China's AI Upstarts: How Z.ai Builds, Benchmarks & Ships in Hours, from ChinaTalk

This podcast is sponsored by Google. Hey folks, I'm Ammar, product and design lead at Google DeepMind. We just launched a revamped Vibe Coding Experience in AI Studio that lets you mix and match AI capabilities to turn your ideas into reality faster than ever. Just describe your app and Gemini will automatically wire up the right models and APIs for you. And if you need a spark, hit I'm feeling lucky and we'll help you get started. Head to ai.studio slash build to create your first app. Hello, and welcome back to the Cognitive Revolution. Today, I'm honored to share a special crosspost from the China Talk podcast, hosted by Jordan Schneider, China Talk analyst Irene Zhang, and Nathan Lambert of AI2 and the Interconnects Substack, featuring a conversation with Zixuan Li, Director of Product and Gen AI Strategy at ZAI, also known as Zipu AI, about the culture, incentives, and constraints shaping Chinese AI development. Now, I imagine that many, even in our AI-obsessed audience, will not be familiar with ZAI, but their model releases demonstrate that they are a significant player worthy of our attention. As of today, their latest GLM 4.6 model holds the number 19 spot on the LM Arena text leaderboard. Its ELO rating is roughly 65 points behind the current leaders, which means that it still wins two out of five head-to-head comparisons with the leaders, and it happens to sit right next to Quinn 3 Max, Kimi K2 Thinking, and DeepSeek's V3.2. Together, these four models, all from China, are indeed the top four open source models available today, though it should be noted that Mistral is not too far behind. On the web development leaderboard, GLM 4.6 does even better, coming in at number nine, making it competitive with GPT 5.1, and meaningfully behind only Gemini 3 and the new Claude 4.5 opus. All that said, this conversation goes way beyond benchmarks and touches on a number of important topics, including why we should understand Chinese companies' open-weight strategy not so much as an ideological commitment, but as a practical marketing tactic from companies that are seeking to gain global mindshare while recognizing that Western enterprises simply can't and won't use their APIs, the culturally distinct AI use cases such as role play that matter in China, and how these drive different fine-tuning priorities than what we typically see from Western companies. The role that Silicon Valley thought leaders play in establishing credibility for Chinese companies, even in their home market. The market for AI talent in China, and why very few people at ZAI even know that Zixuan studied at MIT. Zixuan's view that there is a wall and that further architectural breakthroughs will be needed. The extreme velocity with which ZAI releases models, which often involves shipping within hours of completing training, what, if anything, people in China fear about the AI future, and how Chinese companies generally still view themselves not as peers or rivals with leading American companies, but as upstarts that are just trying to keep pace and who will be quite happy if they can secure for themselves a meaningful niche in the global AI marketplace. This is a fascinating, detail-rich conversation, and regardless of your attitude on U.S.-China competition, it's clear to me that we in the West don't hear nearly enough from the actual builders inside Chinese AI labs. So I appreciate Jordan for allowing me to cross post this episode. And of course, I encourage everyone to subscribe to China Talk online at Chinatalk.media. With that, I hope you find as much value as I did in this behind the scenes look at frontier AI development in China with Zishuan Li of ZAI from the China Talk podcast. this is this jenny who studied uh in the u.s before moving back to china works at drippu ai or z.ai we're gonna let him introduce himself and his role co-hosting today irene jog a long-time china talk analyst as well as nathan lambert of ai2 and the interconnects substack welcome to trying to talk everyone so personally i've known of jipu ai or z.ai for at least over a year and i've been following the work there and then the summer it was like kind of in my mind was like well there's another deep siege when they released glm 4.5 um you'll have to correct me on whether or not it was before or after the kimmy k2 model i guess is after and that was like this the sequence of order of weeks or days after kimmy k2 and it's just like wow like okay there's a lot of people building great models. And it's fun to get to learn about some of them and how it compares between US labs and Chinese labs. And I think a lot of it is more in common than different. Hi, everyone. I'm Zixuan Li from ZDAL AI. And I manage a lot of stuff like global partnerships, ZDAL AI chat, model evaluation, and our API services. So if you have heard of GLM Coding Plan, I'm actually in charge of this thing too. Yeah, nice to meet you, everyone. Yeah, this is the case for introducing yourself. I'd love to hear more about how you ended up working in AI after moving back to China and AI specifically. Yeah, actually, I applied for multiple roles and companies like Moorshot, like other Minimax, but got rejected or got neglected because there are so many resumes going to them every day. and I studied AI for science and also AI safety at MIT. So I do a lot of research on the application and alignment. That's not very relevant to what we're doing right now, but actually it gives me a sense of what's going on in the frontier area. So it helps me a lot, like getting aspects in what OpenAI, anthropics are doing like at that time i think it's it's very innovative to have these sorts of idea and experience was it ever a debate for you whether to stay um in the u.s or did you always know you were going back to china after grad school yeah i already know like i'm going back because my family's here. Yeah, but I got the job after the graduation because it's hard to get a job. I continuously apply for jobs, but finally after one month, yeah, I got an opportunity to interview with ZDAL AI. And at that time, I'm not in charge of the oversee department because our focus is in the domestic area. Domestic chatbot. So I'm responsible for the strategy of developing a domestic chatbot. It's called ChatGLM. Gotcha. So maybe let's do a little bit of Dripoo AI backstory. Kind of when was it founded? How would you place it within the broader landscape of folks or teams developing models in China? Great. So the Junfuan and also like ZDAL AI was founded in 2019. And we are chasing AGI at that time, but not with LLMs, but with some graphic network or graphic compute. So we did a thing like Google Scholar and it's called A-minor. So we use that type of thing to connect all the data resources on the journals and research papers and into a database. And people can easily search and map these scholars and their contributions. So it's very popular at that time. But we shipped it to exploration of large language models in 2020. and we launched our paper GLM in 2021. So that's, I believe, one year ahead of the launch of GPT 3.5. So very, very early stage. And we were one of the first companies to do the exploration of large language models. And after that, we continuously improved the performance of our models, tried new architecture. GLM is a new architecture actually, but we are going to explore more in the future. And I believe that we got famous by the launch of GLM 4.5 and also 4.6 because I think it's very capable in coding, reasoning and agentic tool use. So yeah, that's more useful compared to the previous version and like people may know us through clock code kilo code and other tools so we need to like combine with these top products and get us famous yeah um yeah let's talk a little bit more about the the evolution of uh glm 3.5 um sort of i don't know nathan this is your question why am i asking this question well it's like okay what does it take to transition from the models that you were early to this to things that get international recognition so like i have known of z.ai at your work for years and then it's like snap of the fingers and you're like okay now this model is on everybody's radar that's paying attention and and does this feel like something that was just going to happen for you overnight in developing the models or any like like what What does that feel like when you go through it or how do you get to that moment? Because there's a lot of people that want to do that at their companies. Yeah, that's a very, very interesting point because in 2024, everyone is interested in chatbot arena, right? Like we see GBV4, we should see Gemini performing very well on chatbot arena. So that's our interest because we pay attention to end users' experience. When there are two answers, which one you prefer. So we did a lot of things on that and we did perform very well on Chapel Arena. So ranked maybe sixth or like sixth to ninth on Chapel Arena. But in 2025, with the launch of MANA's Cloud Code, we realized that coding and genetic stuff are more useful or they contribute more economically and also in terms of the efficiency for people. so chat I think we're no longer like put that as our top priorities instead we do more exploration on the coding side on the agent side so we observe the trend and we do a lot of experiments on it yeah we need to follow the trend and also predict the future do you feel like you're better at executing for code versus chatbot arena because like like i like glm 4.5 and like i i think the air version and 4.6 are like extremely renowned for this and i think like the the when you train these models the process can look very similar depending on what your target is and i'm just wondering of like do you feel like it was a shift or it's just like sometimes things work out better than others it's a shift actually yeah we we pay more attention to the coding stuff. And like on the ZDW chat, we are free, right? So nobody's paying for the chat. Yeah, people pay for cloud code use and for agentic stuff. But we just let users chat with the chatbot freely. Yeah, so that's a shift. But we need to like continuously improve the performance in like normal chat. And maybe role-playing, but not our top priority. Hey, we'll continue our interview in a moment after a word from our sponsors. If you're finding value in the cognitive revolution, I think you'd also enjoy Agents of Scale, a new podcast about AI transformation hosted by Zapier CEO Wade Foster. Each episode features a candid conversation with a C-suite leader from companies including Intercom, Replit, Superhuman, Airtable, and Box, who's leading AI across their organization, turning early experiments into lasting change. We recently cross-posted an episode that Wade did with OneMind founder and CEO Amanda Calo about AI-led sales, and I also particularly enjoyed his conversation with John Narona, chief product officer of AI Product Pioneer and recently minted double unicorn, Gamma. From mindset shifts to automation breakthroughs, Agents of Scale tells the stories behind the Enterprise AI wave. Subscribe to Agents of Scale wherever you get your podcasts. Are you still jumping between multiple tools just to update your website? Framer unifies design, content management, and publishing on one canvas. No handoffs, no hassle. Just everything you need to design and publish in one place. Framer already built the fastest way to publish beautiful, production-ready websites, and it's now redefining how we design for the web. With the recent launch of Design Pages, a free canvas-based design tool, Framer is more than a site builder. It's a true all-in-one design platform. From social assets, to campaign visuals, to vectors and icons, all the way to a live site. Framer is where ideas go live, start to finish. And now, they've added a Framer AI layer to make it all faster and easier than ever. With Wireframer, you can skip the blank canvas and get a responsive page with structure and starter content ready to edit. With Workshop, you can create new visual effects, cookie banners, tabs, and more. No coding needed. And with AI plugins, you can connect top models from OpenAI, Anthropic, and Google to generate images, rewrite text, generate alt text, and more. Ready to design, iterate, and publish all in one tool? Start creating for free at framer.com slash design and use code Cognitive for a free month of Framer Pro. That's framer.com slash design. Use promo code cognitive. Framer.com slash design. Promo code cognitive. Rules and restrictions may apply. So yeah, let's talk a little bit about the sort of the talent and the internal culture that allowed you to put out GLM 4.5. What do you think is different about, or what distinguishes Dripway AI from other labs, both in the U.S. and China? I think the first of all, we are more collaborative, like inside the company. So everyone is working on this single target. And maybe we have like some head of separate teams, like pre-trained team, post-trained team. but they are working very closely. They just sit next to each other and working on a single target is trying to build a unified reasoning, agentic, and coding model. So we have built three separate models as we have illustrated in our tech report. And we then distill these three teaching models into one single model is GLM455. So that's our goal. And I believe that's how we built GLM455 more efficiently compared to other companies. And they are super young. And another point is that you mentioned talent. But I believe that nowadays you need to do the research yourself, you need to do the training yourself as the head of the team. So you cannot let others do this stuff for you. Why is that? Because things change really fast. Like maybe during your training there comes QoG4, there comes GBD5, anything can happen. So you need to feel the trend yourself. Like you need to combine the results from experiments, the trends, like what's going on within your competitor's team and to feel the move yourself. It's super important. Like even our founder, he did the experiments himself. He looked at the papers. Yeah, you need to do things simultaneous, not just set goals for people and let others do the stuff for you. Yeah, it's been very fast-paced. I think before we started recording, you were also mentioning that there's a lot of PhD students involved. I was just wondering if these are people that are actively pursuing their PhD or kind of new grads or a mix of all of them. Because I think that I work at a research institute, which is very open source, and we have a lot of full-time students that are part of it. But where you look at other closed labs in the US, there's not nearly as much intermingling with the academic institutions. And I think that could be a really powerful thing if you have this, because there is a lot of extreme talent there. So I'm just wondering if you feel like it's a kind of open door between some academic institutions and your work. Yeah, definitely. There are a lot of ongoing PhD students here. And I believe that they are both chasing their academia and the work of GLM simultaneously. but they can combine them together. If you are doing a really innovative job like training a unified agentic coding model, it's one of your greatest achievements ever. So people won't say, okay, I need to do another research and let me finish this first and then we'll go back to the GLM. They will try to treat GLM as their single biggest achievement. So everyone is really devoted to this stuff. And yeah, we hardly see anyone not devoting to training GOM. Could you talk a little more broadly about the talent market? I mean, you mentioned earlier that you had to put your resume in in a lot of places. What does it look like right now? What's the kind of like hierarchy and what are, you know, what are folks looking, what are employers looking for and what's the talent looking for? So on the research and engineering side, I think they're looking for papers, looking for like GitHub code, looking for competitions. Yeah. And also your experience on using like GPUs, right? Your experience on like training models. But for the non-technical side, they're looking for how you're going to growth the model performance, expand your branding, and also a lot of stuff. If you're going to be a product manager, your live coding skills, your vision on this area, and also how you do the stuff yourself, those are very important. I think it's pretty similar, but you mentioned hierarchy, right? So in terms of hierarchy, large companies choose the people first because they have more money. They can pay more, like Bydes, Alibaba. But for startups, we need people to fight together. You need to fight against other competitors. you need to drive yourself to finish the goals because you don't get paid that much like you need ambition you truly enjoy like working with really young talented people and try to build something like gom like it seems come from nowhere and try to beat other like competitors models. Yeah. How big would you say the T like the um number of people that are actually training the model is I think in the U S it accepted that the core researcher engineering staff normally doesn get to be more than one to 200 people at the likes of like open AI or something. And then there's a lot of support around them in terms of product and distribution and stuff like this. Do you feel like this is similar or there's kind of a core small research team? And a similar, like 100 to 200 people. I think it's, that's enough. Yeah. Because you need to be focused, right? Like there are people preparing data and there are people doing the product stuff. But for the core team, you don't need that much because you need to stay focused. And these people need to be really talented. They cannot make many mistakes. Right. Do you know that's different at bigger companies? I think for bigger companies, there might be different groups. Right. they have like more GPUs and they can do more exploration. For example, like in ByteDance, they are chasing the top performance, not only in text generation, but also in video generation, speech, other areas, so they can allocate the resources to multiple teams. But inside these teams, the core members, I think it's still the same, maybe 10 to 20 and other like 80 or 100 doing the like the training stuff or data preparation um there was a lot made in chinese and and western media about how deep seek like was biased against people who'd studied abroad i'm current i'm curious about the um you know any other broader dynamics you see with relation to sort of returnees versus people who did their whole education in China? I think there's no bias, like because they want the best people. And usually the interviewees like only like stay in China, like perform the best in their interviews. Yeah, They have no bias. But maybe people coming from U.S. or other countries just did worse in their interview. Yeah, I believe that that's not a bias because I think they're judging very well. They have their standards. Maybe their standards are different from what you do around the world. But actually, I believe that that's not an issue. Like even inside China or inside our team is the same, same standard. Because I, like I joined Z.AI after coming back from the US, but I think nobody actually knows. People will never ask like, are you studying abroad or like, have you ever like master degree from MIT? I believe that maybe just 10 people know about this in ZDAR. So there's no bias because people don't care. The worst thing about automation is how often it breaks. You build a structured workflow, carefully map every field from step to step, and it works in testing. But when real data hits or something unexpected happens, the whole thing fails. What started as a time saver is now a fire you have to put out. Tasklet is different. It's an AI agent that runs 24-7. Just describe what you want in plain English, send a daily briefing, triage support emails, or update your CRM. And whatever it is, Tasklet figures out how to make it happen. Tasklet connects to more than 3,000 business tools out of the box, plus any API or MCP server. It can even use a computer to handle anything that can't be done programmatically. Unlike ChatGPT, Tasklet actually does the work for you. And unlike traditional automation software, it just works. No flow charts, no tedious setup, no knowledge silos where only one person understands how it works. Listen to my full interview with Tasklet founder and CEO, Andrew Lee. Try Tasklet for free at tasklet.ai and use code COGREV to get 50% off your first month of any paid plan. That's code COGREV at tasklet.ai. Being an entrepreneur, I can say from personal experience, can be an intimidating, and at times lonely experience. There are so many jobs to be done and often nobody to turn to when things go wrong. That's just one of many reasons that founders absolutely must choose their technology platforms carefully. Pick the right one and the technology can play important roles for you. Pick the wrong one and you might find yourself fighting fires alone. In the e-commerce space, of course, there's never been a better platform than Shopify. Shopify is the commerce platform behind millions of businesses around the world and 10% of all e-commerce in the United States. From household names like Mattel and Gymshark to brands just getting started. With hundreds of ready-to-use templates, Shopify helps you build a beautiful online store to match your brand's style, just as if you had your own design studio. With helpful AI tools that write product descriptions, page headlines, and even enhance your product photography, it's like you have your own content team. And with the ability to easily create email and social media campaigns, you can reach your customers wherever they're scrolling or strolling, just as if you had a full marketing department behind you. Best yet, Shopify is your commerce expert with world-class expertise in everything from managing inventory to international shipping to processing returns and beyond. If you're ready to sell, you're ready for Shopify. Turn your big business idea into cha-ching with Shopify on your side. Sign up for your $1 per month trial and start selling today at shopify.com slash cognitive. Visit shopify.com slash cognitive. Once more, that's shopify.com slash cognitive. Hey, we'll continue our interview in a moment after a word from our sponsors. Let's talk a little bit about open versus closed source in Chinese model developers broadly and with Z.AI in particular. How do you think people, what's the thought process behind so many models going open source in recent years? Yeah, first I think like maybe generally we need to devote more to the research area. like Lama's doing this, Quinn is doing this, and Kim is doing this. We are also doing this. We want to contribute more to the academia and also the exploration of all possibilities. I think that's our top priority, right? But beyond that, as a Chinese company, we need to really be open to get accepted by some companies, right? because people will not use your API to try your models. Maybe they deploy on fireworks, maybe they use it on ROC, and maybe they download to their own chips. So I think it's not easy to get famous in the U.S. States because people just don't accept your API. They need to be stored in the U.S. So I think it's necessary to be open right now for people to use GOR. I mean, this is like, this is what our company does. Like where I work, it's like I wouldn't be able to sign up for the API service of the enterprise, but like I've distilled for multiple Chinese models when I'm training. It's like I'm using multiple models and might come across this. So it's not surprising, but it's good to articulate it. Yeah, we also learn from DeepSeq. because we have closed source version in 2024. Like our flagship model is closed source back then. But when DeepSeek R1 launched, we realized that, oh, you can do this thing simultaneously. You can be really famous for open sourcing your model while get some business return like through API or other stuff or collaboration. yeah you need to expand the cake first and then take a bite of it um maybe taking one step back like why is it so important for chinese model makers to get um i don't know famous in the u.s or just global adoption more broadly yeah because i think there's better ecosystem for developers research still in this United States. Yeah, you need to get accepted by the top researchers or you, right? Because if we don't open source our models, we'll never have opportunity to join this conversation. Yeah, it's also important. Yeah, because we learn from X, from YouTube, Reddit every day and all the Chinese tech media is also paying attention to U.S. KOLs or influencers. This was very surprising, I think, both to Nathan and I, how kind of recursive it was, where the Chinese media covers the models that the Americans are talking about. It's a very curious trend. Yeah, because you have like people like Andrew Kaputh and also Sam Aldama, Elon Musk, they not only talk about their own models, but also what's going on there. So everyone knows. Yeah, if they post a tweet, everyone knows what's going on. what models they're picking, what preferences they have, their views on maybe code versus codecs. All the social media will try to grasp their code idea immediately. So that's very important. We will also learn this from DeepSeek. Frankly speaking, we used to neglect the importance of global economy previously because we think we need to sell our products, sell our APIs directly to Chinese enterprises. But nowadays, Chinese enterprises are still paying attention to the global brand and your global performance. Yeah, this reminded me of something I've been curious about, which is we know the conversation is recursive. We know that the Chinese tech pays attention to what American Silicon Valley is looking at. But is there anything about the AI debate or discourse in China that Western media tends to miss, in your opinion? Are there any kind of issues or debates or things that people are really interested in that people in the English-speaking discourse tend to not understand? Yeah, so I just talked to a professor from Germany yesterday, and he mentioned some models that he knew people are talking about these days, like Lama, Quinn, even Mistral, but not GLM. So there are many people still missing out. out of date in the in the sf circles more people are talking about glm than mistral and arguably llama these days so you've made a lot of progress yeah we've made a lot of progress but we we track we also track um the discussion on reddit other social media but we still see a lot of people talking about what is GOM? Like, is it a good model? Like where it comes from? It comes from nowhere or similar stuff. Yeah, still need to do a lot of things because we only have 20K followers on X. So that's quite few, right? Yeah, so nobody actually get a very deeper understanding of gom compared to other models i think deep seek has like a million it's crazy yeah i like like that's even big for a whole lot of american like like for that it's like for a new american tech company that would even be big it's remarkable yeah also like mistral cohere i think they get much more attention compared to Kimi and Zidaw. Yeah. So we still need to do better in our branding or our engagement in the technical community. You mentioned selling API access to Chinese companies. I don't know. Tell us a little bit about adoption in China, what the sales process is like. Do they all just have VPNs and use cloud code anyways? What's it like trying to do enterprise sales in China? Yeah, so you have two types of enterprises. One is the companies that can't use API because there are companies that need to deploy the model on their own chips, and they cannot accept sending data to other companies, even Zeta or even Alibaba. So that's a requirement. For those companies, they require DeepSeq. So they're a team deploying DeepSeq for them, not from DeepSeq or any companies can deploy DeepSeq for them. And they usually built on top of the DeepSeq model with RAG, like data storage, and workflows and other things. Yeah, and the other one uses APIs, maybe from tech companies and media companies. Yeah, these companies accept API because they need to standardize their workflows. So for the API company, I think they choose between the balance, the performance and the price. So Bydes is doing great in that area. Bydes, I believe, dominates the API services. And also like Queen is still trying to sell their APIs because Queen 3 Max is a close source version. if you have heard of it. So they have open source some models but also keep some thing co-source for selling. For me, we have open source or fraction models so we are always frequently asked what's your service different from the open source version because we can deploy the open source version ourselves so we need better engineering team we need faster decoding speed right so we need to do more on top of just a good model so that might be our unique selling point we need to do searches we need to build our MCP trying to get a competitive advantage over other GLM providers. Is that annoying? I think it's or fun. It's fun. It's fun because I think it's necessary to open source your models. So how you get a bite in that case is really important. Yeah, because we have figured out for a long time. But recently, we found subscription is a good idea. A GLM coding plan. Because with subscription, your users will become more sticky. And yeah, they love this area. Because you don't have to worry about how one prompt consumes in your dialogue. Maybe inside Cloud Code, a round of interaction will consume a million tokens. But you don't have to worry about it. so we'll figure it out for all our users do you think you have meaningful adoption there? because in the US market it's like I could start using Claude Codex Gemini and whatever all for free some basic cursor and that's why I was wondering are people in the US actively, is this a growing market that you think you're going to eat into? because I mean Quinn has one and I might have tried them but I think I'm always like oh, I have my JATGPT subscription. And I'm just wondering if you're like on the ground, if it feels optimistic as something that is like really shifting the needle. It's definitely very optimistic because we don't have to persuade 50% of people doing this. Maybe you only need 5%. Yeah, but 5% is a huge market. If 5% of Cloud Code users ship their model to GLM, it's a huge market. Yeah, and it's growing so fast. But not just for Cloud Code because we're trying new ideas like role-playing. Yeah, many people on Sillatorians are using GLN, Genitry AI, because we did very good in role-playing. Yeah, so we were trying to have more markets, coding markets, agentic markets. maybe one day like man is using our model all right we gotta we gotta take a step back and explain role-playing um what is it how do you make a model that's good at it um what what are people using it for so before gian 4.6 um for models like john 4.5 we are relevant relatively weak in role-playing because we have to train on those data. So we need to create some data and let the model follow the instructions. Because for role-playing, like very long system prompt, if you don't train on that kind of stuff, it will forget who it is and forget all the instruction and just use its general performance. to do the, do the conversation. But for role-playing task, if you've given them very long instructions, it will strictly follow these instructions and like show more emotion or show more, uh, behavior, like before following these instructions. Yeah. So just to be clear, this is, this is, you know, people having a conversation saying, you know i'm a japanese pirate i'm raiding the coast of taiwan and you know 1570 and um i want to plan an attack to defend the fort um and but like people write out like five pages of background um and then you know these are chatbots right so you're having conversations where you're you know it's like playing a like a like a text-based rpg from the 1980s except it's ai and it just generates um all right sorry uh to be clear i'm not sure everyone knows what role playing means um when it comes to when it comes to ai yeah also like we try something very interesting like family guy like we have our own stewie you just give a description of what stewie does in history and then you can create your own Stewie But we perform very well in text generation But if we have some speech model, we can recreate a Stewie. Was there specific kind of pre-training data or RL that you needed to do to get this? or did it just all of a sudden you're like oh wow like this is really good at pretending to be cartoon characters yeah I believe it's mainly post training data okay there's a big discussion of late in the US about people being worried folks are falling in love with AI there's this whole discussion about like AI psychosis where chat GPT kind of like convinces people who like trusted too much to harm themselves. I'm curious your broad sense of that type of discussion in China broadly and then internally in your firm about these sorts of questions where people are using AI to play or for emotional support. Yeah, I just read the post from OpenAI yesterday because they invited a lot of experts on trying to train a model that I think are not addictive. They train data to ask ChatGPT to say it's an AI instead of us saying it's a human being, not letting people attach to ChatGPT anymore. So I've read this, a bunch of people read this, so we can discuss. I think it's internally discussed when people find the relevant stuff. Yeah, we had the news. It's a hot topic and we can discuss. But from a broader audience perspective, I think not many people are looking into this. Yeah, because we are not there yet. If we have a model that can perform like GPT-5, then we can move on to remove the addiction. But still performance is not on par with these top close source models. We need to chase first. Because when we chase these models, we'll shift our focus on the data collection, data preparation and sometimes the model behavior will change dramatically so if we we do some similar things on our previous model it will be outdated in the next version so the performance is still very relevant currently i'm guessing this is somewhere in the rundown but like how it is the how is the balance of optimism versus fear of ai as like a long-term trajectory in your lab versus china generally because i think there's a very big concentration in the u.s of people that worry deeply about the long-term potential of ai whether it's like a powerful entity or concentration of power or other things like there's people that just think that this is the most important technology that has ever been invented. Like, we have to be really serious about it. Like, I'm just wondering where on this kind of spectrum you think the lab has a culture of, or if it's not really something that's debated and you're just like, we're building a useful thing and we're gonna keep making it better. I think developers fear the most. Yeah, because when you use code code, when you use code X, you get the fear in a very concrete way. You can do all the tasks. for you, especially for junior developers. Yeah, but like for writers, for for other managers, I think it's it's simple because we have SaaS, like we have we have like other other technologies helping them already. So large language models, I can teach us another helper for them. So I cannot fear, I feel fear coming out from the general public. But specifically for developers. Yeah, but for developers or data analysts, so they fear the most because they try out the new models, the new products more frequently than the general public. So they can feel the power. So if you don't use tools, because many people use DeepSeq and other chatbots, but DeepSeq can help you brainstorm the ideas, can help you polish your writing, can do the translation for you. But they don't believe that this work can replace them. But for developers, it's a different story. So what are the, what are the main fears? Just people's jobs getting taken away, AI taking over the world. What are, what the, for the people who are worried, what are they worried about? Maybe jobs jobs taken away. Like those are pretty different than the U S there's definitely a huge culture. Like there's not a majority in terms of the people, but a very vocal minority that influences a lot of the thinking of like the risks of AI. beyond well beyond just job loss. Like job loss is almost an assumption for many people in the, in the U S. And then there's like added fears on top of this. And I think that's like, it's a very different media ecosystem and thought ecosystem. Yeah. I definitely know about this because I did the research. But you, you lived here, you were, you lived through some of this obviously. Everyone at MIT are talking about like how we, I will change the world. not on the positive side, but on the negative side. Why do you think this is? Is it that Chinese society is a little more practical or just that job loss shows more imminent or is it because it's less of a market-driven economy? I believe that people just know about DeepSeek because maybe just 1 million people follow the latest trend and there are a billion people do their work daily and not impacted by AI. Yeah, so the more you learn, the more fear you have. And then what's the vibe among these younger engineers that you were talking about, like junior folks who are a little scared? I'm generally just curious what gets them into this work in the first place and what makes them want to work at places like ZAI. At ZAI, I think we lack people. So there's no fear about losing jobs here because we have a lot of things to do. But for other companies, especially the large enterprises, they have maybe 10,000 people doing the similar things like data analytics and also like backend engineering stuff. so they might think if other people are using code code or organic tool maybe they just need 50% of the people yes but they can do nothing they need to wait for their bosses or the founders to make the decision like what's happening at Amazon so for layoffs you can do nothing You just wait for the results. I wanted to jump in here and also ask about translation because their models are very strong in making very contextually rich translations from Chinese to English and deploying onto social media. Could you talk a bit more about the process behind that, if you know, and what's the secret sauce to translating memes? Yeah, exactly. we are doing very great in translation and especially the translation for Chinese and English I think we are on par with Gemini 2.5 Pro yeah but you mentioned memes memes is also one of our weapons because we just prepared the and we understand the culture we can even translate emoji yeah for for example what does that mean how does that work yeah because if you like like 10 center like like like 10 cent emojis to apple emojis no you can if you enter a sentence talking about ai and you use a whale to replace deep seek we might translate this back to deep seek and if you give the sentence about animals and we will translate into a whale yeah you understand the context is this because Chinese internet talk is just so cryptic yeah because people are very they are very novelty they're they're novelty they sometimes use emoji and and there's a lot of companies including some animal names in their brand, in their logo, and we're trying to use that to replace what actually people use. And also people use abbreviation, right? So all those things need to be translated right. I remember a few years ago, there was all this discussion like, oh, it's going to be really hard to train Chinese models to speak colloquially because all the data is behind And, you know, walled gardens and like, you know, you can't get the Tencent. Tencent has the Tencent data. Xiaohongshu has the Xiaohongshu data. Ali has the, I mean, I don't even know what data they have. But like, is that, was that a problem for you guys doing this more kind of like colloquial internet speak? Or is there enough out there and, you know, you can just scrape stuff and figure it out? We need sysetic data, right? We don't have the actual data. We cannot scrape something from other WeChat users per style fog. Yeah. But we know people are talking about, especially in the public area. So in the open area, we can observe what's going on on Xiong Fu, right on TikTok, on other things. We especially pay attention to their comment area because people are really naughty there in their comments. So actually when the TikTok refugees thing happen, we benefit from it because more people or more softwares need auto-translation. And we were trying to conquer some large customers through our translation capabilities. Does anyone train on like Danmu data? Like definitely. definitely yeah all the memes we're trying to collect memes from everywhere especially for our vision model because for the memes they're always in image format I'm trying to understand it with our vision model I think it's very interesting and it's also very necessary because if you cannot translate the comment in a very accurate way, they will not purchase your model. Unlike YouTube, because if you use YouTube's auto-translation, it won't grasp the exact meaning, because people will just need to understand, oh, this English version is about this and i can read it in chinese oh 80 is enough for me but for apps like x um reno xiong fu wechat you need to understand like 100 of the the common area is it a challenge to balance like i mean not let alone culture but data across like you like you're marketing to western users as well and you have your domestic market like is that a technical challenge to feel like you have to do both it's excellently i think it's a it's a challenge but we we can do very well in chinese and english and we're trying to explore more in french and even hindu so we have data on hindu So we can perform very well in, I believe in like 20 languages. But beyond that, we're still exploring the data, their software. So we need to register on their software to see what people are doing out there. Yeah, sometimes it's hard to figure out. are out and trying to like learn from Gemini and GPT five. Like why do why they do so great in translation? Uh, can we talk a little bit about compute? There's all these rumors. We're recording this, uh, October 29th, the evening us time. Are you excited to buy some black wells if, uh, they come on the market in the next few weeks? Blackwell is great because not only the chip but also FP4, right? So FP4 can reduce a lot of cost and yeah, we're trying to use the best we can get. Yeah, that's a strategy. I think it's pretty clear. Yeah. For the model training side, for the architecture, we use the best. full of chips. Not the best, but I think we do the best trade-off between the performance and the cost. Do you guys train outside of China as well or only on domestic clouds? Yeah, we do the inference from outside China. Yeah, but all the training is going out here. How do you feel about Huawei chips and software? Are they are they going to make it? Yeah, we are going to because we have multiple models like GLM 4.6 and the upcoming 4.6 Air and also our previous version. So we need to find the best use case for all sorts of chips, domestic chips and Ndda chips. Yeah, we need to classify the use case because for one customer, maybe it needs like 30 tokens per second and for another customers need 80 tokens per second. So maybe for one customer or one use case some chips are enough and for others we need better chips and better inference techniques. Do you try to do any API sales or just enterprise sales in general outside the US or China? because we mentioned having a lot of languages and whatnot. Do you see any use cases coming from other places? We have two platforms. Like inside China, our platform is called Big Model. Big Model, it's like a large language model. It's a simple translation, bigmodel.cn. And we also have a z.ai. It's called API.z.ai. It's our overseas platform. So I'm actually in charge of API.z.ai. So all of our services are hosted in Singapore. Yeah, so actually I'm an employee of a Singaporean company. Sorry, I wasn't clear. I meant, do you see much demand coming from, for the 4Z.ai coming from non-US countries, like other countries? A lot of countries, like India, Indonesia, even Norway and also Brazil yeah but it depends on who's using Reddit who's X because we basically did our growth on X on Reddit maybe some on YouTube like if people are watching watching these materials or videos like they will purchase it but we we're trying to do telegram or other things. So it might shift the proportion of our users. Yeah, but India, Indonesia are huge markets. So, but there are more revenue coming from US compared to other countries because they pay more. So they buy like pro plan, MX plan instead of the light plan. So yeah, in terms of users, I think India has the most number of users, but like U.S. market generate 50% of overseas revenue. Jordan, are you on the Chinese plans yet? What's your AI bill? How do we diversify this internationally? I'm on like $500 a month. It's not good. I don't know. just charge it to the firm charge it to the allen estate nathan come on we gotta gonna save you um irene what was your what was your question earlier um building off of what we were talking about earlier with like turns out walled gardens didn't matter which is that um with whether zeta ai has any thoughts about doing ai search on the chinese internet and what that would look like in China where there increasingly is no unified open internet? I think that's a challenge also for US product builders because Google don't have a API, search API, and Bing is trying to stop its search API. So there are other third-party providers like SERP, and they basically just scrape the data and they quickly send a request to Google and scrape the page, right? So also very challenging for builders like Proclaxity and even ChatGPT. So we need to do the technical side nowadays is using our own technology or trying to grasp multiple resources from different platforms. I think that's very reasonable. And there are other technologies like Manus, like they just browse the internet themselves without using API. I think that's more doable these days when you want to see multiple resources and trying to distinguish the best use case, the best resources, you need to really log into an account and see the data yourself, read the page yourself instead of just using whatever API gives you. Nathan, maybe you want to ask some broader research direction type stuff? or whatever else is on your mind where are you planning to take your models next i think less in domain but like how do you make models better given that everybody has limited compute and data resources and we're changing from chat to agents and it's just like how do you like how far out do you think or do you think about like the very short problems or like there just so many direction so you can take it I don know Yeah I can give some names of the idea we are exploring right now like on policy training on policy reinforced learning because we are quite mature in all policy reinforcement learning. But for the on policy learning, we still need to explore more. And also multi-agents. Yes, so when you look at the ZIH head, it actually acts like a single agent. So one model, do the search itself, comes back, and do another round of search, then come back, and it can generate slides, generate presentation, or generate a poster, things like that. it's all performed by a single actor the the one gom 4.6 but maybe for i think our models are like do we think you have to change our models a lot in order to do this because i think like so much of 2025 has been changing the training stack away from like we are a chatbot to now we are an agent and it's just like i like what do you think we should change the most about our models given that that is like it's almost like the air model the faster model might be more useful because you can have more of them and things like this. So that's the reason why we need to do a very solid evaluation because we have different product solutions. And currently, the single agent works very well on our platform, but we need to do more to try out different ideas and see whether we can improve the speed, the performance with multi-agent architecture and yeah, other possibilities. Because for single agents, it has better context management because you have the best model that can see all the contacts ahead of the current conversation and follow the instructions maybe up there. But for like multi-agents, you need to compress the context for each agent and that might lose some context. Or like orchestration is hard, where it's like if you give four agents the same context, they might all try the same thing and they might not work together well and stuff like this. Yeah, and maybe even one agent has hallucination, it will ruin all of the research. But also we are trying to make a longer context window and a longer effective context window because we all know that you said like your model can do 1 million context window but actually it just performed very well inside 60k or like 100k. You can release whatever size of context window you want and that it's like whether or not it actually works. Yeah. Like do you see like how much, I guess another question that people we debate a lot is like how much do you think it's going to be scaling the kind of transformers that we have which is like making a long context better like just improving the data versus if there's like fundamental walls that this is approaching kind of like the low-hanging fruit question like do you think there's a ton to keep improving it is it kind of easy to find the things to do and you just don't have time it's not easy it's not easy we believe it's the RT tech thing. Yeah, data can improve, but it cannot cross the wall. There is a wall. So we need better architecture, pre-training data, and also post-training data. Do you think you're starting to hit this wall, or do you kind of see it coming already? Is this something you're forecasting, or you're like, oh, this specific thing, data alone is not solving it for us? Because people in the US that are trading these models just don't talk about it. They're like, I don't know. I can't say it. So I'm like curious. And it's like the models I train are smaller. I think our biggest models like 30 B scale. So it's like when you scale up, you start to see very different limits of what's happening. But we need we need to do some experiments. Like GOM is a 355 billion parameter model, right? But we cannot do experiments with this large model, like we need to do experiments with some smaller models, maybe 9 billion parameter or 30 billion parameter. And we test our hypothesis. 90% of the time we we just failed. Because the experiments you cannot win every time. But you need to do a lot of scientific stuff to finally get the right answer. Yeah, so If you're talking about whether the GLM 4.6 architecture will hit the wall, actually there's a wall. But we need to shift our focus and start from maybe a new architecture or a new framework for doing this stuff. So it sounds like doing these bigger runs where I don't know, it's not necessarily barely making it, but definitely stressful for you. Yes, it's stressful, but we're going to use some engineering stuff to try to compress the text windows to make our users happy because you don't normally need that much. You don't normally need 1 million tokens. Yeah. So if it cannot perform very well, you can compress the context window to 60K or like 30K to make it work. Zisran, you mentioned earlier that you guys, all the inferences abroad but training is at home. What's behind the rationale of that decision there? I think the rationale is very simple because we provide services to oversee customers. So I think it's a requirement to store the data overseas. Right. So it's a very strict policy for our ZDAL AI endpoint. We change that privacy policy every month to make it stricter and more coherent to people's expectation. But for inference, for the training, I think it's more simple. Because we don't have many resources. We only have these resources and we need to utilize it. But like doing it on Nebius or AWS and, you know, Malaysia or Singapore, it's too expensive. It's too slow. You guys already have enough chips at home. What's the, what's the thinking there? I think it's not, it's not very slow. It's fast because we don't, we not only change the, the location of GPU, but also CPU. you. Yeah, and like database. So if they're all in Singapore is still very fast. But if you like, have to go back from Singapore to to mainland China and then go back to Singapore, it will be slow. But but on the training side, it's on the training side in particular, on the training side. Um, I think it's it's very simple, because we were not open there. We're not a topic, we don't have to like choose between Amazon and Google and their own infra. So they're doing very complicated stuff. But for us, I think we're still in the initial stage. Yeah, don't have many like complicated structure with these, these large inference providers. So things just very simple here. For now. For now, for now. Irene? or Nathan, any more training questions before Irene wraps us up? Only sensitive questions that no one, that I don't expect to have an answer to. How big's your next model? How many GPUs do you have? It's like, I don't know. It's not a real question. It's just the curiosity. So for our next generation, we are going to launch 4K6 Air and I don't know whether it's called maybe Mini. It's a 30 billion parameter model. So it becomes a lot smaller in a couple of weeks. And I think that's all for 2025. And for 2026, we are still doing experiments. Like what I said, I'm trying to explore more, but we are doing experiments on smaller models. So they will not be putting into practice in 2026, but it gives a lot of idea on how we're going to train the next generation. Yeah, we'll see. So when this podcast launched, I believe we already had like 4.6 Air, 4.6 Mini, and also Vision Model, the next like 4.6 Vision Model. Yeah, I guess a good question is like, how long does it take from when the model is done training till you've released it? Like, what is your thought process on getting it out fast? get it fast, get it fast, several hours. In several hours. Yeah. So we just open source it. I love it. When we finish the training, we do some like evaluation. And after the evaluation, just so we don't have some arrangements like sending the endpoint to LM Arena or to artificial analysis this and trying to let it evaluate first and then release the model. We don't have this and we don't have a nano banana thing that trying to make it famous before it's launched because we are very transparent and we believe that if you want to open source the model, the open source itself is the biggest event. So you're going to try to time it to some market or anything? because we're trying to do some market hit thing. From my side, I want to make it longer. We want like a week for me to collaborate with inference providers, like benchmark companies, coding agents, and let everyone try out the model before it's released. But from the company's perspective, if open source is the most important thing, you need to only prepare for the materials for open source. You need the benchmarks. You need maybe a tech blog. And it's very stressful for me because I need to negotiate with multiple partners within several hours. Like we have a new model it's coming in two hours maybe three hours, maybe you're sleeping. But this is huge. yeah sorry don't don't give you enough time for you to like connect to the model or into the integration but yeah we're trying to post your your tweet afterwards can you talk a little bit about ours i mean this was uh we now have america we got our own thing zero zero two uh what is zero zero yeah what nathan midnight to midnight with a two-hour break so dumb i think hours vary a lot even inside the company like someone would just leave the company at 7 p.m someone will never leave the company yeah for me i work 18 hours a day because I need to negotiate with like US large firm CEOs or the founder of coding agents I need to discuss with fireworks with Alamarina and with maybe Kilo Code their CEOs so I need to follow their time and do the meetings maybe at 2am 3am, it's all possible. Yeah, but like for our researchers or for the engineering team, I think your brains can only work maybe eight hours a day. So if you feel tired, you need to get some rest. Yeah, I think it's impossible to ask a top researcher to work 40 hours a day because yeah, that means that you are working really not efficiently or you just attend meetings, right? Because if you join meetings, you can join like 20 hours a day. You just sit here and listen to other people talking. But if you want to read papers and do experiments and write code, I think eight hours is enough. It's very sensible. My PhD advisor always said that you can totally change the world if you do four hours a day of top technical work just like go go walk in the sun after that you did a good job because i asked um a couple final questions then um i've always been curious i've always wanted to ask chinese ai folks this because i feel like the conversation on value propositions can be really western centric how do you explain the kind of value of your work to let's say like kids in high school in Beijing or your grandmother or my work yeah or the is the is work or the industry's work how do you explain the value of that to other people like to kids to older people in China it's hard it's hard I can only say I do the similar thing like deep seek so we we are just a company like deep seek we do the similar thing because deep seek is so famous everyone in high school and even in kindergarten knows about DeepSeek. For other companies even Quinn, yeah let's say Quinn cannot explain itself to kindergarten kids or high school students. Yeah because you can see I'm just an alternative of DeepSeek or other stuff but for like developers is simple. We are like one of the best coding LLMs you can find, especially in China. Yeah, but for high school students, they always ask like, we have DeepSeek, like what are we doing? So why we need you? So are you doing a similar thing? Or are you better? Are you faster? Like, if I'm not using DeepSeek, have Dobo, I have other apps. Why do I need your app? Yeah, that's very top. So, we still need to improve the model performance. I think that's the top priority. Yeah, product is the second or your experience, the product experience is the second. But without a solid model, like nobody will try to pay attention to you because like we are at the same level only the most famous one get all the attention so you think the salience of ai models generationally society came straight out deep seek and the kind of nationalism associated with that yeah i think there there is a hype yeah it got so famous even in china so we got unknown even here. I believe that a lot of students in Tsinghua University haven't tried GIOI or even haven't heard of this company. Because everybody renews, but not everyone goes to this building to visit GIOI. But deep seas are all over the news and social media. So it's really tough to explain over contribution, over value, because like we said, we have a genetic model or genetic tool use model. So what is tool? What is search? Yeah. But we're trying to do more in the future. Do you think Chinese society is starting to find AI to be more valuable or scary? Valuable. Yeah. Because we are not there yet. So AI is not so strong to make people fear about it. because there's still hallucination, still not following the instructions. So all those issues make people feel, oh, it's still very silly for me. Or there's an agent, but it has hallucination, how can I use it? Yeah, so still a lot of issues to solve before it gets more fearful or terrified for people. So we end every episode with a song. Does Drupal have like a theme song? Or what do people listen to when they code around the office? No, actually, because our founder loves Rangin. like he is a he's a pro in marathon yeah what's his time what's his marathon time marathon time like below 3 hours it's a solid because the founder of Moorshot really loves song but like our founder doesn't have much interest in so for our for anniversary we have a half marathon to celebrate the anniversary. It's crazy. I gotta go do this. I'm gonna go run the ZAI half marathon next year. So I have an intern who did 3 hours and 15 minutes. It's a girl to finish the half marathon. It is crazy. I just waited for her at the finished live. She's almost there. It's super crazy, but yeah, we need to work very long hours, so energy is very important. So no music, just work. I don't know if this makes you a good boss for waiting or a terrible boss for making her do it in first place. Those interns, man, they gotta earn their slot, show their dedication. I love that. She is actually the product manager of Z.A at chat. So she built this. Good. She earned it. Well, after making her do the half marathon, I'm glad you guys gave her a job at the end. She ate two hamburgers after that. Okay, good. Good. All right. Well, this was really fun. Thank you so much for joining the show. If you're finding value in the show, we'd appreciate it if you'd take a moment to share it with friends, post online, write a review on Apple Podcasts or Spotify, or just leave us a comment on YouTube. Of course, we always welcome your feedback, guests and topic suggestions, and sponsorship inquiries, either via our website, CognitiveRevolution.ai, or by DMing me on your favorite social network. The Cognitive Revolution is part of the Turpentine Network, a network of podcasts, which is now part of A16Z, where experts talk technology, business, economics, geopolitics, culture, and more. We're produced by AI Podcasting. If you're looking for podcast production help for everything from the moment you stop recording to the moment your audience starts listening, check them out and see my endorsement at AIpodcast.ing. And thank you to everyone who listens for being part of the Cognitive Revolution.