Open Philanthropy. You can read more about ideas that matter to Holden at his blog Cold Takes.
Further reading:
Staff
Music
Affiliates

2,672 Listeners

1,705 Listeners

26,380 Listeners

4,270 Listeners

2,461 Listeners

1,532 Listeners

907 Listeners

124 Listeners

4,025 Listeners

941 Listeners

4,167 Listeners

101 Listeners

551 Listeners

18 Listeners

147 Listeners