Behind the DeepSeek Hype, AI is Learning to Reason

Your Undivided Attention

Center for Humane Technology

Tristan Harris, Socialjustice, Tech Podcast, Character Ai, Little Tech, Ai History, Silicon Valley, Privacy, Daniel Barcay, Addiction, Ai Addiction, Chat Bots, Children And Tech, Tech Policy, Responsibleai, Tech, New Ai Shows, Screen Time, Open Ai, Elections, Kids Tech, Google, Ai And Kids, Politicsandai, Politics, Anthropic, Dataprivacy, Humans, Tech And Relationships, Us Politics, Ai And Relationships, Aiandhumanrights, Civictech, Aiinsociety, Surveillance, Sam Altman, Technopoly, Humancenteredai, Breakdown Of Trust, Ai And Work, Ai And The Future, Democracy, Futureofwork, Tech Politics, Tech Ethics, Future, Tech Addiction, Asi, Kids Phone Addiction, Best Ai Shows, Ai Regulations, Meta, Digitalgovernance, Bigtech, Ai And Happiness, Machinelearning, Screentime, Relationships, Ai Welfare, Ai Podcast, Cognitive Liberty, Infinite Scroll, Ai And Education, Kids And Ai, Ai Politics, Apple, Digitaldemocracy, Claude, Llms, Societalimpact, Artificial General Intelligence, Agi, Machines, Us Society, Politicaltechnology, Disinformation, Ai And Rights, Elon Musk, Government, Aiaccountability, Polarization, Jon Haidt, Algorithmicbias, Ai Personhood, Kids Online Safety, Superintelligence, Techandsociety, Automation, Design Ethics, News, Time Well Spent, Tech News, Society & Culture, Humane Design, Technology, Cht, Artificial Intelligence, Center For Humane Technology, The Social Dilemma Netflix, Philosophy, Human Downgrading, Aza Raskin, Attention Economy, Ethical Technology

4.8 • 1.5K Ratings

🗓️ 20 February 2025

⏱️ 32 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

DeepSeek's breakthrough in AI sent markets reeling. But behind the headlines lies a crucial shift: AI that can actually reason and think. As labs race toward self-improving AI, the real question isn't how fast we can go, but how to steer this power for the benefit of us all.

Transcript

Click on a timestamp to play from that location

0:00.0	Hey everyone, it's Aza. Welcome back to your undivided attention. So today we are going to be doing actually a bit of a special episode. It's going to be me here with our co-founder Randy Fernando, who was at Nvidia for seven years. And what we really want to do
0:22.3	is give you some insights into the latest set of AI models that came out. So these are
0:28.8	Open AIs 03, Deep Seaks R1, and actually they're following on Open AIs 01 from a little, a couple
0:36.3	months ago. And we want to talk about what makes them a big deal,
0:41.3	why we have switched into a new paradigm
0:44.1	and how these models get trained
0:45.5	and what's going on behind the scenes.
0:47.7	So first, Randy, thanks for joining me.
0:50.1	Glad to be here.
0:51.5	First place to start is, you know,
0:52.9	this new model from China, deep sea car one, it dropped.
0:56.1	And it ended up creating this frenzy in media.
1:00.0	It shook global markets.
1:02.9	The hype is quieted down.
1:04.3	And actually, you know, I think that the drop in global markets was very irrational.
1:09.8	But let's talk a little bit now about what makes this a key inflection point in AI tech.
1:17.6	I think there were several things, right?
1:20.4	And I'm not sure exactly which order to go, but I'll just name a few.
1:25.0	One was low-cost, high-performance reasoning. Like, it actually performed well and people
1:30.8	used it and that was really impressive. Now, there are some asteris about the cost because
1:38.0	the cost didn't account for the GPUs, the salaries.
1:42.0	Just to jump in, there's a widely reported number that between $5 to $6 million,
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from Center for Humane Technology, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of Center for Humane Technology and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.