DeepSeek, every mention
90 scenes (2025), the whole family · ← back to DeepSeek
tap a year for its mentions
every year 2025 anyone Alex Kantrowitz 209Ranjan Roy 23Grace Shao 21M.G. Siegler 17Dwarkesh Patel 4Reed Albergotti 3Dylan Patel 3Condoleezza Rice 3Amjad Masad 3Stephen Morris 2
Verbatim, from the transcripts: the passages where DeepSeek comes up
Sam Altman: How OpenAI Wins, ChatGPT’s Future, AI Buildout Logic, IPO in 2026?
- ▶ 1:15 Sam Altman This happened to us in the past that happened earlier this year with deep seek. 2 times in the scene
Capital One's Prem Natarajan: Why We're Building Our AI From The Ground Up
- ▶ 28:44 Alex Kantrowitz And, uh, and one of the things that I've been thinking about is, you know, when this deep seek moment happened, which it was like, okay, open source had not exceeded the closed source models, but it seemed like it had caught up to the… 2 times in the scene
Groq and Cohere CFOs: Can Artificial Intelligence Be Profitable?
- ▶ 6:00 Alex Kantrowitz I can't believe it's been this year, uh, that deep sea came out and seemed to drive the cost of reasoning models down, uh, by something like a 40th of what open AI was, uh, was offering.
Google Pushes OpenAI, Bezos Returns, AI’s No. 1 Hit
- ▶ 6:31 Ranjan Roy Their entire, across the entire industry right now, like that, is this a deep seek moment the way Google actually built Gemini three?
Can We Trust Silicon Valley With Superintelligence? — With Nick Clegg
- ▶ 34:17 Sir Nick Clegg And that I see Chesky the other day saying that Airbnb rely very heavily on deep seek for their own work and so on.
AGI or Bust, OpenAI’s $1 Trillion Gamble, Apple’s Next CEO?
- ▶ 8:07 Ranjan Roy I saw some paper around, I think it was called like a tiny recursive model can actually achieve very simple, similar results as deep seek.
- ▶ 15:58 Ranjan Roy And, and obviously, I think that's why Deep Seek was such a moment, because it was like, okay, completely separate teams and people actually are
Is Generative AI a Cybersecurity Disaster Waiting to Happen? — With Yinon Costica
- ▶ 38:52 Alex Kantrowitz And also you had a very interesting report on deep seek when it came out.
- ▶ 39:07 Alex Kantrowitz Um, let's talk about, uh, DeepSeek right away, because there was an interesting release that Wiz put out when DeepSeek was, um, sort of all the rage, and it got a lot of headlines. 16 times in the scene
Vibe Coding: Everything You Need To Know — With Amjad Masad
- ▶ 39:06 Amjad Masad I mean, Deep Seek showed that the models are getting more efficient, and if, you know, Deep Seek open source was able to make it, you better believe that the labs are also getting more efficient. 2 times in the scene
- ▶ 39:40 Alex Kantrowitz And of course, this deep seek moment was a big moment where we found out that this seeming, uh, small hedge fund in China with some GPUs was able to engineer 5 times in the scene
- ▶ 55:03 Amjad Masad And I think that really matters, you know, between O-one preview and, and Deep Seek was like two or three months, um, between, I mean, the biggest one was, was this Kimmy K-II one that we just talked about that, that was like maybe nine…
Anthropic CEO Dario Amodei: AI's Potential, OpenAI Rivalry, GenAI Business, Doomerism
- ▶ 39:19 Dario Amodei Like if we talk about Deep Seek, I don't think it mattered that Deep Seek is open source.
Countdown to GPT-5, OpenAI’s Stargate Sputters, AI Math Wars
- ▶ 26:32 Stephen Morris We remember the other thing that happened that week in January when I was in Davos was DeepSeek's release, which was built on the top of other open source products, and they basically said you can be
Are AI's Economics Unsustainable? — With Ed Zitron
- ▶ 53:26 Alex Kantrowitz I mean, the, the, the folks who built DeepSeq, it seems like that was a big part of the way that they were able to make it cheaper. 2 times in the scene
- ▶ 53:50 Ed Zitron Perplexity releasing, like, a 1776 version of R-One without the Tiananmen Square thing.
Grok's AI Lovebot, Aqui-Hire-Sition Backlash, OpenAI's ChatGPT Agent Debuts
- ▶ 44:11 Alex Kantrowitz Uh, moonshot, this Chinese lab claimed the model outperforms mainstream open source models in some areas, including deep seeks V three and rival capabilities of leading us models, such as those from anthropic and certain functions as… 3 times in the scene
- ▶ 50:34 unnamed speaker I don't think this is as exciting as DeepSeq for me.
Condoleezza Rice: U.S. Tech at Risk Amid University Cuts, China Threat
- ▶ 0:37 Alex Kantrowitz You can make an argument that they're leading in humanoid robots, and they just came for our AI industry with DeepSeq.
- ▶ 3:50 Condoleezza Rice Well, our labs are closed for commercial purposes, but I think when you look at the amount of work that is done at the frontiers of these technologies in universities, ah, we, we publish just about everything openly, and as you know, many… 3 times in the scene
Dwarkesh Patel: AI Continuous Improvement, Intelligence Explosion, Memory, Frontier Lab Competition
- ▶ 23:01 Dwarkesh Patel I mean, deep seek, we know it's trained on five million dollars supposedly, and it's better than GPT four, right?
- ▶ 1:04:36 Dwarkesh Patel Um, I don't know what the mood is now that deep seek has made such a big splash, uh, whether that's changed people's minds. 2 times in the scene
Sam Altman’s Gentle Singularity, Zuck’s AI Power Play, Burning Of The Waymos
- ▶ 31:05 Alex Kantrowitz Uh, they were surpassed by Deep Seek, and then Lama four was not up to expectations.
- ▶ 39:59 Alex Kantrowitz Did DeepSeqs signal a top for services like this?
Mass AI-Driven Unemployment, NVIDIA Surges, CEOs' AI Avatars Unleashed
- ▶ 21:30 Ranjan Roy Could change, and it appears to be changing, but as of today, May 30th, twenty-twenty-five, everyone's playing the same game, and it's a compute game, so that's definitely not changing, and deep seek, but that, this is a six to 12 to 24… 7 times in the scene
- ▶ 23:32 Alex Kantrowitz There's a new upgrade to DeepSeq's R-one model that the company says, uh, reasons better and hallucinates less.
- ▶ 25:30 Alex Kantrowitz And let me ask you this as deep seek, um, continues to rise. 2 times in the scene
- ▶ 32:50 Alex Kantrowitz And after Deep Seek, we haven't seen anything like that or less, we've seen less of that.
Google DeepMind CTO: Advancing AI Frontier, New Reasoning Methods, Video Generation’s Potential
- ▶ 26:48 Alex Kantrowitz We saw a deep seek, uh, push the, the, um, state of the art forward.
Is ChatGPT The Last Website?, Grok’s System Prompt, Meta’s llama Fiasco
- ▶ 27:20 Alex Kantrowitz And let's say, you know, this shows that this could definitely show up in a deep seek or a model that comes from a different country or a place with a different, different values than you, as opposed to one at home.
Are We at the End of Ai Progress? — With Gary Marcus
- ▶ 29:53 Gary Marcus So DeepSeek came out and OpenAI dropped its prices quite a bit. 2 times in the scene
- ▶ 38:39 Alex Kantrowitz We talked about the fact that, like, AI can now walk virologists through, uh, how to create or enhance the function of viruses, and we're starting to see some of these AI programs, like you mentioned, Deep Seek, be available to everybody,…
Generative AI 101: Tokens, Pre-training, Fine-tuning, Reasoning — With SemiAnalysis CEO Dylan Patel
- ▶ 26:44 Alex Kantrowitz So I think one of the things that people focused on with deep seek was that it was just able to be much more efficient in the way that it generates answers. 6 times in the scene
- ▶ 29:14 Dylan Patel And likewise, when we look at, uh, from GPT-IV to DeepSeq VIII, uh, it's fallen roughly 600 X in cost. 2 times in the scene
- ▶ 33:46 Dylan Patel Like, Deep Seek is, like, much cheaper, it's much more efficient, um, but it's roughly the same capabilities as GPT-IV.
Why China's AI Strategy Just Might Work — With Grace Shao
- ▶ 0:00 Alex Kantrowitz Was DeepSeq an outlier or a sign that China's AI industry is a force to be reckoned with and a potential global leader in artificial intelligence? 2 times in the scene
- ▶ 0:57 Alex Kantrowitz Alright, let's start off with this question that we've been wondering about here, which is whether Deep Seek was an outlier, whether it was just a smart group of people who came up with a real AI innovation, or the tip of the spear,… 6 times in the scene
- ▶ 8:23 Grace Shao So now for China on top of all of that, and then Deepsea completely broke the moat where like said and told everyone like, look, you don't actually need to pay for LLMs.
- ▶ 12:54 Grace Shao In fact, they're the only and first one that have actually integrated Deep Seek into their products, um, into their WeChat, which, which obviously, um, all of you know about. 5 times in the scene
- ▶ 15:53 Alex Kantrowitz So I, I wanna, I don't wanna let my original question go without going a little bit deeper into it, asking whether DeepSeq was an outlier, whether there's more, uh, that China is doing. 8 times in the scene
- ▶ 24:54 Grace Shao We have Kimmy, we have Doubao, we have like, you know, we have, uh, like obviously DeepSeep directly, et cetera. 6 times in the scene
- ▶ 26:29 Alex Kantrowitz Huawei used the R-one model to upgrade its Siri-like assistant.
- ▶ 36:53 Alex Kantrowitz You know, it is, it does seem to me, and I think this is important to talk about, that there has been a, after DeepSeek, like, you know, people saw DeepSeek, and then we started talking about a U.S. context. 4 times in the scene
- ▶ 51:14 Alex Kantrowitz Um, outside of Deep Seek. 3 times in the scene
- ▶ 57:03 Alex Kantrowitz Some people say, is this the next deep seek? 7 times in the scene
Google Cloud CEO Thomas Kurian on AI Competition, Agents, And Tariffs
- ▶ 8:22 Thomas Kurian Uh, Lama, Mistral, Deep Seek, uh, a variety of them.
- ▶ 26:26 Alex Kantrowitz You mentioned deep seek. 2 times in the scene
OpenAI Raises $40 billion, Is AI a Letdown?, Musk Sells X to xAI
- ▶ 17:09 Ranjan Roy We saw it with the deep seek effect itself that it should be getting more efficient and cheaper.
- ▶ 20:05 Alex Kantrowitz SimilarWeb also says Anthropix Claude reached 3.3 million average daily visits in March, and DeepSeek had 16.5, uh, in this million visits in the same month.
- ▶ 32:25 Ranjan Roy And obviously DeepSeek created the entire conversation around the need to, or the potential around open weight models.
OpenAI’s Ghibli Moment, CoreWeave's IPO Letdown, End of Silicon Valley’s Monopoly?
- ▶ 11:56 Alex Kantrowitz I think they do feel pressure from places like deep seek from other open source AI companies.
- ▶ 13:18 Brian McCullough Their, their thunder had been stolen by Deep Seek, by others, especially in the AI and, and tech communities. 2 times in the scene
- ▶ 24:21 Alex Kantrowitz I mean, you know, we saw that, you know, the mod, the models had become more efficient, and then, you know, we had this deep seek moment.
- ▶ 51:30 Alex Kantrowitz I mean, deep seek, and we're going to have a China episode coming up as well, going into all the different, um, and, and don't sleep on the Middle East because you're, you're right.
AI's Rising Risks: Hacking, Virology, Loss of Control — With Dan Hendrycks
- ▶ 29:59 Alex Kantrowitz Um, or even, let's say you just take an open source model, like Deep Seek, download it, and then run it without safeguards. 2 times in the scene
- ▶ 36:56 Alex Kantrowitz Maybe not DeepSeq. 2 times in the scene
- ▶ 46:33 Alex Kantrowitz Then, of course, we see the release of Deep Seek and some other AI applications from China, and everyone's trying to build the better AI so that they have the soft power, like we spoke about, uh, to effectively, you know, a control, like,…
- ▶ 1:11:46 Alex Kantrowitz I mean, we just saw Deep Seek, uh, not to, you know, go back to it all the time, but it effectively equaled the cutting edge at,
Why Can't AI Make Its Own Discoveries? — With Yann LeCun
- ▶ 10:19 Alex Kantrowitz And, um, when DeepSeek came out, one of the big screenshots that was going around was someone asking DeepSeek for a novel insight on the human condition. 2 times in the scene
- ▶ 54:34 Alex Kantrowitz Okay, Jan, I have to get you out of here where we are over time, but can you give me, like, 60 seconds, your reaction, uh, to DeepSeek and sort of has open source overtaken the proprietary models at this point? 3 times in the scene
Gruber Slams Apple's Siri Failure, Tech Stocks In Freefall, Vibecoding & OpenAI's Agents
- ▶ 49:12 Reed Albergotti I tried, I tried Deep Seek. 3 times in the scene
Roblox CEO: We Want AI To Generate Full Games
- ▶ 20:21 David Baszucki We, we have also gotten very good at running relatively complex models efficiently, and, and I, uh, you know, deep seeks in the news that, that's kind of, you know,
How Amazon Rebuilt Alexa From The Ground Up
- ▶ 11:07 Alex Kantrowitz It's been part of what DeepSeek DeepSeq has used to become much more efficient, uh, in its, in its reasoning, for instance, because instead of lighting up the entire large language model it's is deciding to light up certain areas. 2 times in the scene
OpenAI's Chief Research Officer on GPT 4.5's Debut, Scaling Laws, And Teaching EQ to Models
- ▶ 3:37 Alex Kantrowitz There's a lot of buzz about deep seek.
- ▶ 9:41 Alex Kantrowitz Uh, but, uh, it makes me think of deep seek. 4 times in the scene
Will Big Tech CEOs Win Their All-In Bet on Trump? With Saagar Enjeti
- ▶ 27:09 Alex Kantrowitz So on this show, we've been talking a lot about, uh, DeepSeek from a technology, from a technology standpoint. 2 times in the scene
- ▶ 27:38 Alex Kantrowitz So Sagar, I'm curious, you know, you've been dissecting what deep seek means for, uh, us global AI policy. 8 times in the scene
- ▶ 33:26 Alex Kantrowitz Yes, they definitely trained with a lot more money than they said they did to build this R one model. 2 times in the scene
Boom Times For ChatGPT, OpenAI’s Deep Research, AI Super Bowl
- ▶ 7:29 Alex Kantrowitz DeepSea came out. 10 times in the scene
- ▶ 13:51 Alex Kantrowitz So, uh, it is really clear that DeepSeek put them on their heels, and they said as much in this AMA that included Sam Altman, CEO of OpenAI. 10 times in the scene
- ▶ 17:28 Ranjan Roy Yeah, no, I mean, on one hand, recognizing and trying, the rational point of view would be, they see Deep Seek, they see R-one, and they're just recognizing that this is the state of the industry, potentially we need to go open source,… 3 times in the scene
- ▶ 23:00 Ranjan Roy But in reality, I still, I think DeepSeq shows, has shown us, and that number is not five or six million dollars to actually build and train the whole model, but the actual training part of it was six million dollars.
- ▶ 32:45 Alex Kantrowitz Um, and you know, it says AI spend sores among deep seek challenge. 4 times in the scene
- ▶ 36:30 Alex Kantrowitz In months, we've seen dramatic improvements from OpenAI's O-one family to their new O-three models, and that's where DeepSeq factors in, and DeepSeq has its R-one model that everyone went crazy about. 2 times in the scene
- ▶ 46:34 Ranjan Roy And so, it's down 12% since the DeepSeek announcement, and it's back up a little bit. 2 times in the scene
NVIDIA VP Rev Lebaredian Talks Plan To Build AI That Understands The Real World
- ▶ 5:51 Rev Lebaredian Um, so what, what we've seen here with, with DeepSeq is, is a great advancement that's on that same curve that we've been on for, for a decade now.
Reid Hoffman: Why The AI Investment Will Pay Off
- ▶ 17:59 Reid Hoffman Um, there's, you know, Deep Seek and, and others within China, and so the, the competitive sphere,
- ▶ 21:08 Alex Kantrowitz Um, let's just talk quickly about deep seek since you brought it up. 4 times in the scene
How DeepSeek Changes AI Research & Silicon Valley w/ M.G. Siegler
- ▶ 0:00 Alex Kantrowitz It's time for a bonus episode exclusively about DeepSeq R-One as the Chinese open source AI model roils markets and threatens to upend the generative AI industry. 3 times in the scene
- ▶ 0:16 Alex Kantrowitz We're doing a bonus edition today exclusively on DeepSeq, what it means for the AI industry, what it means for markets. 4 times in the scene
- ▶ 4:25 M.G. Siegler For all different sorts of things, you know, you, you noted how much cheaper it is, um, to run than say open AI's models and, and, you know, over the weekend, just reading all of these sort of reports about the model and how many, um,… 5 times in the scene
- ▶ 6:25 Alex Kantrowitz And you mentioned that some startups are already swapping in, uh, deep seek R one for the models they're using right now. 3 times in the scene
- ▶ 10:59 Alex Kantrowitz So, MG, could you tell us just at a really high level, what DeepSeek has done to be able to get these results? 4 times in the scene
- ▶ 12:00 M.G. Siegler And, you know, they've just used the process of distillation to, you know, effectively bring those bigger versions of the, the, um, sort of state-of-the-art models and distill them down into, um, you know, smaller models, which eventually… 2 times in the scene
- ▶ 15:41 Alex Kantrowitz And now what DeepSeek has shown is that you can actually do all this without that. 4 times in the scene
- ▶ 20:34 Alex Kantrowitz And I, and let's get into the business and I'm, I'm smiling here because you're making me think of, we have Reid Hoffman on the show on Wednesday and I interviewed him before our one came out.
- ▶ 25:16 M.G. Siegler Meta is more interesting because their open source philosophy, open weight philosophy and model is so similar to what Deep Seek has done, right? 5 times in the scene
- ▶ 32:48 Alex Kantrowitz Like you look at the chain of thought, even in deep seek, and you're just like, how is a computer, you know, quote unquote thinking through this stuff, but the economy needs to take hold of this powerful technology and make use of it and… 3 times in the scene
- ▶ 38:56 Alex Kantrowitz And that's effectively what deep seek is going to put to the test. 3 times in the scene
Google DeepMind CEO Demis Hassabis: The Path To AGI, Deceptive AIs, Building a Virtual Cell
- ▶ 52:41 Alex Kantrowitz I mean, Deep Seek is this new model getting-