PodParley Podparley
PodParley Podparley

Are we ready for human-level AI by 2030? Anthropic's co-founder answers

Anthropic's co-founder and chief scientist Jared Kaplan discusses AI's rapid evolution, the shorter-than-expected timeline to human-level AI, and how Claude's "thinking time" feature represents a new frontier in AI reasoning capabilities.

Listen to this episode

0:00 / 0:00

Summary

Anthropic's co-founder and chief scientist Jared Kaplan discusses AI's rapid evolution, the shorter-than-expected timeline to human-level AI, and how Claude's "thinking time" feature represents a new frontier in AI reasoning capabilities.

First published

04/01/2025

Genres

technology business investing news

Duration

52 minutes

Parent Podcast

Azeem Azhar's Exponential View

View Podcast

Share this episode

Similar Episodes

No similar episodes found

Similar Podcasts

No similar podcasts found

Episode Description

<p>Anthropic's co-founder and chief scientist Jared Kaplan discusses AI's rapid evolution, the shorter-than-expected timeline to human-level AI, and how Claude's "thinking time" feature represents a new frontier in AI reasoning capabilities.</p><p><strong>In this episode you'll hear:</strong></p><ul><li>Why Jared believes human-level AI is now likely to arrive in 2-3 years instead of by 2030</li><li>How AI models are developing the ability to handle increasingly complex tasks that would take humans hours or days</li><li>The importance of constitutional AI and interpretability research as essential guardrails for increasingly powerful systems</li></ul><p><strong>Our new show </strong></p><p>This was originally recorded for "Friday with Azeem Azhar", a new show that takes place every Friday at 9am PT and 12pm ET on Exponential View. You can tune in through my Substack linked below. The format is experimental and we'd love your feedback, so feel free to comment or email your thoughts to our team at <a href="mailto:[email protected]">[email protected]</a>.</p><p><strong>Timestamps:</strong></p><p>(00:00) Episode trailer</p><p>(01:27) Jared's updated prediction for reaching human-level intelligence</p><p>(08:12) What will limit scaling laws?</p><p>(11:13) How long will we wait between model generations?</p><p>(16:27) Why test-time scaling is a big deal</p><p>(21:59) There’s no reason why DeepSeek can’t be competitive algorithmically</p><p>(25:31) Has Anthropic changed their approach to safety vs speed?</p><p>(30:08) Managing the paradoxes of AI progress</p><p>(32:21) Can interpretability and monitoring really keep AI safe?</p><p>(39:43) Are model incentives misaligned with public interests?</p><p>(42:36) How should we prepare for electricity-level impact?</p><p>(51:15) What Jared is most excited about in the next 12 months</p><p><strong>Jared's links:</strong></p><ul><li>Anthropic: <a href="https://www.anthropic.com/">https://www.anthropic.com/</a></li></ul><p><strong>Azeem's links: </strong></p><ul><li>Substack: <a href="https://www.exponentialview.co/">https://www.exponentialview.co/</a></li><li>Website: <a href="https://www.azeemazhar.com/">https://www.azeemazhar.com/</a></li><li>LinkedIn: <a href="https://www.linkedin.com/in/azhar?originalSubdomain=uk">https://www.linkedin.com/in/azhar</a></li><li>Twitter/X: <a href="https://x.com/azeem">https://x.com/azeem</a></li></ul><p>Produced by supermix.io</p><br/> <p>Hosted by Simplecast, an AdsWizz company. See <a href="https://pcm.adswizz.com">pcm.adswizz.com</a> for information about our collection and use of personal data for advertising.</p>

just now