All Segments
Everyone Was Wrong About Intelligence – Dario Amodei (Anthropic CEO)
5 segments available
Full Episode: https://youtu.be/Nlkk3glap_U (August 2023) Transcript: https://www.dwarkeshpatel.com/dario-amodei Apple Podcasts: https://apple.co/3rZOzPA Spotify: https://spoti.fi/3QwMXXU Follow me on Twitter: https://twitter.com/dwarkesh_sp
Segments Timeline
The Unpredictability of Intelligence Predictions
Dario Amodei discusses the challenges of predicting advancements in AI models and intelligence. He reflects on the unpredictability of commercial breakthroughs and the varying capabilities of models across different tasks, emphasizing that intelligence is not a simple spectrum but a complex array of skills and expertise.
"I feel like these scaling laws have been very predictable but then when you say like well you know when when is there going to be a commercial explosion in these models or what's the form it's going t..."
Rethinking Intelligence as a Spectrum
Amodei challenges the traditional view of intelligence as a linear spectrum. He argues that intelligence encompasses various domains and skills, highlighting the surprising breadth of human capabilities and the limitations of current AI models in replicating these diverse skills.
"for different tasks right like you know like write write a sonnet you know in the style of cor MC McCarthy or something like I don't know I'm not very creative so I couldn't do that but like you know ..."
The Discrepancy in AI Performance
In this segment, Amodei reflects on the impressive capabilities of AI models like GPT-3, noting their ability to perform complex tasks while still lacking true human-level intelligence. He discusses the paradox of AI's vast knowledge versus its inability to make novel connections or discoveries.
"spectrum is also wide if you asked me 10 years ago that's not what I would have expected at all but uh I think that's very much the way it's turned out one thing that's been surprising is like I thoug..."
The Limits of AI Creativity
Amodei explores the concept of creativity in AI, acknowledging that while models can generate creative outputs, they have yet to achieve significant scientific breakthroughs. He emphasizes the need for higher skill levels in AI to leverage their extensive knowledge effectively.
"in the current generation and potentially for generations to come what explains discrepancy between super impressive performance in these benchmarks and in just like the things you could describe vers..."
Biology vs. Physics: The Complexity of Discovery
In this concluding segment, Amodei contrasts the fields of biology and physics regarding AI's potential for discovery. He argues that while AI models possess vast knowledge, their current skill levels may not be sufficient to synthesize this information into meaningful discoveries, particularly in the complex domain of biology.
"four or more orders of magnitude of data if you compare to you know number of words human a human sees as they're developing to age 18 we have to admit that that's a weird thing that doesn't match up ..."