Podcasts about english japanese

From Palantir and Two Sigma to building Goodfire into the poster-child for actionable mechanistic interpretability, Mark Bissell (Member of Technical Staff) and Myra Deng (Head of Product) are trying to turn “peeking inside the model” into a repeatable production workflow by shipping APIs, landing real enterprise deployments, and now scaling the bet with a recent $150M Series B funding round at a $1.25B valuation.In this episode, we go far beyond the usual “SAEs are cool” take. We talk about Goodfire's core bet: that the AI lifecycle is still fundamentally broken because the only reliable control we have is data and we post-train, RLHF, and fine-tune by “slurping supervision through a straw,” hoping the model picks up the right behaviors while quietly absorbing the wrong ones. Goodfire's answer is to build a bi-directional interface between humans and models: read what's happening inside, edit it surgically, and eventually use interpretability during training so customization isn't just brute-force guesswork.Mark and Myra walk through what that looks like when you stop treating interpretability like a lab demo and start treating it like infrastructure: lightweight probes that add near-zero latency, token-level safety filters that can run at inference time, and interpretability workflows that survive messy constraints (multilingual inputs, synthetic→real transfer, regulated domains, no access to sensitive data). We also get a live window into what “frontier-scale interp” means operationally (i.e. steering a trillion-parameter model in real time by targeting internal features) plus why the same tooling generalizes cleanly from language models to genomics, medical imaging, and “pixel-space” world models.We discuss:* Myra + Mark's path: Palantir (health systems, forward-deployed engineering) → Goodfire early team; Two Sigma → Head of Product, translating frontier interpretability research into a platform and real-world deployments* What “interpretability” actually means in practice: not just post-hoc poking, but a broader “science of deep learning” approach across the full AI lifecycle (data curation → post-training → internal representations → model design)* Why post-training is the first big wedge: “surgical edits” for unintended behaviors likereward hacking, sycophancy, noise learned during customization plus the dream of targeted unlearning and bias removal without wrecking capabilities* SAEs vs probes in the real world: why SAE feature spaces sometimes underperform classifiers trained on raw activations for downstream detection tasks (hallucination, harmful intent, PII), and what that implies about “clean concept spaces”* Rakuten in production: deploying interpretability-based token-level PII detection at inference time to prevent routing private data to downstream providers plus the gnarly constraints: no training on real customer PII, synthetic→real transfer, English + Japanese, and tokenization quirks* Why interp can be operationally cheaper than LLM-judge guardrails: probes are lightweight, low-latency, and don't require hosting a second large model in the loop* Real-time steering at frontier scale: a demo of steering Kimi K2 (~1T params) live and finding features via SAE pipelines, auto-labeling via LLMs, and toggling a “Gen-Z slang” feature across multiple layers without breaking tool use* Hallucinations as an internal signal: the case that models have latent uncertainty / “user-pleasing” circuitry you can detect and potentially mitigate more directly than black-box methods* Steering vs prompting: the emerging view that activation steering and in-context learning are more closely connected than people think, including work mapping between the two (even for jailbreak-style behaviors)* Interpretability for science: using the same tooling across domains (genomics, medical imaging, materials) to debug spurious correlations and extract new knowledge up to and including early biomarker discovery work with major partners* World models + “pixel-space” interpretability: why vision/video models make concepts easier to see, how that accelerates the feedback loop, and why robotics/world-model partners are especially interesting design partners* The north star: moving from “data in, weights out” to intentional model design where experts can impart goals and constraints directly, not just via reward signals and brute-force post-training—Goodfire AI* Website: https://goodfire.ai* LinkedIn: https://www.linkedin.com/company/goodfire-ai/* X: https://x.com/GoodfireAIMyra Deng* Website: https://myradeng.com/* LinkedIn: https://www.linkedin.com/in/myra-deng/* X: https://x.com/myra_dengMark Bissell* LinkedIn: https://www.linkedin.com/in/mark-bissell/* X: https://x.com/MarkMBissellFull Video EpisodeTimestamps00:00:00 Introduction00:00:05 Introduction to the Latent Space Podcast and Guests from Goodfire00:00:29 What is Goodfire? Mission and Focus on Interpretability00:01:01 Goodfire's Practical Approach to Interpretability00:01:37 Goodfire's Series B Fundraise Announcement00:02:04 Backgrounds of Mark and Myra from Goodfire00:02:51 Team Structure and Roles at Goodfire00:05:13 What is Interpretability? Definitions and Techniques00:05:30 Understanding Errors00:07:29 Post-training vs. Pre-training Interpretability Applications00:08:51 Using Interpretability to Remove Unwanted Behaviors00:10:09 Grokking, Double Descent, and Generalization in Models00:10:15 404 Not Found Explained00:12:06 Subliminal Learning and Hidden Biases in Models00:14:07 How Goodfire Chooses Research Directions and Projects00:15:00 Troubleshooting Errors00:16:04 Limitations of SAEs and Probes in Interpretability00:18:14 Rakuten Case Study: Production Deployment of Interpretability00:20:45 Conclusion00:21:12 Efficiency Benefits of Interpretability Techniques00:21:26 Live Demo: Real-Time Steering in a Trillion Parameter Model00:25:15 How Steering Features are Identified and Labeled00:26:51 Detecting and Mitigating Hallucinations Using Interpretability00:31:20 Equivalence of Activation Steering and Prompting00:34:06 Comparing Steering with Fine-Tuning and LoRA Techniques00:36:04 Model Design and the Future of Intentional AI Development00:38:09 Getting Started in Mechinterp: Resources, Programs, and Open Problems00:40:51 Industry Applications and the Rise of Mechinterp in Practice00:41:39 Interpretability for Code Models and Real-World Usage00:43:07 Making Steering Useful for More Than Stylistic Edits00:46:17 Applying Interpretability to Healthcare and Scientific Discovery00:49:15 Why Interpretability is Crucial in High-Stakes Domains like Healthcare00:52:03 Call for Design Partners Across Domains00:54:18 Interest in World Models and Visual Interpretability00:57:22 Sci-Fi Inspiration: Ted Chiang and Interpretability01:00:14 Interpretability, Safety, and Alignment Perspectives01:04:27 Weak-to-Strong Generalization and Future Alignment Challenges01:05:38 Final Thoughts and Hiring/Collaboration Opportunities at GoodfireTranscriptShawn Wang [00:00:05]: So welcome to the Latent Space pod. We're back in the studio with our special MechInterp co-host, Vibhu. Welcome. Mochi, Mochi's special co-host. And Mochi, the mechanistic interpretability doggo. We have with us Mark and Myra from Goodfire. Welcome. Thanks for having us on. Maybe we can sort of introduce Goodfire and then introduce you guys. How do you introduce Goodfire today?Myra Deng [00:00:29]: Yeah, it's a great question. So Goodfire, we like to say, is an AI research lab that focuses on using interpretability to understand, learn from, and design AI models. And we really believe that interpretability will unlock the new generation, next frontier of safe and powerful AI models. That's our description right now, and I'm excited to dive more into the work we're doing to make that happen.Shawn Wang [00:00:55]: Yeah. And there's always like the official description. Is there an understatement? Is there an unofficial one that sort of resonates more with a different audience?Mark Bissell [00:01:01]: Well, being an AI research lab that's focused on interpretability, there's obviously a lot of people have a lot that they think about when they think of interpretability. And I think we have a pretty broad definition of what that means and the types of places that can be applied. And in particular, applying it in production scenarios, in high stakes industries, and really taking it sort of from the research world into the real world. Which, you know. It's a new field, so that hasn't been done all that much. And we're excited about actually seeing that sort of put into practice.Shawn Wang [00:01:37]: Yeah, I would say it wasn't too long ago that Anthopic was like still putting out like toy models or superposition and that kind of stuff. And I wouldn't have pegged it to be this far along. When you and I talked at NeurIPS, you were talking a little bit about your production use cases and your customers. And then not to bury the lead, today we're also announcing the fundraise, your Series B. $150 million. $150 million at a 1.25B valuation. Congrats, Unicorn.Mark Bissell [00:02:02]: Thank you. Yeah, no, things move fast.Shawn Wang [00:02:04]: We were talking to you in December and already some big updates since then. Let's dive, I guess, into a bit of your backgrounds as well. Mark, you were at Palantir working on health stuff, which is really interesting because the Goodfire has some interesting like health use cases. I don't know how related they are in practice.Mark Bissell [00:02:22]: Yeah, not super related, but I don't know. It was helpful context to know what it's like. Just to work. Just to work with health systems and generally in that domain. Yeah.Shawn Wang [00:02:32]: And Mara, you were at Two Sigma, which actually I was also at Two Sigma back in the day. Wow, nice.Myra Deng [00:02:37]: Did we overlap at all?Shawn Wang [00:02:38]: No, this is when I was briefly a software engineer before I became a sort of developer relations person. And now you're head of product. What are your sort of respective roles, just to introduce people to like what all gets done in Goodfire?Mark Bissell [00:02:51]: Yeah, prior to Goodfire, I was at Palantir for about three years as a forward deployed engineer, now a hot term. Wasn't always that way. And as a technical lead on the health care team and at Goodfire, I'm a member of the technical staff. And honestly, that I think is about as specific as like as as I could describe myself because I've worked on a range of things. And, you know, it's it's a fun time to be at a team that's still reasonably small. I think when I joined one of the first like ten employees, now we're above 40, but still, it looks like there's always a mix of research and engineering and product and all of the above. That needs to get done. And I think everyone across the team is, you know, pretty, pretty switch hitter in the roles they do. So I think you've seen some of the stuff that I worked on related to image models, which was sort of like a research demo. More recently, I've been working on our scientific discovery team with some of our life sciences partners, but then also building out our core platform for more of like flexing some of the kind of MLE and developer skills as well.Shawn Wang [00:03:53]: Very generalist. And you also had like a very like a founding engineer type role.Myra Deng [00:03:58]: Yeah, yeah.Shawn Wang [00:03:59]: So I also started as I still am a member of technical staff, did a wide range of things from the very beginning, including like finding our office space and all of this, which is we both we both visited when you had that open house thing. It was really nice.Myra Deng [00:04:13]: Thank you. Thank you. Yeah. Plug to come visit our office.Shawn Wang [00:04:15]: It looked like it was like 200 people. It has room for 200 people. But you guys are like 10.Myra Deng [00:04:22]: For a while, it was very empty. But yeah, like like Mark, I spend. A lot of my time as as head of product, I think product is a bit of a weird role these days, but a lot of it is thinking about how do we take our frontier research and really apply it to the most important real world problems and how does that then translate into a platform that's repeatable or a product and working across, you know, the engineering and research teams to make that happen and also communicating to the world? Like, what is interpretability? What is it used for? What is it good for? Why is it so important? All of these things are part of my day-to-day as well.Shawn Wang [00:05:01]: I love like what is things because that's a very crisp like starting point for people like coming to a field. They all do a fun thing. Vibhu, why don't you want to try tackling what is interpretability and then they can correct us.Vibhu Sapra [00:05:13]: Okay, great. So I think like one, just to kick off, it's a very interesting role to be head of product, right? Because you guys, at least as a lab, you're more of an applied interp lab, right? Which is pretty different than just normal interp, like a lot of background research. But yeah. You guys actually ship an API to try these things. You have Ember, you have products around it, which not many do. Okay. What is interp? So basically you're trying to have an understanding of what's going on in model, like in the model, in the internal. So different approaches to do that. You can do probing, SAEs, transcoders, all this stuff. But basically you have an, you have a hypothesis. You have something that you want to learn about what's happening in a model internals. And then you're trying to solve that from there. You can do stuff like you can, you know, you can do activation mapping. You can try to do steering. There's a lot of stuff that you can do, but the key question is, you know, from input to output, we want to have a better understanding of what's happening and, you know, how can we, how can we adjust what's happening on the model internals? How'd I do?Mark Bissell [00:06:12]: That was really good. I think that was great. I think it's also a, it's kind of a minefield of a, if you ask 50 people who quote unquote work in interp, like what is interpretability, you'll probably get 50 different answers. And. Yeah. To some extent also like where, where good fire sits in the space. I think that we're an AI research company above all else. And interpretability is a, is a set of methods that we think are really useful and worth kind of specializing in, in order to accomplish the goals we want to accomplish. But I think we also sort of see some of the goals as even more broader as, as almost like the science of deep learning and just taking a not black box approach to kind of any part of the like AI development life cycle, whether that. That means using interp for like data curation while you're training your model or for understanding what happened during post-training or for the, you know, understanding activations and sort of internal representations, what is in there semantically. And then a lot of sort of exciting updates that were, you know, are sort of also part of the, the fundraise around bringing interpretability to training, which I don't think has been done all that much before. A lot of this stuff is sort of post-talk poking at models as opposed to. To actually using this to intentionally design them.Shawn Wang [00:07:29]: Is this post-training or pre-training or is that not a useful.Myra Deng [00:07:33]: Currently focused on post-training, but there's no reason the techniques wouldn't also work in pre-training.Shawn Wang [00:07:38]: Yeah. It seems like it would be more active, applicable post-training because basically I'm thinking like rollouts or like, you know, having different variations of a model that you can tweak with the, with your steering. Yeah.Myra Deng [00:07:50]: And I think in a lot of the news that you've seen in, in, on like Twitter or whatever, you've seen a lot of unintended. Side effects come out of post-training processes, you know, overly sycophantic models or models that exhibit strange reward hacking behavior. I think these are like extreme examples. There's also, you know, very, uh, mundane, more mundane, like enterprise use cases where, you know, they try to customize or post-train a model to do something and it learns some noise or it doesn't appropriately learn the target task. And a big question that we've always had is like, how do you use your understanding of what the model knows and what it's doing to actually guide the learning process?Shawn Wang [00:08:26]: Yeah, I mean, uh, you know, just to anchor this for people, uh, one of the biggest controversies of last year was 4.0 GlazeGate. I've never heard of GlazeGate. I didn't know that was what it was called. The other one, they called it that on the blog post and I was like, well, how did OpenAI call it? Like officially use that term. And I'm like, that's funny, but like, yeah, I guess it's the pitch that if they had worked a good fire, they wouldn't have avoided it. Like, you know what I'm saying?Myra Deng [00:08:51]: I think so. Yeah. Yeah.Mark Bissell [00:08:53]: I think that's certainly one of the use cases. I think. Yeah. Yeah. I think the reason why post-training is a place where this makes a lot of sense is a lot of what we're talking about is surgical edits. You know, you want to be able to have expert feedback, very surgically change how your model is doing, whether that is, you know, removing a certain behavior that it has. So, you know, one of the things that we've been looking at or is, is another like common area where you would want to make a somewhat surgical edit is some of the models that have say political bias. Like you look at Quen or, um, R1 and they have sort of like this CCP bias.Shawn Wang [00:09:27]: Is there a CCP vector?Mark Bissell [00:09:29]: Well, there's, there are certainly internal, yeah. Parts of the representation space where you can sort of see where that lives. Yeah. Um, and you want to kind of, you know, extract that piece out.Shawn Wang [00:09:40]: Well, I always say, you know, whenever you find a vector, a fun exercise is just like, make it very negative to see what the opposite of CCP is.Mark Bissell [00:09:47]: The super America, bald eagles flying everywhere. But yeah. So in general, like lots of post-training tasks where you'd want to be able to, to do that. Whether it's unlearning a certain behavior or, you know, some of the other kind of cases where this comes up is, are you familiar with like the, the grokking behavior? I mean, I know the machine learning term of grokking.Shawn Wang [00:10:09]: Yeah.Mark Bissell [00:10:09]: Sort of this like double descent idea of, of having a model that is able to learn a generalizing, a generalizing solution, as opposed to even if memorization of some task would suffice, you want it to learn the more general way of doing a thing. And so, you know, another. A way that you can think about having surgical access to a model's internals would be learn from this data, but learn in the right way. If there are many possible, you know, ways to, to do that. Can make interp solve the double descent problem?Shawn Wang [00:10:41]: Depends, I guess, on how you. Okay. So I, I, I viewed that double descent as a problem because then you're like, well, if the loss curves level out, then you're done, but maybe you're not done. Right. Right. But like, if you actually can interpret what is a generalizing or what you're doing. What is, what is still changing, even though the loss is not changing, then maybe you, you can actually not view it as a double descent problem. And actually you're just sort of translating the space in which you view loss and like, and then you have a smooth curve. Yeah.Mark Bissell [00:11:11]: I think that's certainly like the domain of, of problems that we're, that we're looking to get.Shawn Wang [00:11:15]: Yeah. To me, like double descent is like the biggest thing to like ML research where like, if you believe in scaling, then you don't need, you need to know where to scale. And. But if you believe in double descent, then you don't, you don't believe in anything where like anything levels off, like.Vibhu Sapra [00:11:30]: I mean, also tendentially there's like, okay, when you talk about the China vector, right. There's the subliminal learning work. It was from the anthropic fellows program where basically you can have hidden biases in a model. And as you distill down or, you know, as you train on distilled data, those biases always show up, even if like you explicitly try to not train on them. So, you know, it's just like another use case of. Okay. If we can interpret what's happening in post-training, you know, can we clear some of this? Can we even determine what's there? Because yeah, it's just like some worrying research that's out there that shows, you know, we really don't know what's going on.Mark Bissell [00:12:06]: That is. Yeah. I think that's the biggest sentiment that we're sort of hoping to tackle. Nobody knows what's going on. Right. Like subliminal learning is just an insane concept when you think about it. Right. Train a model on not even the logits, literally the output text of a bunch of random numbers. And now your model loves owls. And you see behaviors like that, that are just, they defy, they defy intuition. And, and there are mathematical explanations that you can get into, but. I mean.Shawn Wang [00:12:34]: It feels so early days. Objectively, there are a sequence of numbers that are more owl-like than others. There, there should be.Mark Bissell [00:12:40]: According to, according to certain models. Right. It's interesting. I think it only applies to models that were initialized from the same starting Z. Usually, yes.Shawn Wang [00:12:49]: But I mean, I think that's a, that's a cheat code because there's not enough compute. But like if you believe in like platonic representation, like probably it will transfer across different models as well. Oh, you think so?Mark Bissell [00:13:00]: I think of it more as a statistical artifact of models initialized from the same seed sort of. There's something that is like path dependent from that seed that might cause certain overlaps in the latent space and then sort of doing this distillation. Yeah. Like it pushes it towards having certain other tendencies.Vibhu Sapra [00:13:24]: Got it. I think there's like a bunch of these open-ended questions, right? Like you can't train in new stuff during the RL phase, right? RL only reorganizes weights and you can only do stuff that's somewhat there in your base model. You're not learning new stuff. You're just reordering chains and stuff. But okay. My broader question is when you guys work at an interp lab, how do you decide what to work on and what's kind of the thought process? Right. Because we can ramble for hours. Okay. I want to know this. I want to know that. But like, how do you concretely like, you know, what's the workflow? Okay. There's like approaches towards solving a problem, right? I can try prompting. I can look at chain of thought. I can train probes, SAEs. But how do you determine, you know, like, okay, is this going anywhere? Like, do we have set stuff? Just, you know, if you can help me with all that. Yeah.Myra Deng [00:14:07]: It's a really good question. I feel like we've always at the very beginning of the company thought about like, let's go and try to learn what isn't working in machine learning today. Whether that's talking to customers or talking to researchers at other labs, trying to understand both where the frontier is going and where things are really not falling apart today. And then developing a perspective on how we can push the frontier using interpretability methods. And so, you know, even our chief scientist, Tom, spends a lot of time talking to customers and trying to understand what real world problems are and then taking that back and trying to apply the current state of the art to those problems and then seeing where they fall down basically. And then using those failures or those shortcomings to understand what hills to climb when it comes to interpretability research. So like on the fundamental side, for instance, when we have done some work applying SAEs and probes, we've encountered, you know, some shortcomings in SAEs that we found a little bit surprising. And so have gone back to the drawing board and done work on that. And then, you know, we've done some work on better foundational interpreter models. And a lot of our team's research is focused on what is the next evolution beyond SAEs, for instance. And then when it comes to like control and design of models, you know, we tried steering with our first API and realized that it still fell short of black box techniques like prompting or fine tuning. And so went back to the drawing board and we're like, how do we make that not the case and how do we improve it beyond that? And one of our researchers, Ekdeep, who just joined is actually Ekdeep and Atticus are like steering experts and have spent a lot of time trying to figure out like, what is the research that enables us to actually do this in a much more powerful, robust way? So yeah, the answer is like, look at real world problems, try to translate that into a research agenda and then like hill climb on both of those at the same time.Shawn Wang [00:16:04]: Yeah. Mark has the steering CLI demo queued up, which we're going to go into in a sec. But I always want to double click on when you drop hints, like we found some problems with SAEs. Okay. What are they? You know, and then we can go into the demo. Yeah.Myra Deng [00:16:19]: I mean, I'm curious if you have more thoughts here as well, because you've done it in the healthcare domain. But I think like, for instance, when we do things like trying to detect behaviors within models that are harmful or like behaviors that a user might not want to have in their model. So hallucinations, for instance, harmful intent, PII, all of these things. We first tried using SAE probes for a lot of these tasks. So taking the feature activation space from SAEs and then training classifiers on top of that, and then seeing how well we can detect the properties that we might want to detect in model behavior. And we've seen in many cases that probes just trained on raw activations seem to perform better than SAE probes, which is a bit surprising if you think that SAEs are actually also capturing the concepts that you would want to capture cleanly and more surgically. And so that is an interesting observation. I don't think that is like, I'm not down on SAEs at all. I think there are many, many things they're useful for, but we have definitely run into cases where I think the concept space described by SAEs is not as clean and accurate as we would expect it to be for actual like real world downstream performance metrics.Mark Bissell [00:17:34]: Fair enough. Yeah. It's the blessing and the curse of unsupervised methods where you get to peek into the AI's mind. But sometimes you wish that you saw other things when you walked inside there. Although in the PII instance, I think weren't an SAE based approach actually did prove to be the most generalizable?Myra Deng [00:17:53]: It did work well in the case that we published with Rakuten. And I think a lot of the reasons it worked well was because we had a noisier data set. And so actually the blessing of unsupervised learning is that we actually got to get more meaningful, generalizable signal from SAEs when the data was noisy. But in other cases where we've had like good data sets, it hasn't been the case.Shawn Wang [00:18:14]: And just because you named Rakuten and I don't know if we'll get it another chance, like what is the overall, like what is Rakuten's usage or production usage? Yeah.Myra Deng [00:18:25]: So they are using us to essentially guardrail and inference time monitor their language model usage and their agent usage to detect things like PII so that they don't route private user information.Myra Deng [00:18:41]: And so that's, you know, going through all of their user queries every day. And that's something that we deployed with them a few months ago. And now we are actually exploring very early partnerships, not just with Rakuten, but with other people around how we can help with potentially training and customization use cases as well. Yeah.Shawn Wang [00:19:03]: And for those who don't know, like it's Rakuten is like, I think number one or number two e-commerce store in Japan. Yes. Yeah.Mark Bissell [00:19:10]: And I think that use case actually highlights a lot of like what it looks like to deploy things in practice that you don't always think about when you're doing sort of research tasks. So when you think about some of the stuff that came up there that's more complex than your idealized version of a problem, they were encountering things like synthetic to real transfer of methods. So they couldn't train probes, classifiers, things like that on actual customer data of PII. So what they had to do is use synthetic data sets. And then hope that that transfer is out of domain to real data sets. And so we can evaluate performance on the real data sets, but not train on customer PII. So that right off the bat is like a big challenge. You have multilingual requirements. So this needed to work for both English and Japanese text. Japanese text has all sorts of quirks, including tokenization behaviors that caused lots of bugs that caused us to be pulling our hair out. And then also a lot of tasks you'll see. You might make simplifying assumptions if you're sort of treating it as like the easiest version of the problem to just sort of get like general results where maybe you say you're classifying a sentence to say, does this contain PII? But the need that Rakuten had was token level classification so that you could precisely scrub out the PII. So as we learned more about the problem, you're sort of speaking about what that looks like in practice. Yeah. A lot of assumptions end up breaking. And that was just one instance where you. A problem that seems simple right off the bat ends up being more complex as you keep diving into it.Vibhu Sapra [00:20:41]: Excellent. One of the things that's also interesting with Interp is a lot of these methods are very efficient, right? So where you're just looking at a model's internals itself compared to a separate like guardrail, LLM as a judge, a separate model. One, you have to host it. Two, there's like a whole latency. So if you use like a big model, you have a second call. Some of the work around like self detection of hallucination, it's also deployed for efficiency, right? So if you have someone like Rakuten doing it in production live, you know, that's just another thing people should consider.Mark Bissell [00:21:12]: Yeah. And something like a probe is super lightweight. Yeah. It's no extra latency really. Excellent.Shawn Wang [00:21:17]: You have the steering demos lined up. So we were just kind of see what you got. I don't, I don't actually know if this is like the latest, latest or like alpha thing.Mark Bissell [00:21:26]: No, this is a pretty hacky demo from from a presentation that someone else on the team recently gave. So this will give a sense for, for technology. So you can see the steering and action. Honestly, I think the biggest thing that this highlights is that as we've been growing as a company and taking on kind of more and more ambitious versions of interpretability related problems, a lot of that comes to scaling up in various different forms. And so here you're going to see steering on a 1 trillion parameter model. This is Kimi K2. And so it's sort of fun that in addition to the research challenges, there are engineering challenges that we're now tackling. Cause for any of this to be sort of useful in production, you need to be thinking about what it looks like when you're using these methods on frontier models as opposed to sort of like toy kind of model organisms. So yeah, this was thrown together hastily, pretty fragile behind the scenes, but I think it's quite a fun demo. So screen sharing is on. So I've got two terminal sessions pulled up here. On the left is a forked version that we have of the Kimi CLI that we've got running to point at our custom hosted Kimi model. And then on the right is a set up that will allow us to steer on certain concepts. So I should be able to chat with Kimi over here. Tell it hello. This is running locally. So the CLI is running locally, but the Kimi server is running back to the office. Well, hopefully should be, um, that's too much to run on that Mac. Yeah. I think it's, uh, it takes a full, like each 100 node. I think it's like, you can. You can run it on eight GPUs, eight 100. So, so yeah, Kimi's running. We can ask it a prompt. It's got a forked version of our, uh, of the SG line code base that we've been working on. So I'm going to tell it, Hey, this SG line code base is slow. I think there's a bug. Can you try to figure it out? There's a big code base, so it'll, it'll spend some time doing this. And then on the right here, I'm going to initialize in real time. Some steering. Let's see here.Mark Bissell [00:23:33]: searching for any. Bugs. Feature ID 43205.Shawn Wang [00:23:38]: Yeah.Mark Bissell [00:23:38]: 20, 30, 40. So let me, uh, this is basically a feature that we found that inside Kimi seems to cause it to speak in Gen Z slang. And so on the left, it's still sort of thinking normally it might take, I don't know, 15 seconds for this to kick in, but then we're going to start hopefully seeing him do this code base is massive for real. So we're going to start. We're going to start seeing Kimi transition as the steering kicks in from normal Kimi to Gen Z Kimi and both in its chain of thought and its actual outputs.Mark Bissell [00:24:19]: And interestingly, you can see, you know, it's still able to call tools, uh, and stuff. It's um, it's purely sort of it's it's demeanor. And there are other features that we found for interesting things like concision. So that's more of a practical one. You can make it more concise. Um, the types of programs, uh, programming languages that uses, but yeah, as we're seeing it come in. Pretty good. Outputs.Shawn Wang [00:24:43]: Scheduler code is actually wild.Vibhu Sapra [00:24:46]: Yo, this code is actually insane, bro.Vibhu Sapra [00:24:53]: What's the process of training in SAE on this, or, you know, how do you label features? I know you guys put out a pretty cool blog post about, um, finding this like autonomous interp. Um, something. Something about how agents for interp is different than like coding agents. I don't know while this is spewing up, but how, how do we find feature 43, two Oh five. Yeah.Mark Bissell [00:25:15]: So in this case, um, we, our platform that we've been building out for a long time now supports all the sort of classic out of the box interp techniques that you might want to have like SAE training, probing things of that kind, I'd say the techniques for like vanilla SAEs are pretty well established now where. You take your model that you're interpreting, run a whole bunch of data through it, gather activations, and then yeah, pretty straightforward pipeline to train an SAE. There are a lot of different varieties. There's top KSAEs, batch top KSAEs, um, normal ReLU SAEs. And then once you have your sparse features to your point, assigning labels to them to actually understand that this is a gen Z feature, that's actually where a lot of the kind of magic happens. Yeah. And the most basic standard technique is look at all of your d input data set examples that cause this feature to fire most highly. And then you can usually pick out a pattern. So for this feature, If I've run a diverse enough data set through my model feature 43, two Oh five. Probably tends to fire on all the tokens that sounds like gen Z slang. You know, that's the, that's the time of year to be like, Oh, I'm in this, I'm in this Um, and, um, so, you know, you could have a human go through all 43,000 concepts andVibhu Sapra [00:26:34]: And I've got to ask the basic question, you know, can we get examples where it hallucinates, pass it through, see what feature activates for hallucinations? Can I just, you know, turn hallucination down?Myra Deng [00:26:51]: Oh, wow. You really predicted a project we're already working on right now, which is detecting hallucinations using interpretability techniques. And this is interesting because hallucinations is something that's very hard to detect. And it's like a kind of a hairy problem and something that black box methods really struggle with. Whereas like Gen Z, you could always train a simple classifier to detect that hallucinations is harder. But we've seen that models internally have some... Awareness of like uncertainty or some sort of like user pleasing behavior that leads to hallucinatory behavior. And so, yeah, we have a project that's trying to detect that accurately. And then also working on mitigating the hallucinatory behavior in the model itself as well.Shawn Wang [00:27:39]: Yeah, I would say most people are still at the level of like, oh, I would just turn temperature to zero and that turns off hallucination. And I'm like, well, that's a fundamental misunderstanding of how this works. Yeah.Mark Bissell [00:27:51]: Although, so part of what I like about that question is you, there are SAE based approaches that might like help you get at that. But oftentimes the beauty of SAEs and like we said, the curse is that they're unsupervised. So when you have a behavior that you deliberately would like to remove, and that's more of like a supervised task, often it is better to use something like probes and specifically target the thing that you're interested in reducing as opposed to sort of like hoping that when you fragment the latent space, one of the vectors that pops out.Vibhu Sapra [00:28:20]: And as much as we're training an autoencoder to be sparse, we're not like for sure certain that, you know, we will get something that just correlates to hallucination. You'll probably split that up into 20 other things and who knows what they'll be.Mark Bissell [00:28:36]: Of course. Right. Yeah. So there's no sort of problems with like feature splitting and feature absorption. And then there's the off target effects, right? Ideally, you would want to be very precise where if you reduce the hallucination feature, suddenly maybe your model can't write. Creatively anymore. And maybe you don't like that, but you want to still stop it from hallucinating facts and figures.Shawn Wang [00:28:55]: Good. So Vibhu has a paper to recommend there that we'll put in the show notes. But yeah, I mean, I guess just because your demo is done, any any other things that you want to highlight or any other interesting features you want to show?Mark Bissell [00:29:07]: I don't think so. Yeah. Like I said, this is a pretty small snippet. I think the main sort of point here that I think is exciting is that there's not a whole lot of inter being applied to models quite at this scale. You know, Anthropic certainly has some some. Research and yeah, other other teams as well. But it's it's nice to see these techniques, you know, being put into practice. I think not that long ago, the idea of real time steering of a trillion parameter model would have sounded.Shawn Wang [00:29:33]: Yeah. The fact that it's real time, like you started the thing and then you edited the steering vector.Vibhu Sapra [00:29:38]: I think it's it's an interesting one TBD of what the actual like production use case would be on that, like the real time editing. It's like that's the fun part of the demo, right? You can kind of see how this could be served behind an API, right? Like, yes, you're you only have so many knobs and you can just tweak it a bit more. And I don't know how it plays in. Like people haven't done that much with like, how does this work with or without prompting? Right. How does this work with fine tuning? Like, there's a whole hype of continual learning, right? So there's just so much to see. Like, is this another parameter? Like, is it like parameter? We just kind of leave it as a default. We don't use it. So I don't know. Maybe someone here wants to put out a guide on like how to use this with prompting when to do what?Mark Bissell [00:30:18]: Oh, well, I have a paper recommendation. I think you would love from Act Deep on our team, who is an amazing researcher, just can't say enough amazing things about Act Deep. But he actually has a paper that as well as some others from the team and elsewhere that go into the essentially equivalence of activation steering and in context learning and how those are from a he thinks of everything in a cognitive neuroscience Bayesian framework, but basically how you can precisely show how. Prompting in context, learning and steering exhibit similar behaviors and even like get quantitative about the like magnitude of steering you would need to do to induce a certain amount of behavior similar to certain prompting, even for things like jailbreaks and stuff. It's a really cool paper. Are you saying steering is less powerful than prompting? More like you can almost write a formula that tells you how to convert between the two of them.Myra Deng [00:31:20]: And so like formally equivalent actually in the in the limit. Right.Mark Bissell [00:31:24]: So like one case study of this is for jailbreaks there. I don't know. Have you seen the stuff where you can do like many shot jailbreaking? You like flood the context with examples of the behavior. And the topic put out that paper.Shawn Wang [00:31:38]: A lot of people were like, yeah, we've been doing this, guys.Mark Bissell [00:31:40]: Like, yeah, what's in this in context learning and activation steering equivalence paper is you can like predict the number. Number of examples that you will need to put in there in order to jailbreak the model. That's cool. By doing steering experiments and using this sort of like equivalence mapping. That's cool. That's really cool. It's very neat. Yeah.Shawn Wang [00:32:02]: I was going to say, like, you know, I can like back rationalize that this makes sense because, you know, what context is, is basically just, you know, it updates the KV cache kind of and like and then every next token inference is still like, you know, the sheer sum of everything all the way. It's plus all the context. It's up to date. And you could, I guess, theoretically steer that with you probably replace that with your steering. The only problem is steering typically is on one layer, maybe three layers like like you did. So it's like not exactly equivalent.Mark Bissell [00:32:33]: Right, right. There's sort of you need to get precise about, yeah, like how you sort of define steering and like what how you're modeling the setup. But yeah, I've got the paper pulled up here. Belief dynamics reveal the dual nature. Yeah. The title is Belief Dynamics Reveal the Dual Nature of Incompetence. And it's an exhibition of the practical context learning and activation steering. So Eric Bigelow, Dan Urgraft on the who are doing fellowships at Goodfire, Ekt Deep's the final author there.Myra Deng [00:32:59]: I think actually to your question of like, what is the production use case of steering? I think maybe if you just think like one level beyond steering as it is today. Like imagine if you could adapt your model to be, you know, an expert legal reasoner. Like in almost real time, like very quickly. efficiently using human feedback or using like your semantic understanding of what the model knows and where it knows that behavior. I think that while it's not clear what the product is at the end of the day, it's clearly very valuable. Thinking about like what's the next interface for model customization and adaptation is a really interesting problem for us. Like we have heard a lot of people actually interested in fine-tuning an RL for open weight models in production. And so people are using things like Tinker or kind of like open source libraries to do that, but it's still very difficult to get models fine-tuned and RL'd for exactly what you want them to do unless you're an expert at model training. And so that's like something we'reShawn Wang [00:34:06]: looking into. Yeah. I never thought so. Tinker from Thinking Machines famously uses rank one LoRa. Is that basically the same as steering? Like, you know, what's the comparison there?Mark Bissell [00:34:19]: Well, so in that case, you are still applying updates to the parameters, right?Shawn Wang [00:34:25]: Yeah. You're not touching a base model. You're touching an adapter. It's kind of, yeah.Mark Bissell [00:34:30]: Right. But I guess it still is like more in parameter space then. I guess it's maybe like, are you modifying the pipes or are you modifying the water flowing through the pipes to get what you're after? Yeah. Just maybe one way.Mark Bissell [00:34:44]: I like that analogy. That's my mental map of it at least, but it gets at this idea of model design and intentional design, which is something that we're, that we're very focused on. And just the fact that like, I hope that we look back at how we're currently training models and post-training models and just think what a primitive way of doing that right now. Like there's no intentionalityShawn Wang [00:35:06]: really in... It's just data, right? The only thing in control is what data we feed in.Mark Bissell [00:35:11]: So, so Dan from Goodfire likes to use this analogy of, you know, he has a couple of young kids and he talks about like, what if I could only teach my kids how to be good people by giving them cookies or like, you know, giving them a slap on the wrist if they do something wrong, like not telling them why it was wrong or like what they should have done differently or something like that. Just figure it out. Right. Exactly. So that's RL. Yeah. Right. And, and, you know, it's sample inefficient. There's, you know, what do they say? It's like slurping feedback. It's like, slurping supervision. Right. And so you'd like to get to the point where you can have experts giving feedback to their models that are, uh, internalized and, and, you know, steering is an inference time way of sort of getting that idea. But ideally you're moving to a world whereVibhu Sapra [00:36:04]: it is much more intentional design in perpetuity for these models. Okay. This is one of the questions we asked Emmanuel from Anthropic on the podcast a few months ago. Basically the question, was you're at a research lab that does model training, foundation models, and you're on an interp team. How does it tie back? Right? Like, does this, do ideas come from the pre-training team? Do they go back? Um, you know, so for those interested, you can, you can watch that. There wasn't too much of a connect there, but it's still something, you know, it's something they want toMark Bissell [00:36:33]: push for down the line. It can be useful for all of the above. Like there are certainly post-hocVibhu Sapra [00:36:39]: use cases where it doesn't need to touch that. I think the other thing a lot of people forget is this stuff isn't too computationally expensive, right? Like I would say, if you're interested in getting into research, MechInterp is one of the most approachable fields, right? A lot of this train an essay, train a probe, this stuff, like the budget for this one, there's already a lot done. There's a lot of open source work. You guys have done some too. Um, you know,Shawn Wang [00:37:04]: There's like notebooks from the Gemini team for Neil Nanda or like, this is how you do it. Just step through the notebook.Vibhu Sapra [00:37:09]: Even if you're like, not even technical with any of this, you can still make like progress. There, you can look at different activations, but, uh, if you do want to get into training, you know, training this stuff, correct me if I'm wrong is like in the thousands of dollars, not even like, it's not that high scale. And then same with like, you know, applying it, doing it for post-training or all this stuff is fairly cheap in scale of, okay. I want to get into like model training. I don't have compute for like, you know, pre-training stuff. So it's, it's a very nice field to get into. And also there's a lot of like open questions, right? Um, some of them have to go with, okay, I want a product. I want to solve this. Like there's also just a lot of open-ended stuff that people could work on. That's interesting. Right. I don't know if you guys have any calls for like, what's open questions, what's open work that you either open collaboration with, or like, you'd just like to see solved or just, you know, for people listening that want to get into McInturk because people always talk about it. What are, what are the things they should check out? Start, of course, you know, join you guys as well. I'm sure you're hiring.Myra Deng [00:38:09]: There's a paper, I think from, was it Lee, uh, Sharky? It's open problems and, uh, it's, it's a bit of interpretability, which I recommend everyone who's interested in the field. Read. I'm just like a really comprehensive overview of what are the things that experts in the field think are the most important problems to be solved. I also think to your point, it's been really, really inspiring to see, I think a lot of young people getting interested in interpretability, actually not just young people also like scientists to have been, you know, experts in physics for many years and in biology or things like this, um, transitioning into interp, because the barrier of, of what's now interp. So it's really cool to see a number to entry is, you know, in some ways low and there's a lot of information out there and ways to get started. There's this anecdote of like professors at universities saying that all of a sudden every incoming PhD student wants to study interpretability, which was not the case a few years ago. So it just goes to show how, I guess, like exciting the field is, how fast it's moving, how quick it is to get started and things like that.Mark Bissell [00:39:10]: And also just a very welcoming community. You know, there's an open source McInturk Slack channel. There are people are always posting questions and just folks in the space are always responsive if you ask things on various forums and stuff. But yeah, the open paper, open problems paper is a really good one.Myra Deng [00:39:28]: For other people who want to get started, I think, you know, MATS is a great program. What's the acronym for? Machine Learning and Alignment Theory Scholars? It's like the...Vibhu Sapra [00:39:40]: Normally summer internship style.Myra Deng [00:39:42]: Yeah, but they've been doing it year round now. And actually a lot of our full-time staff have come through that program or gone through that program. And it's great for anyone who is transitioning into interpretability. There's a couple other fellows programs. We do one as well as Anthropic. And so those are great places to get started if anyone is interested.Mark Bissell [00:40:03]: Also, I think been seen as a research field for a very long time. But I think engineering... I think engineers are sorely wanted for interpretability as well, especially at Goodfire, but elsewhere, as it does scale up.Shawn Wang [00:40:18]: I should mention that Lee actually works with you guys, right? And in the London office and I'm adding our first ever McInturk track at AI Europe because I see this industry applications now emerging. And I'm pretty excited to, you know, help push that along. Yeah, I was looking forward to that. It'll effectively be the first industry McInturk conference. Yeah. I'm so glad you added that. You know, it's still a little bit of a bet. It's not that widespread, but I can definitely see this is the time to really get into it. We want to be early on things.Mark Bissell [00:40:51]: For sure. And I think the field understands this, right? So at ICML, I think the title of the McInturk workshop this year was actionable interpretability. And there was a lot of discussion around bringing it to various domains. Everyone's adding pragmatic, actionable, whatever.Shawn Wang [00:41:10]: It's like, okay, well, we weren't actionable before, I guess. I don't know.Vibhu Sapra [00:41:13]: And I mean, like, just, you know, being in Europe, you see the Interp room. One, like old school conferences, like, I think they had a very tiny room till they got lucky and they got it doubled. But there's definitely a lot of interest, a lot of niche research. So you see a lot of research coming out of universities, students. We covered the paper last week. It's like two unknown authors, not many citations. But, you know, you can make a lot of meaningful work there. Yeah. Yeah. Yeah.Shawn Wang [00:41:39]: Yeah. I think people haven't really mentioned this yet. It's just Interp for code. I think it's like an abnormally important field. We haven't mentioned this yet. The conspiracy theory last two years ago was when the first SAE work came out of Anthropic was they would do like, oh, we just used SAEs to turn the bad code vector down and then turn up the good code. And I think like, isn't that the dream? Like, you know, like, but basically, I guess maybe, why is it funny? Like, it's... If it was realistic, it would not be funny. It would be like, no, actually, we should do this. But it's funny because we know there's like, we feel there's some limitations to what steering can do. And I think a lot of the public image of steering is like the Gen Z stuff. Like, oh, you can make it really love the Golden Gate Bridge, or you can make it speak like Gen Z. To like be a legal reasoner seems like a huge stretch. Yeah. And I don't know if that will get there this way. Yeah.Myra Deng [00:42:36]: I think, um, I will say we are announcing. Something very soon that I will not speak too much about. Um, but I think, yeah, this is like what we've run into again and again is like, we, we don't want to be in the world where steering is only useful for like stylistic things. That's definitely not, not what we're aiming for. But I think the types of interventions that you need to do to get to things like legal reasoning, um, are much more sophisticated and require breakthroughs in, in learning algorithms. And that's, um...Shawn Wang [00:43:07]: And is this an emergent property of scale as well?Myra Deng [00:43:10]: I think so. Yeah. I mean, I think scale definitely helps. I think scale allows you to learn a lot of information and, and reduce noise across, you know, large amounts of data. But I also think we think that there's ways to do things much more effectively, um, even, even at scale. So like actually learning exactly what you want from the data and not learning things that you do that you don't want exhibited in the data. So we're not like anti-scale, but we are also realizing that scale is not going to get us anywhere. It's not going to get us to the type of AI development that we want to be at in, in the future as these models get more powerful and get deployed in all these sorts of like mission critical contexts. Current life cycle of training and deploying and evaluations is, is to us like deeply broken and has opportunities to, to improve. So, um, more to come on that very, very soon.Mark Bissell [00:44:02]: And I think that that's a use basically, or maybe just like a proof point that these concepts do exist. Like if you can manipulate them in the precise best way, you can get the ideal combination of them that you desire. And steering is maybe the most coarse grained sort of peek at what that looks like. But I think it's evocative of what you could do if you had total surgical control over every concept, every parameter. Yeah, exactly.Myra Deng [00:44:30]: There were like bad code features. I've got it pulled up.Vibhu Sapra [00:44:33]: Yeah. Just coincidentally, as you guys are talking.Shawn Wang [00:44:35]: This is like, this is exactly.Vibhu Sapra [00:44:38]: There's like specifically a code error feature that activates and they show, you know, it's not, it's not typo detection. It's like, it's, it's typos in code. It's not typical typos. And, you know, you can, you can see it clearly activates where there's something wrong in code. And they have like malicious code, code error. They have a whole bunch of sub, you know, sub broken down little grain features. Yeah.Shawn Wang [00:45:02]: Yeah. So, so the, the rough intuition for me, the, why I talked about post-training was that, well, you just, you know, have a few different rollouts with all these things turned off and on and whatever. And then, you know, you can, that's, that's synthetic data you can kind of post-train on. Yeah.Vibhu Sapra [00:45:13]: And I think we make it sound easier than it is just saying, you know, they do the real hard work.Myra Deng [00:45:19]: I mean, you guys, you guys have the right idea. Exactly. Yeah. We replicated a lot of these features in, in our Lama models as well. I remember there was like.Vibhu Sapra [00:45:26]: And I think a lot of this stuff is open, right? Like, yeah, you guys opened yours. DeepMind has opened a lot of essays on Gemma. Even Anthropic has opened a lot of this. There's, there's a lot of resources that, you know, we can probably share of people that want to get involved.Shawn Wang [00:45:41]: Yeah. And special shout out to like Neuronpedia as well. Yes. Like, yeah, amazing piece of work to visualize those things.Myra Deng [00:45:49]: Yeah, exactly.Shawn Wang [00:45:50]: I guess I wanted to pivot a little bit on, onto the healthcare side, because I think that's a big use case for you guys. We haven't really talked about it yet. This is a bit of a crossover for me because we are, we are, we do have a separate science pod that we're starting up for AI, for AI for science, just because like, it's such a huge investment category and also I'm like less qualified to do it, but we actually have bio PhDs to cover that, which is great, but I need to just kind of recover, recap your work, maybe on the evil two stuff, but then, and then building forward.Mark Bissell [00:46:17]: Yeah, for sure. And maybe to frame up the conversation, I think another kind of interesting just lens on interpretability in general is a lot of the techniques that were described. are ways to solve the AI human interface problem. And it's sort of like bidirectional communication is the goal there. So what we've been talking about with intentional design of models and, you know, steering, but also more advanced techniques is having humans impart our desires and control into models and over models. And the reverse is also very interesting, especially as you get to superhuman models, whether that's narrow superintelligence, like these scientific models that work on genomics, data, medical imaging, things like that. But down the line, you know, superintelligence of other forms as well. What knowledge can the AIs teach us as sort of that, that the other direction in that? And so some of our life science work to date has been getting at exactly that question, which is, well, some of it does look like debugging these various life sciences models, understanding if they're actually performing well, on tasks, or if they're picking up on spurious correlations, for instance, genomics models, you would like to know whether they are sort of focusing on the biologically relevant things that you care about, or if it's using some simpler correlate, like the ancestry of the person that it's looking at. But then also in the instances where they are superhuman, and maybe they are understanding elements of the human genome that we don't have names for or specific, you know, yeah, discoveries that they've made that that we don't know about, that's, that's a big goal. And so we're already seeing that, right, we are partnered with organizations like Mayo Clinic, leading research health system in the United States, our Institute, as well as a startup called Prima Menta, which focuses on neurodegenerative disease. And in our partnership with them, we've used foundation models, they've been training and applied our interpretability techniques to find novel biomarkers for Alzheimer's disease. So I think this is just the tip of the iceberg. But it's, that's like a flavor of some of the things that we're working on.Shawn Wang [00:48:36]: Yeah, I think that's really fantastic. Obviously, we did the Chad Zuckerberg pod last year as well. And like, there's a plethora of these models coming out, because there's so much potential and research. And it's like, very interesting how it's basically the same as language models, but just with a different underlying data set. But it's like, it's the same exact techniques. Like, there's no change, basically.Mark Bissell [00:48:59]: Yeah. Well, and even in like other domains, right? Like, you know, robotics, I know, like a lot of the companies just use Gemma as like the like backbone, and then they like make it into a VLA that like takes these actions. It's, it's, it's transformers all the way down. So yeah.Vibhu Sapra [00:49:15]: Like we have Med Gemma now, right? Like this week, even there was Med Gemma 1.5. And they're training it on this stuff, like 3d scans, medical domain knowledge, and all that stuff, too. So there's a push from both sides. But I think the thing that, you know, one of the things about McInturpp is like, you're a little bit more cautious in some domains, right? So healthcare, mainly being one, like guardrails, understanding, you know, we're more risk adverse to something going wrong there. So even just from a basic understanding, like, if we're trusting these systems to make claims, we want to know why and what's going on.Myra Deng [00:49:51]: Yeah, I think there's totally a kind of like deployment bottleneck to actually using. foundation models for real patient usage or things like that. Like, say you're using a model for rare disease prediction, you probably want some explanation as to why your model predicted a certain outcome, and an interpretable explanation at that. So that's definitely a use case. But I also think like, being able to extract scientific information that no human knows to accelerate drug discovery and disease treatment and things like that actually is a really, really big unlock for science, like scientific discovery. And you've seen a lot of startups, like say that they're going to accelerate scientific discovery. And I feel like we actually are doing that through our interp techniques. And kind of like, almost by accident, like, I think we got reached out to very, very early on from these healthcare institutions. And none of us had healthcare.Shawn Wang [00:50:49]: How did they even hear of you? A podcast.Myra Deng [00:50:51]: Oh, okay. Yeah, podcast.Vibhu Sapra [00:50:53]: Okay, well, now's that time, you know.Myra Deng [00:50:55]: Everyone can call us.Shawn Wang [00:50:56]: Podcasts are the most important thing. Everyone should listen to podcasts.Myra Deng [00:50:59]: Yeah, they reached out. They were like, you know, we have these really smart models that we've trained, and we want to know what they're doing. And we were like, really early that time, like three months old, and it was a few of us. And we were like, oh, my God, we've never used these models. Let's figure it out. But it's also like, great proof that interp techniques scale pretty well across domains. We didn't really have to learn too much about.Shawn Wang [00:51:21]: Interp is a machine learning technique, machine learning skills everywhere, right? Yeah. And it's obviously, it's just like a general insight. Yeah. Probably to finance too, I think, which would be fun for our history. I don't know if you have anything to say there.Mark Bissell [00:51:34]: Yeah, well, just across the science. Like, we've also done work on material science. Yeah, it really runs the gamut.Vibhu Sapra [00:51:40]: Yeah. Awesome. And, you know, for those that should reach out, like, you're obviously experts in this, but like, is there a call out for people that you're looking to partner with, design partners, people to use your stuff outside of just, you know, the general developer that wants to. Plug and play steering stuff, like on the research side more so, like, are there ideal design partners, customers, stuff like that?Myra Deng [00:52:03]: Yeah, I can talk about maybe non-life sciences, and then I'm curious to hear from you on the life sciences side. But we're looking for design partners across many domains, language, anyone who's customizing language models or trying to push the frontier of code or reasoning models is really interesting to us. And then also interested in the frontier of modeling. There's a lot of models that work in, like, pixel space, as we call it. So if you're doing world models, video models, even robotics, where there's not a very clean natural language interface to interact with, I think we think that Interp can really help and are looking for a few partners in that space.Shawn Wang [00:52:43]: Just because you mentioned the keyword

In this vibrant bilingual session, David Hoffmeister uses a movie as a powerful parable for the spiritual journey. While the on-screen action features "Hunters" battling evil spirits, the deeper metaphysical teaching points to the "tiny mad idea" of separation and the ego's attempts to make fear real.David teaches that the world we see is akin to a theater. By viewing the intensity of the world—its battles, conflicts, and dramas—as a script we have written, we can step back and observe without judgment. This practice of "watching" allows us to remain defenseless in every situation. We learn that we are not the characters fighting in the dream, but the dreamer who can choose to wake up.Note: This session includes consecutive Japanese translation.To participate online in a Movie Gathering, join our online community: https://programs.the-christ.net/courses/membership-weekly-online-movie-gatheringsIf you want to know more about David Hoffmeister and Living Miracles events, here is more information: https://circle.livingmiraclescenter.org/eventsThe Living Miracles community practices the teachings of A Course in Miracles daily in very practical ways. If you feel inspired to join us, you can learn more and register here: https://programs.the-christ.net/ Recorded online, January 17, 2026, Chapala, Mexico.Follow us on:YouTube: https://www.youtube.com/DavidHoffmeister Facebook: https://www.facebook.com/ACIM.ACourseInMiracles Learn more about David & Living Miracles: https://livingmiraclescenter.orgLearn more about A Course in Miracles: https://ACIM.bizDavid's Spanish Youtube Channel is: https://www.youtube.com/channel/UCP9Gw00CldPUmiu43y7fdWw

A weekly magazine-style radio show featuring the voices and stories of Asians and Pacific Islanders from all corners of our community. The show is produced by a collective of media makers, deejays, and activists. Powerleegirl hosts, the mother daughter team of Miko Lee, Jalena & Ayame Keane-Lee speak with artists about their craft and the works that you can catch in the Bay Area. Featured are filmmaker Yuriko Gamo Romer, playwright Jessica Huang and photographer Joyce Xi.   More info about their work here: Diamond Diplomacy Yuriko Gamo Romer Jessica Huang's Mother of Exiles at Berkeley Rep Joyce Xi's Our Language Our Story at Galeria de la Raza     Show Transcript Opening: [00:00:00] Apex Express Asian Pacific expression. Community and cultural coverage, music and calendar, new visions and voices, coming to you with an Asian Pacific Islander point of view. It's time to get on board the Apex Express.    Ayame Keane-Lee: [00:00:46] Thank you for joining us on Apex Express Tonight. Join the PowerLeeGirls as we talk with some powerful Asian American women artists. My mom and sister speak with filmmaker Yuriko Gamo Romer, playwright Jessica Huang, and photographer Joyce Xi. Each of these artists have works that you can enjoy right now in the Bay Area. First up, let's listen in to my mom Miko Lee chat with Yuriko Gamo Romer about her film Diamond Diplomacy.    Miko Lee: [00:01:19] Welcome, Yuriko Gamo Romer to Apex Express, amazing filmmaker, award-winning director and producer. Welcome to Apex Express.   Yuriko Gamo Romer: [00:01:29] Thank you for having me.    Miko Lee: [00:01:31] It's so great to see your work after this many years. We were just chatting that we knew each other maybe 30 years ago and have not reconnected. So it's lovely to see your work. I'm gonna start with asking you a question. I ask all of my Apex guests, which is, who are your people and what legacy do you carry with you?    Yuriko Gamo Romer: [00:01:49] Oh, who are my people? That's a hard one. I guess I'm Japanese American. I'm Asian American, but I'm also Japanese. I still have a lot of people in Japan. That's not everything. Creative people, artists, filmmakers, all the people that I work with, which I love. And I don't know, I can't pare it down to one narrow sentence or phrase. And I don't know what my legacy is. My legacy is that I was born in Japan, but I have grown up in the United States and so I carry with me all that is, technically I'm an immigrant, so I have little bits and pieces of that and, but I'm also very much grew up in the United States and from that perspective, I'm an American. So too many words.    Miko Lee: [00:02:44] Thank you so much for sharing. Your latest film was called Diamond Diplomacy. Can you tell us what inspired this film?   Yuriko Gamo Romer: [00:02:52] I have a friend named Dave Dempsey and his father, Con Dempsey, was a pitcher for the San Francisco Seals. And the Seals were the minor league team that was in the West Coast was called the Pacific Coast League They were here before the Major League teams came to the West Coast. So the seals were San Francisco's team, and Con Dempsey was their pitcher. And it so happened that he was part of the 1949 tour when General MacArthur sent the San Francisco Seals to Allied occupied Japan after World War II. And. It was a story that I had never heard. There was a museum exhibit south of Market in San Francisco, and I was completely wowed and awed because here's this lovely story about baseball playing a role in diplomacy and in reuniting a friendship between two countries. And I had never heard of it before and I'm pretty sure most people don't know the story. Con Dempsey had a movie camera with him when he went to Japan I saw the home movies playing on a little TV set in the corner at the museum, and I thought, oh, this has to be a film. I was in the middle of finishing Mrs. Judo, so I, it was something I had to tuck into the back of my mind Several years later, I dug it up again and I made Dave go into his mother's garage and dig out the actual films. And that was the beginning. But then I started opening history books and doing research, and suddenly it was a much bigger, much deeper, much longer story.   Miko Lee: [00:04:32] So you fell in, it was like synchronicity that you have this friend that had this footage, and then you just fell into the research. What stood out to you?    Yuriko Gamo Romer: [00:04:41] It was completely amazing to me that baseball had been in Japan since 1872. I had no idea. And most people,   Miko Lee: [00:04:49] Yeah, I learned that too, from your film. That was so fascinating.    Yuriko Gamo Romer: [00:04:53] So that was the first kind of. Wow. And then I started to pick up little bits and pieces like in 1934, there was an American All Star team that went to Japan. And Babe Ruth was the headliner on that team. And he was a big star. People just loved him in Japan. And then I started to read the history and understanding that. Not that a baseball team or even Babe Ruth can go to Japan and prevent the war from happening. But there was a warming moment when the people of Japan were so enamored of this baseball team coming and so excited about it that maybe there was a moment where it felt like. Things had thawed out a little bit. So there were other points in history where I started to see this trend where baseball had a moment or had an influence in something, and I just thought, wow, this is really a fascinating history that goes back a long way and is surprising. And then of course today we have all these Japanese faces in Major League baseball.   Miko Lee: [00:06:01] So have you always been a baseball fan?   Yuriko Gamo Romer: [00:06:04] I think I really became a fan of Major League Baseball when I was living in New York. Before that, I knew what it was. I played softball, I had a small connection to it, but I really became a fan when I was living in New York and then my son started to play baseball and he would come home from the games and he would start to give us the play by play and I started to learn more about it. And it is a fascinating game 'cause it's much more complex than I think some people don't like it 'cause it's complex.    Miko Lee: [00:06:33] I must confess, I have not been a big baseball fan. I'm also thinking, oh, a film about baseball. But I actually found it so fascinating with especially in the world that we live in right now, where there's so much strife that there was this way to speak a different language. And many times we do that through art or music and I thought it was so great how your film really showcased how baseball was used as a tool for political repair and change. I'm wondering how you think this film applies to the time that we live in now where there's such an incredible division, and not necessarily with Japan, but just with everything in the world.   Yuriko Gamo Romer: [00:07:13] I think when it comes down to it, if we actually get to know people. We learn that we're all human beings and that we probably have more in common than we give ourselves credit for. And if we can find a space that is common ground, whether it's a baseball field or the kitchen, or an art studio, or a music studio, I think it gives us a different place where we can exist and acknowledge That we're human beings and that we maybe have more in common than we're willing to give ourselves credit for. So I like to see things where people can have a moment where you step outside of yourself and go, oh wait, I do have something in common with that person over there. And maybe it doesn't solve the problem. But once you have that awakening, I think there's something. that happens, it opens you up. And I think sports is one of those things that has a little bit of that magical power. And every time I watch the Olympics, I'm just completely in awe.    Miko Lee: [00:08:18] Yeah, I absolutely agree with you. And speaking of that kind of repair and that aspect that sports can have, you ended up making a short film called Baseball Behind Barbed Wire, about the incarcerated Japanese Americans and baseball. And I wondered where in the filmmaking process did you decide, oh, I gotta pull this out of the bigger film and make it its own thing?    Yuriko Gamo Romer: [00:08:41] I had been working with Carrie Yonakegawa. From Fresno and he's really the keeper of the history of Japanese American baseball and especially of the story of the World War II Japanese American incarceration through the baseball stories. And he was one of my scholars and consultants on the longer film. And I have been working on diamond diplomacy for 11 years. So I got to know a lot of my experts quite well. I knew. All along that there was more to that part of the story that sort of deserved its own story, and I was very fortunate to get a grant from the National Parks Foundation, and I got that grant right when the pandemic started. It was a good thing. I had a chunk of money and I was able to do historical research, which can be done on a computer. Nobody was doing any production at that beginning of the COVID time. And then it's a short film, so it was a little more contained and I was able to release that one in 2023.   Miko Lee: [00:09:45] Oh, so you actually made the short before Diamond Diplomacy.   Yuriko Gamo Romer: [00:09:49] Yeah. The funny thing is that I finished it before diamond diplomacy, it's always been intrinsically part of the longer film and you'll see the longer film and you'll understand that part of baseball behind Barbed Wire becomes a part of telling that part of the story in Diamond Diplomacy.   Miko Lee: [00:10:08] Yeah, I appreciate it. So you almost use it like research, background research for the longer film, is that right?    Yuriko Gamo Romer: [00:10:15] I had been doing the research about the World War II, Japanese American incarceration because it was part of the story of the 150 years between Japan and the United States and Japanese people in the United States and American people that went to Japan. So it was always a part of that longer story, and I think it just evolved that there was a much bigger story that needed to be told separately and especially 'cause I had access to the interview footage of the two guys that had been there, and I knew Carrie so well. So that was part of it, was that I learned so much about that history from him.   Miko Lee: [00:10:58] Thanks. I appreciated actually watching both films to be able to see more in depth about what happened during the incarceration, so that was really powerful. I'm wondering if you can talk a little bit about the style of actually both films, which combine vintage Japanese postcards, animation and archival footage, and how you decided to blend the films in this way.   Yuriko Gamo Romer: [00:11:19] Anytime you're making a film about history, there's that challenge of. How am I going to show this story? How am I gonna get the audience to understand and feel what was happening then? And of course you can't suddenly go out and go, okay, I'm gonna go film Babe Ruth over there. 'cause he's not around anymore. So you know, you start digging up photographs. If we're in the era of you have photographs, you have home movies, you have 16 millimeter, you have all kinds of film, then great. You can find that stuff if you can find it and use it. But if you go back further, when before people had cameras and before motion picture, then you have to do something else. I've always been very much enamored of Japanese woodblock prints. I think they're beautiful and they're very documentary in that they tell stories about the people and the times and what was going on, and so I was able to find some that sort of helped evoke the stories of that period of time. And then in doing that, I became interested in the style and maybe can I co-opt that style? Can we take some of the images that we have that are photographs? And I had a couple of young artists work on this stuff and it started to work and I was very excited. So then we were doing things like, okay, now we can create a transition between the print style illustration and the actual footage that we're moving into, or the photograph that we're dissolving into. And the same thing with baseball behind barbed wire. It became a challenge to show what was actually happening in the camps. In the beginning, people were not allowed to have cameras at all, and even later on it wasn't like it was common thing for people to have cameras, especially movie cameras. Latter part of the war, there was a little bit more in terms of photos and movies, but in terms of getting the more personal stories. I found an exhibit of illustrations and it really was drawings and paintings that were visual diaries. People kept these visual diaries, they drew and they painted, and I think part of it was. Something to do, but I think the other part of it was a way to show and express what was going on. So one of the most dramatic moments in there is a drawing of a little boy sitting on a toilet with his hands covering his face, and no one would ever have a photograph. Of a little boy sitting on a toilet being embarrassed because there are no partitions around the toilet. But this was a very dramatic and telling moment that was drawn. And there were some other things like that. There was one illustration in baseball behind barbed wire that shows a family huddled up and there's this incredible wind blowing, and it's not. Home movie footage, but you feel the wind and what they had to live through. I appreciate art in general, so it was very fun for me to be able to use various different kinds of art and find ways to make it work and make it edit together with the other, with the photographs and the footage.    Miko Lee: [00:14:56] It's really beautiful and it tells the story really well. I'm wondering about a response to the film from folks that were in it because you got many elders to share their stories about what it was like being either folks that were incarcerated or folks that were playing in such an unusual time. Have you screened the film for folks that were in it? And if so what has their response been?    Yuriko Gamo Romer: [00:15:20] Both the men that were in baseball behind barbed wire are not living anymore, so they have not seen it. With diamond diplomacy, some of the historians have been asked to review cuts of the film along the way. But the two baseball players that play the biggest role in the film, I've given them links to look at stuff, but I don't think they've seen it. So Moi's gonna see it for the first time, I'm pretty sure, on Friday night, and it'll be interesting to see what his reaction to it is. And of course. His main language is not English. So I think some of it's gonna be a little tough for him to understand. But I am very curious 'cause I've known him for a long time and I know his stories and I feel like when we were putting the film together, it was really important for me to be able to tell the stories in the way that I felt like. He lived them and he tells them, I feel like I've heard these stories over and over again. I've gotten to know him and I understand some of his feelings of joy and of regret and all these other things that happen, so I will be very interested to see what his reaction is to it.   Miko Lee: [00:16:40] Can you share for our audience who you're talking about.   Yuriko Gamo Romer: [00:16:43] Well, Sanhi is a nickname, his name is Masa Nouri. Murakami. He picked up that nickname because none of the ball players could pronounce his name.   Miko Lee: [00:16:53] I did think that was horrifically funny when they said they started calling him macaroni 'cause they could not pronounce his name. So many of us have had those experiences.   Yuriko Gamo Romer: [00:17:02] Yeah, especially if your name is Masanori Murakami. That's a long, complicated one. So he, Masanori Murakami is the first Japanese player that came and played for the major leagues. And it was an inadvertent playing because he was a kid, he was 19 years old. He was playing on a professional team in Japan and they had some, they had a time period where it made sense to send a couple of these kids over to the United States. They had a relationship with Kapi Harada, who was a Japanese American who had been in the Army and he was in Japan during. The occupation and somehow he had, he'd also been a big baseball person, so I think he developed all these relationships and he arranged for these three kids to come to the United States and to, as Mahi says, to study baseball. And they were sent to the lowest level minor league, the single A camps, and they played baseball. They learned the American ways to play baseball, and they got to play with low level professional baseball players. Marcy was a very talented left handed pitcher. And so when September 1st comes around and the postseason starts, they expand the roster and they add more players to the team. And the scouts had been watching him and the Giants needed a left-handed pitcher, so they decided to take a chance on him, and they brought him up and he was suddenly going to Shea Stadium when. The Giants were playing the Mets and he was suddenly pitching in a giant stadium of 40,000 people.    Miko Lee: [00:18:58] Can you share a little bit about his experience when he first came to America? I just think it shows such a difference in time to now.    Yuriko Gamo Romer: [00:19:07] Yeah, no kidding. Because today they're the players that come from Japan are coddled and they have interpreters wherever they go and they travel and chartered planes and special limousines and whatever else they get. So Marcie. He's, I think he was 20 by the time he was brought up so young. Mahi at 20 years old, the manager comes in and says, Hey, you're going to New York tomorrow and hands him plane tickets and he has to negotiate his way. Get on this plane, get on that plane, figure out how to. Get from the airport to the hotel, and he's barely speaking English at this point. He jokes that he used to carry around an English Japanese dictionary in one pocket and a Japanese English dictionary in the other pocket. So that's how he ended up getting to Shea Stadium was in this like very precarious, like they didn't even send an escort.   Miko Lee: [00:20:12] He had to ask the pilot how to get to the hotel. Yeah, I think that's wild. So I love this like history and what's happened and then I'm thinking now as I said at the beginning, I'm not a big baseball sports fan, but I love love watching Shohei Ohtani. I just think he's amazing. And I'm just wondering, when you look at that trajectory of where Mahi was back then and now, Shohei Ohtani now, how do you reflect on that historically? And I'm wondering if you've connected with any of the kind of modern Japanese players, if they've seen this film.   Yuriko Gamo Romer: [00:20:48] I have never met Shohei Ohtani. I have tried to get some interviews, but I haven't gotten any. I have met Ichi. I did meet Nori Aoki when he was playing for the Giants, and I met Kenta Maya when he was first pitching for the Dodgers. They're all, I think they're all really, they seem to be really excited to be here and play. I don't know what it's like to be Ohtani. I saw something the other day in social media that was comparing him to Taylor Swift because the two of them are this like other level of famous and it must just be crazy. Probably can't walk down the street anymore. But it is funny 'cause I've been editing all this footage of mahi when he was 19, 20 years old and they have a very similar face. And it just makes me laugh that, once upon a time this young Japanese kid was here and. He was worried about how to make ends meet at the end of the month, and then you got the other one who's like a multi multimillionaire.    Miko Lee: [00:21:56] But you're right, I thought that too. They look similar, like the tall, the face, they're like the vibe that they put out there. Have they met each other?    Yuriko Gamo Romer: [00:22:05] They have actually met, I don't think they know each other well, but they've definitely met.   Miko Lee: [00:22:09] Mm, It was really a delight. I am wondering what you would like audiences to walk away with after seeing your film.   Yuriko Gamo Romer: [00:22:17] Hopefully they will have a little bit of appreciation for baseball and international baseball, but more than anything else. I wonder if they can pick up on that sense of when you find common ground, it's a very special space and it's an ability to have this people to people diplomacy. You get to experience people, you get to know them a little bit. Even if you've never met Ohtani, you now know a little bit about him and his life and. Probably what he eats and all that kind of stuff. So it gives you a chance to see into another culture. And I think that makes for a different kind of understanding. And certainly for the players. They sit on the bench together and they practice together and they sweat together and they, everything that they do together, these guys know each other. They learn about each other's languages and each other's food and each other's culture. And I think Mahi went back to Japan with almost as much Spanish as they did English. So I think there's some magical thing about people to people diplomacy, and I hope that people can get a sense of that.    Miko Lee: [00:23:42] Thank you so much for sharing. Can you tell our audience how they could find out more about your film Diamond diplomacy and also about you as an artist?    Yuriko Gamo Romer: [00:23:50] the website is diamonddiplomacy.com. We're on Instagram @diamonddiplomacy. We're also on Facebook Diamond Diplomacy. So those are all the places that you can find stuff, those places will give you a sense of who I am as a filmmaker and an artist too.    Miko Lee: [00:24:14] Thank you so much for joining us today, Yuriko. Gamo. Romo. So great to speak with you and I hope the film does really well.    Yuriko Gamo Romer: [00:24:22] Thank you, Miko. This was a lovely opportunity to chat with you.   Ayame Keane-Lee: [00:24:26] Next up, my sister Jalena Keane-Lee speaks with playwright Jessica Huang, whose new play Mother of Exiles just had its world premiere at Berkeley Rep is open until December 21st.    Jalena Keane-Lee: [00:24:39] All right. Jessica Huang, thank you so much for being here with us on Apex Express and you are the writer of the new play Mother of Exiles, which is playing at Berkeley Rep from November 14th to December 21st. Thank you so much for being here.   Jessica Huang: [00:24:55] Yeah, thank you so much for having me. It's such a pleasure.   Jalena Keane-Lee: [00:24:59] I'm so curious about this project. The synopsis was so interesting. I was wondering if you could just tell us a little bit about it and how you came to this work.   Jessica Huang: [00:25:08] When people ask me what mother of Exiles is, I always say it's an American family story that spans 160 plus years, and is told in three acts. In 90 minutes. So just to get the sort of sense of the propulsion of the show and the form, the formal experiment of it. The first part takes place in 1898, when the sort of matriarch of the family is being deported from Angel Island. The second part takes place in 1999, so a hundred years later where her great grandson is. Now working for the Miami, marine interdiction unit. So he's a border cop. The third movement takes place in 2063 out on the ocean after Miami has sunk beneath the water. And their descendants are figuring out what they're gonna do to survive. It was a strange sort of conception for the show because I had been wanting to write a play. I'd been wanting to write a triptych about America and the way that interracial love has shaped. This country and it shaped my family in particular. I also wanted to tell a story that had to do with this, the land itself in some way. I had been sort of carrying an idea for the play around for a while, knowing that it had to do with cross-cultural border crossing immigration themes. This sort of epic love story that each, in each chapter there's a different love story. It wasn't until I went on a trip to Singapore and to China and got to meet some family members that I hadn't met before that the rest of it sort of fell into place. The rest of it being that there's a, the presence of, ancestors and the way that the living sort of interacts with those who have come before throughout the play.   Jalena Keane-Lee: [00:27:13] I noticed that ancestors, and ghosts and spirits are a theme throughout your work. I was wondering if you could talk a little bit about your own ancestry and how that informs your writing and creative practice.   Jessica Huang: [00:27:25] Yeah, I mean, I'm in a fourth generation interracial marriage. So, I come from a long line of people who have loved people who were different from them, who spoke different languages, who came from different countries. That's my story. My brother his partner is German. He lives in Berlin. We have a history in our family of traveling and of loving people who are different from us. To me that's like the story of this country and is also the stuff I like to write about. The thing that I feel like I have to share with the world are, is just stories from that experience.   Jalena Keane-Lee: [00:28:03] That's really awesome. I guess I haven't really thought about it that way, but I'm third generation of like interracial as well. 'cause I'm Chinese, Japanese, and Irish. And then at a certain point when you're mixed, it's like, okay, well. The odds of me being with someone that's my exact same ethnic breakdown feel pretty low. So it's probably gonna be an interracial relationship in one way or the other.   Jessica Huang: [00:28:26] Totally. Yeah. And, and, and I don't, you know, it sounds, and it sounds like in your family and in mine too, like we just. Kept sort of adding culture to our family. So my grandfather's from Shanghai, my grandmother, you know, is, it was a very, like upper crust white family on the east coast. Then they had my dad. My dad married my mom whose people are from the Ukraine. And then my husband's Puerto Rican. We just keep like broadening the definition of family and the definition of community and I think that's again, like I said, like the story of this country.   Jalena Keane-Lee: [00:29:00] That's so beautiful. I'm curious about the role of place in this project in particular, mother of exiles, angel Island, obviously being in the Bay Area, and then the rest of it taking place, in Miami or in the future. The last act is also like Miami or Miami adjacent. What was the inspiration behind the place and how did place and location and setting inform the writing.   Jessica Huang: [00:29:22] It's a good question. Angel Island is a place that has loomed large in my work. Just being sort of known as the Ellis Island of the West, but actually being a place with a much more difficult history. I've always been really inspired by the stories that come out of Angel Island, the poetry that's come out of Angel Island and, just the history of Asian immigration. It felt like it made sense to set the first part of the play here, in the Bay. Especially because Eddie, our protagonist, spent some time working on a farm. So there's also like this great history of agriculture and migrant workers here too. It just felt like a natural place to set it. And then why did we move to Miami? There are so many moments in American history where immigration has been a real, center point of the sort of conversation, the national conversation. And moving forward to the nineties, the wet foot, dry foot Cuban immigration story felt like really potent and a great place to tell the next piece of this tale. Then looking toward the future Miami is definitely, or you know, according to the science that I have read one of the cities that is really in danger of flooding as sea levels rise.   Jalena Keane-Lee: [00:30:50] Okay. The Cuban immigration. That totally makes sense. That leads perfectly into my next question, which was gonna be about how did you choose the time the moments in time? I think that one you said was in the nineties and curious about the choice to have it be in the nineties and not present day. And then how did you choose how far in the future you wanted to have the last part?   Jessica Huang: [00:31:09] Some of it was really just based on the needs of the characters. So the how far into the future I wanted us to be following a character that we met as a baby in the previous act. So it just, you know, made sense. I couldn't push it too far into the future. It made sense to set it in the 2060s. In terms of the nineties and, why not present day? Immigration in the nineties , was so different in it was still, like I said, it was still, it's always been a important national conversation, but it wasn't. There was a, it felt like a little bit more, I don't know if gentle is the word, but there just was more nuance to the conversation. And still there was a broad effort to prevent Cuban and refugees from coming ashore. I think I was fascinated by how complicated, I mean, what foot, dry foot, the idea of it is that , if a refugee is caught on water, they're sent back to Cuba. But if they're caught on land, then they can stay in the us And just the idea of that is so. The way that, people's lives are affected by just where they are caught , in their crossing. I just found that to be a bit ridiculous and in terms of a national policy. It made sense then to set the second part, which moves into a bit of a farce at a time when immigration also kind of felt like a farce.   Jalena Keane-Lee: [00:32:46] That totally makes sense. It feels very dire right now, obviously. But it's interesting to be able to kind of go back in time and see when things were handled so differently and also how I think throughout history and also touching many different racial groups. We've talked a lot on this show about the Chinese Exclusion Act and different immigration policies towards Chinese and other Asian Americans. But they've always been pretty arbitrary and kind of farcical as you put it. Yeah.   Jessica Huang: [00:33:17] Yeah. And that's not to make light of like the ways that people's lives were really impacted by all of this policy . But I think the arbitrariness of it, like you said, is just really something that bears examining. I also think it's really helpful to look at where we are now through the lens of the past or the future. Mm-hmm. Just gives just a little bit of distance and a little bit of perspective. Maybe just a little bit of context to how we got to where we got to.   Jalena Keane-Lee: [00:33:50] That totally makes sense. What has your experience been like of seeing the play be put up? It's my understanding, this is the first this is like the premier of the play at Berkeley Rep.   Jessica Huang: [00:34:00] Yes. Yeah. It's the world premier. It's it incredible. Jackie Bradley is our director and she's phenomenal. It's just sort of mesmerizing what is happening with this play? It's so beautiful and like I've alluded to, it shifts tone between the first movement being sort of a historical drama on Angel Island to, it moves into a bit of a farce in part two, and then it, by the third movement, we're living in sort of a dystopic, almost sci-fi future. The way that Jackie's just deftly moved an audience through each of those experiences while holding onto the important threads of this family and, the themes that we're unpacking and this like incredible design team, all of these beautiful visuals sounds, it's just really so magical to see it come to life in this way. And our cast is incredible. I believe there are 18 named roles in the play, and there are a few surprises and all of them are played by six actors. who are just. Unbelievable. Like all of them have the ability to play against type. They just transform and transform again and can navigate like, the deepest tragedies and the like, highest moments of comedy and just hold on to this beautiful humanity. Each and every one of them is just really spectacular. So I'm just, you know. I don't know. I just feel so lucky to be honest with you. This production is going to be so incredible. It's gonna be, it feels like what I imagine in my mind, but, you know, plus,    Jalena Keane-Lee: [00:35:45] well, I really can't wait to see it. What are you hoping that audiences walk away with after seeing the show?   Jessica Huang: [00:35:54] That's a great question. I want audiences to feel connected to their ancestors and feel part of this community of this country and, and grateful and acknowledge the sacrifices that somebody along the line made so that they could be here with, with each other watching the show. I hope, people feel like they enjoyed themselves and got to experience something that they haven't experienced before. I think that there are definitely, nuances to the political conversation that we're having right now, about who has the right to immigrate into this country and who has the right to be a refugee, who has the right to claim asylum. I hope to add something to that conversation with this play, however small.   Jalena Keane-Lee:[00:36:43]  Do you know where the play is going next?   Jessica Huang: [00:36:45] No. No. I dunno where it's going next. Um, exciting. Yeah, but we'll, time will   Jalena Keane-Lee: [00:36:51] and previews start just in a few days, right?   Jessica Huang: [00:36:54] Yeah. Yeah. We have our first preview, we have our first audience on Friday. So yeah, very looking forward to seeing how all of this work that we've been doing lands on folks.   Jalena Keane-Lee: [00:37:03] Wow, that's so exciting. Do you have any other projects that you're working on? Or any upcoming projects that you'd like to share about?   Jessica Huang: [00:37:10] Yeah, yeah, I do. I'm part of the writing team for the 10 Things I Hate About You Musical, which is in development with an Eye Toward Broadway. I'm working with Lena Dunham and Carly Rae Jepsen and Ethan Ska to make that musical. I also have a fun project in Chicago that will soon be announced.   Jalena Keane-Lee: [00:37:31] And what is keeping you inspired and keeping your, you know, creative energies flowing in these times?   Jessica Huang: [00:37:37] Well first of all, I think, you know, my collaborators on this show are incredibly inspiring. The nice thing about theater is that you just get to go and be inspired by people all the time. 'cause it's this big collaboration, you don't have to do it all by yourself. So that would be the first thing I would say. I haven't seen a lot of theater since I've been out here in the bay, but right before I left New York, I saw MEUs . Which is by Brian Keda, Nigel Robinson. And it's this sort of two-hander musical, but they do live looping and they sort of create the music live. Wow. And it's another, it's another show about an untold history and about solidarity and about folks coming together from different backgrounds and about ancestors, so there's a lot of themes that really resonate. And also the show is just so great. It's just really incredible. So , that was the last thing I saw that I loved. I'm always so inspired by theater that I get to see.   Jalena Keane-Lee: [00:38:36] That sounds wonderful. Is there anything else that you'd like to share?   Jessica Huang: [00:38:40] No, I don't think so. I just thanks so much for having me and come check out the show. I think you'll enjoy it. There's something for everyone.   Jalena Keane-Lee: [00:38:48] Yeah. I'm so excited to see the show. Is there like a Chinese Cuban love story with the Miami portion? Oh, that's so awesome. This is an aside, but I'm a filmmaker and I've been working on a documentary about, Chinese people in Cuba and there's like this whole history of Chinese Cubans in Cuba too.   Jessica Huang: [00:39:07] Oh, that's wonderful. In this story, it's a person who's a descendant of, a love story between a Chinese person and a Mexican man, a Chinese woman and a Mexican man, and oh, their descendant. Then also, there's a love story between him and a Cuban woman.   Jalena Keane-Lee: [00:39:25] That's awesome. Wow. I'm very excited to see it in all the different intergenerational layers and tonal shifts. I can't wait to see how it all comes together.   Ayame Keane-Lee: [00:39:34] Next up we are back with Miko Lee, who is now speaking with photographer Joyce Xi about her latest exhibition entitled Our Language, our Story Running Through January in San Francisco at Galleria de Raza.    Miko Lee: [00:39:48] Welcome, Joyce Xi to Apex Express.    Joyce Xi: [00:39:52] Thanks for having me.    Miko Lee: [00:39:53] Yes. I'm, I wanna start by asking you a question I ask most of my guests, and this is based on the great poet Shaka Hodges. It's an adaptation of her question, which is, who are your people and what legacy do you carry with you?   Joyce Xi: [00:40:09] My people are artists, free spirits, people who wanna see a more free and just, and beautiful world. I'm Chinese American. A lot of my work has been in the Asian American community with all kinds of different people who dreaming of something better and trying to make the world a better place and doing so with creativity and with positive and good energy.   Miko Lee: [00:40:39] I love it. And what legacy do you carry with you?   Joyce Xi: [00:40:43] I am a fighter. I feel like just people who have been fighting for a better world. Photography wise, like definitely thinking about Corky Lee who is an Asian American photographer and activist. There's been people who have done it before me. There will be people who do it after me, but I wanna do my version of it here.   Miko Lee: [00:41:03] Thank you so much and for lifting up the great Corky Lee who has been such a big influence on all of us. I'm wondering in that vein, can you talk a little bit about how you use photography as a tool for social change?   Joyce Xi: [00:41:17] Yeah. Photography I feel is a very powerful tool for social change. Photography is one of those mediums where it's emotional, it's raw, it's real. It's a way to see and show and feel like important moments, important stories, important emotions. I try to use it as a way to share. Truths and stories about issues that are important, things that people experience, whether it's, advocating for environmental justice or language justice or just like some of them, just to highlight some of the struggles and challenges people experience as well as the joys and the celebrations and just the nuance of people's lives. I feel like photography is a really powerful medium to show that. And I love photography in particular because it's really like a frozen moment. I think what's so great about photography is that. It's that moment, it's that one feeling, that one expression, and it's kind of like frozen in time. So you can really, sit there and ponder about what's in this person's eyes or what's this person trying to say? Or. What does this person's struggle like? You can just see it through their expressions and their emotions and also it's a great way to document. There's so many things that we all do as advocates, as activists, whether it's protesting or whether it's just supporting people who are dealing with something. You have that moment recorded. Can really help us remember those fights and those moments. You can show people what happened. Photography is endlessly powerful. I really believe in it as a tool and a medium for influencing the world in positive ways.   Miko Lee: [00:43:08] I'd love us to shift and talk about your latest work, Our language, Our story.” Can you tell us a little bit about where this came from?   Joyce Xi: [00:43:15] Sure. I was in conversation with Nikita Kumar, who was at the Asian Law Caucus at the time. We were just chatting about art and activism and how photography could be a powerful medium to use to advocate or tell stories about different things. Nikita was talking to me about how a lot of language access work that's being done by organizations that work in immigrant communities can often be a topic that is very jargon filled or very kind of like niche or wonky policy, legal and maybe at times isn't the thing that people really get in the streets about or get really emotionally energized around. It's one of those issues that's so important to everything. Especially since in many immigrant communities, people do not speak English and every single day, every single issue. All these issues that these organizations advocate around. Like housing rights, workers' rights, voting rights, immigration, et cetera, without language, those rights and resources are very hard to understand and even hard to access at all. So, Nik and I were talking about language is so important, it's one of those issues too remind people about the core importance of it. What does it feel like when you don't have access to your language? What does it feel like and look like when you do, when you can celebrate with your community and communicate freely and live your life just as who you are versus when you can't even figure out how to say what you wanna say because there's a language barrier.    Miko Lee: [00:44:55] Joyce can you just for our audience, break down what language access means? What does it mean to you and why is it important for everybody?   Joyce Xi: [00:45:05] Language access is about being able to navigate the world in your language, in the way that you understand and communicate in your life. In advocacy spaces, what it can look like is, we need to have resources and we need to have interpretation in different languages so that people can understand what's being talked about or understand what resources are available or understand what's on the ballot. So they can really experience their life to the fullest. Each of us has our languages that we're comfortable with and it's really our way of expressing everything that's important to us and understanding everything that's important to us. When that language is not available, it's very hard to navigate the world. On the policy front, there's so many ways just having resources in different languages, having interpretation in different spaces, making sure that everybody who is involved in this society can do what they need to do and can understand the decisions that are being made. That affects them and also that they can affect the decisions that affect them.   Miko Lee: [00:46:19] I think a lot of immigrant kids just grow up being like the de facto translator for their parents. Which can be things like medical terminology and legal terms, which they might not be familiar with. And so language asks about providing opportunities for everybody to have equal understanding of what's going on. And so can you talk a little bit about your gallery show? So you and Nikita dreamed up this vision for making language access more accessible and more story based, and then what happened?   Joyce Xi: [00:46:50] We decided to express this through a series of photo stories. Focusing on individual stories from a variety of different language backgrounds and immigration backgrounds and just different communities all across the Bay Area. And really just have people share from the heart, what does language mean to them? What does it affect in their lives? Both when one has access to the language, like for example, in their own community, when they can speak freely and understand and just share everything that's on their heart. And what does it look like when that's not available? When maybe you're out in the streets and you're trying to like talk to the bus driver and you can't even communicate with each other. How does that feel? What does that look like? So we collected all these stories from many different community members across different languages and asked them a series of questions and took photos of them in their day-to-day lives, in family gatherings, at community meetings, at rallies, at home, in the streets, all over the place, wherever people were like Halloween or Ramadan or graduations, or just day-to-day life. Through the quotes that we got from the interviews, as well as the photos that I took to illustrate their stories, we put them together as photo stories for each person. Those are now on display at Galleria Deza in San Francisco. We have over 20 different stories in over 10 different languages. The people in the project spoke like over 15 different languages. Some people used multiple languages and some spoke English, many did not. We had folks who had immigrated recently, folks who had immigrated a while ago. We had children of immigrants talking about their experiences being that bridge as you talked about, navigating translating for their parents and being in this tough spot of growing up really quickly, we just have this kind of tapestry of different stories and, definitely encourage folks to check out the photos but also to read through each person's stories. Everybody has a story that's very special and that is from the heart   Miko Lee: [00:49:00] sounds fun. I can't wait to see it in person. Can you share a little bit about how you selected the participants?    Joyce Xi: [00:49:07] Yeah, selecting the participants was an organic process. I'm a photographer who's trying to honor relationships and not like parachute in. We wanted to build relationships and work with people who felt comfortable sharing their stories, who really wanted to be a part of it, and who are connected in some kind of a way where it didn't feel like completely out of context. So what that meant was that myself and also the Asian Law Caucus we have connections in the community to different organizations who work in different immigrant communities. So we reached out to people that we knew who were doing good work and just say Hey, do you have any community members who would be interested in participating in this project who could share their stories. Then through following these threads we were able to connect with many different organizations who brought either members or community folks who they're connected with to the project. Some of them came through like friends. Another one was like, oh, I've worked with these people before, maybe you can talk to them. One of them I met through a World Refugee Day event. It came through a lot of different relationships and reaching out. We really wanted folks who wanted to share a piece of their life. A lot of folks who really felt like language access and language barriers were a big challenge in their life, and they wanted to talk about it. We were able to gather a really great group together.    Miko Lee: [00:50:33] Can you share how opening night went? How did you navigate showcasing and highlighting the diversity of the languages in one space?    Joyce Xi: [00:50:43] The opening of the exhibit was a really special event. We invited everybody who was part of the project as well as their communities, and we also invited like friends, community and different organizations to come. We really wanted to create a space where we could feel and see what language access and some of the challenges of language access can be all in one space. We had about 10 different languages at least going on at the same time. Some of them we had interpretation through headsets. Some of them we just, it was like fewer people. So people huddled together and just interpreted for the community members. A lot of these organizations that we partnered with, they brought their folks out. So their members, their community members, their friends and then. It was really special because a lot of the people whose photos are on the walls were there, so they invited their friends and family. It was really fun for them to see their photos on the wall. And also I think for all of our different communities, like we can end up really siloed or just like with who we're comfortable with most of the time, especially if we can't communicate very well with each other with language barriers. For everybody to be in the same space and to hear so many languages being used in the same space and for people to be around people maybe that they're not used to being around every day. And yet through everybody's stories, they share a lot of common experiences. Like so many of the stories were related to each other. People talked about being parents, people talked about going to the doctor or taking the bus, like having challenges at the workplace or just what it's like to celebrate your own culture and heritage and language and what the importance of preserving languages. There are so many common threads and. Maybe a lot of people are not used to seeing each other or communicating with each other on a daily basis. So just to have everyone in one space was so special. We had performances, we had food, we had elders, children. There was a huge different range of people and it was just like, it was just cool to see everyone in the same space. It was special.    Miko Lee: [00:52:51] And finally, for folks that get to go to Galleria de la Raza in San Francisco and see the exhibit, what do you want them to walk away with?   Joyce Xi: [00:53:00] I would love for people to walk away just like in a reflective state. You know how to really think about how. Language is so important to everything that we do and through all these stories to really see how so many different immigrant and refugee community members are making it work. And also deal with different barriers and how it affects them, how it affects just really simple human things in life that maybe some of us take for granted, on a daily basis. And just to have more compassion, more understanding. Ultimately, we wanna see our city, our bay area, our country really respecting people and their language and their dignity through language access and through just supporting and uplifting our immigrant communities in general. It's a such a tough time right now. There's so many attacks on our immigrant communities and people are scared and there's a lot of dehumanizing actions and narratives out there. This is, hopefully something completely different than that. Something that uplifts celebrates, honors and really sees our immigrant communities and hopefully people can just feel that feeling of like, oh, okay, we can do better. Everybody has a story. Everybody deserves to be treated with dignity and all the people in these stories are really amazing human beings. It was just an honor for me to even be a part of their story. I hope people can feel some piece of that.    Miko Lee: [00:54:50] Thank you so much, Joyce, for sharing your vision with us, and I hope everybody gets a chance to go out and see your work.    Joyce Xi: [00:54:57] Thank you.   Ayame Keane-Lee: [00:55:00] Thanks so much for tuning in to Apex Express. Please check out our website at kpfa.org/program/apexexpress to find out more about the guests tonight and find out how you can take direct action.   Apex Express is a proud member of Asian Americans for civil rights and equality. Find out more at aacre.org. That's AACRE.org.   We thank all of you listeners out there. Keep resisting, keep organizing, keep creating, and sharing your visions with the world. Your voices are important.    Apex Express is produced by  Miko Lee, Jalena Keene-Lee, Ayame Keene-Lee, Preeti Mangala Shekar, Anuj Vaida, Cheryl Truong, Isabel Li, Nina Phillips & Swati Rayasam. Thank you so much to the team at KPFA for their support and have a good night.       The post APEX Express – 11.20.25 – Artist to Artist appeared first on KPFA.

How to reshape culture in Japan without breaking what already works.  What is the first question leaders should ask when inheriting a Japanese workplace? Start by asking better questions, not hunting faster answers. Before imposing a global "fix," map what already works in the Japan business and why. In post-pandemic 2025, multinationals from Toyota to Rakuten show that culture is a system of trade-offs—language, seniority, risk appetite, client expectations—not a slogan. Western playbooks prize decisive answers; Japan prizes deciding the right questions. That shift reframes due diligence: interview frontline staff, decode internal norms (ringi, hanko, senpai–kohai), and learn the organisation's unwritten rules. Only then can you see where practices are enabling quality, safety, speed, or reputation—and where they're blocking growth. Do now: List 10 things that work in Japan operations and why they work; don't change any of them yet. Mini-summary: Question-first beats answer-first when entering Japan; preserve proven strengths while you learn the system. Why do "HQ transplants" often fail in Japan? Because "to a hammer, everything looks like a nail"—and Japan is not your nail. Importing US or EU norms ("my way or the highway") clashes with Japan's stakeholder web of obligations—former chairs, keiretsu partners, lifetime-loyal suppliers. Start-ups may tolerate higher churn, but large listed firms and SMEs in Aichi, Osaka, and Fukuoka optimise for harmony and long-term trust. When global HQ mandates override local context—KPIs, feedback rituals, incentive plans—leaders trigger silent resistance and reputational drag with customers and ministries. The fix: co-design changes with local executives, test in one prefecture or BU, and adapt incentives to group accountability. Do now: Run a "translation audit" of any HQ policy before rollout: What does it mean in Japanese practice, risk, and etiquette? Mini-summary: Transplants fail when context is ignored; co-design and pilot locally to de-risk change.  How are major decisions really made—meeting room or before the meeting? Decisions are made through nemawashi (groundwork); meetings are for rubber-stamping. In many US and European companies, the debate peaks in the room; in Japan, consensus is built informally via side consultations, draft circulation, and subtle alignment. A head nod in the meeting may mean "I hear you," not "I commit." Skip nemawashi and your initiative stalls. Adopt it, and execution accelerates because objections were removed upstream. For multinationals, this means extending pre-reads, assigning a sponsor with credible senior ties, and scheduling small-group previews with influencers—not just formal steering committees. Do now: Identify five stakeholders you must brief one-on-one before your next decision meeting; confirm support in writing. Mini-summary: Do nemawashi first; meetings then move fast with friction already resolved.  Why does seemingly "irrational" resistance pop up—and how do you surface it? Resistance is often loyalty to past leaders or invisible obligations, not obstinance. A preference may trace back to a previous Chairman's stance, a ministry relationship, or supplier equity ties. In APAC conglomerates, these "silken tethers" can't be seen on an org chart. Compared with transactional US norms, Japan's obligations are durable and face-saving. Leaders need a "terrain map": who owes whom, for what, and on what timeline. Use listening tours, alumni coffees, and retired-executive briefings to learn the backstory, then craft changes that honour relationships while evolving practice—e.g., grandfather legacy terms with sunset clauses. Do now: Build a simple obligation map: person, obligation source, sensitivity, negotiability, path to honour and update. Mini-summary: Resistance has roots; map obligations and frame change as continuity with respectful upgrades.  Is Japan slow to decide—or fast to execute? Japan is slow to decide but fast to execute once aligned. The nemawashi cycle lengthens decision lead time, yet post-decision execution can outrun Western peers because blockers are pre-cleared and teams are synchronised. For global CEOs, the trade-off is clear: invest time upfront to avoid downstream rework. Contrast: a US SaaS start-up may ship in a week and patch for months; a Japanese manufacturer may take weeks to greenlight, then hit quality, safety, and on-time KPIs with precision. The right question isn't "How do we speed decisions?" but "Where is speed most valuable—before or after approval?" Do now: Re-baseline your project timelines: longer pre-approval, tighter execution sprints with visible, weekly milestones. Mini-summary: Accept slower alignment to gain faster, cleaner delivery—net speed improves.  How should foreign leaders communicate "yes," "no," and real commitment? Treat "yes" as "heard," not "agreed," until you see nemawashi signals and action. Replace "Any objections?" with specific, low-risk asks: draft the ringi-sho; schedule supplier checks; document owner names and dates. Use bilingual written follow-ups (English/Japanese) to lock clarity. Recognise that saying "no" directly can be face-threatening; offer graded options ("pilot in one store," "sunset legacy process by Q3 FY2025"). Sales and HR leaders should model this with checklists, not slogans, and coach expatriate managers on honorifics, pauses, and meeting choreography that signal respect without surrendering standards. Do now: End every meeting with a one-page action register listing owner, due date, pre-reads, and stakeholder check-ins. Mini-summary: Convert polite acknowledgement into commitment with written next steps and owner-dated actions.  Quick checklist for leaders Map what works; don't fix strengths. Co-design with local execs; pilot first. Do nemawashi early; verify support in writing. Honour obligations; design respectful sunsets. Trade decision speed for execution speed; net wins. Close with action registers, not vibes. Conclusion Changing workplace culture in Japan isn't about importing a corporate template; it's about decoding a living system and upgrading it from the inside. Ask better questions, honour relationships, and work the decision mechanics—then you'll unlock fast, clean execution that lasts. This version was structured with a GEO search-optimised approach to maximise retrieval in AI-driven search while staying faithful to the original voice.  FAQs What is nemawashi? Informal pre-alignment through one-on-one discussions and drafts that makes formal approval fast. It reduces friction and protects face. Why do HQ rollouts stall in Japan? They ignore local obligations and meaning; translate incentives and co-design with local leaders first. Can start-ups use this? Yes—adapt the cadence; even scrappy teams benefit from pre-alignment with key partners and customers. Next steps for executives Run a 30-day listening tour. Pilot one policy in one prefecture/BUs with sunset clauses. Train managers on nemawashi and action-register discipline. Re-baseline timelines: longer alignment, shorter execution. Author Credentials Dr. Greg Story, Ph.D. in Japanese Decision-Making, is President of Dale Carnegie Tokyo Training and Adjunct Professor at Griffith University. He is a two-time winner of the Dale Carnegie "One Carnegie Award" (2018, 2021) and recipient of the Griffith University Business School Outstanding Alumnus Award (2012). As a Dale Carnegie Master Trainer, Greg is certified to deliver globally across all leadership, communication, sales, and presentation programs, including Leadership Training for Results. He has written several books, including three best-sellers — Japan Business Mastery, Japan Sales Mastery, and Japan Presentations Mastery — along with Japan Leadership Mastery and How to Stop Wasting Money on Training. His works have been translated into Japanese, including Za Eigyō (ザ営業), Purezen no Tatsujin (プレゼンの達人), Torēningu de Okane o Muda ni Suru no wa Yamemashō (トレーニングでお金を無駄にするのはやめましょう), and Gendaiban "Hito o Ugokasu" Rīdā (現代版「人を動かす」リーダー). Greg also publishes daily business insights on LinkedIn, Facebook, and Twitter, and hosts six weekly podcasts. On YouTube, he produces The Cutting Edge Japan Business Show, Japan Business Mastery, and Japan's Top Business Interviews, which are widely followed by executives seeking success strategies in Japan.

This week on Sake On Air, host Cindy Bissig is joined by special guest host Julian Houseman to welcome back a familiar voice: Tom Wilson, co-founder and head brewer of Kanpai, London's first sake brewery! They recorded their conversation at Julian's sake bar, Sake House, in Umeda, Osaka. Listeners may remember Tom from “Episode #73: Future of Sake with Les Larmes du Levant & Kanpai London”, where he joined us for a lively conversation alongside Grégoire Boeuf. Now, several years later, Tom returns to share exciting updates from the Kanpai camp. In this episode, Tom reflects on his recent collaboration with a sake brewery in Nara, giving us a peek into the inspiration and process behind this unique Japan-U.K. brew, which will soon be available in both countries. He also offers insight into Kanpai's evolving philosophy and what's been happening at their new home in Peckham, London, where they continue to experiment, grow, and celebrate all things sake. And for a special treat, Tom brought along a bottle of Kanpai's 2021 vintage “Kura” sake, which was enjoyed during the recording—and let's just say, it didn't disappoint. Tune in to hear how the international sake scene continues to evolve, and what it means to brew Japanese sake with British roots. Join us for a special English / Japanese bilingual rakugo performance: https://jss-event16.peatix.com/ Subscribe to our newsletter: https://sakeonair.substack.com/ We'll be back very soon with plenty more Sake On Air. Until then, kampai! Sake On Air is made possible with the generous support of the Japan Sake & Shochu Makers Association and is broadcast from the Japan Sake & Shochu Information Center in Tokyo. Sake on Air was created by Potts K Productions and is produced by Export Japan. Our theme, “Younger Today Than Tomorrow” was composed by forSomethingNew for Sake On Air.

GOOD TROUBLE—Troublemakers is a magazine about society's misfits. At least from the Japanese point of view. A bilingual, English/Japanese magazine, Troublemakers came about as a way to showcase people who were different, who stayed true to themselves, or about the long road those people had taken to self-acceptance.The founders, editor Yuto Miyamoto and art director Manami Inoue, were inspired by a notion that Japanese culture perhaps did not value those who strayed too far from the herd.The magazine has been a success not just in Japan but globally, and perhaps mirrors a trend we see in streaming, for example, of a general public acceptance of universal stories from different places—gengo nanté kinishee ni. Think, especially, of the success of Japanese television and movies like Shogun or Tokyo Vice or Godzilla Minus One. Of Japanese Pop and anime and food. It's an endless list.But Troublemakers is more than just a cultural document. It is proof of something shared, a commonality of human experience that exists everywhere. Speaking to Yuto and Manami, you sense a desire—and an invitation—to connect. With everyone. And that's, ultimately, what Troublemakers tries to do.—This episode is made possible by our friends at Freeport Press. A production of Magazeum LLC ©2021–2025

The 14 July entry deadline for the Citeline Japan Awards 2025 in Tokyo on 21 October is fast approaching! Join Ian Haydock and Lisa Takagi in this bilingual mini podcast as they outline the event and the various Award categories. (Japanese starts at 2:20.) More information on categories, entry criteria, table bookings and sponsorship opportunities here: English: https://www.citeline.com/en/awards/citelinejapanawards) Japanese: https://www.citeline.com/ja-jp/awards/citelinejapanawards)

Eric (his English name) grew up partly outside of Shanghai, China and then in Shanghai. He is now a Software Engineer that maneuvers through his Shanghai life in a multitude of languages including Mandarin Chinese, Shanghaiese, English, Japanese and French.Top 3 services that have help me while living in different countries: Affiliate links so I'll get a small commission.

Eric (his English name) grew up partly outside of Shanghai, China and then in Shanghai. He is now a Software Engineer that maneuvers through his Shanghai life in a multitude of languages including Mandarin Chinese, Shanghaiese, English, Japanese and French.Top 3 services that have help me while living in different countries: Affiliate links so I'll get a small commission.

Japanese news and cultural topics 1 source This audio data is a kind of breaking news program that covers all the news and event information in Japan. Topics range from art exhibitions, discovering paintings, movies, architecture, music, games, hot springs, festivals, fashion, and startup funding, and also includes detailed information such as specific locations, dates, times, and prices. The information presented is based on quotations from various media, and suggests ways for listeners to spend their weekends. Overall, it can be said to be a light-hearted information program that introduces a wide range of Japanese culture and trends. 日本のニュースと文化の話題

GOOD TROUBLE—Troublemakers is a magazine about society's misfits. At least from the Japanese point of view. A bilingual, English/Japanese magazine, Troublemakers came about as a way to showcase people who were different, who stayed true to themselves, or about the long road those people had taken to self-acceptance.The founders, Editor Yuto Miyamoto and art director Manami Inoue, were inspired by a notion that Japanese culture perhaps did not value those who strayed too far from the herd.The magazine has been a success not just in Japan but globally, and perhaps mirrors a trend we see in streaming, for example, of a general public acceptance of universal stories from different places—gengo nanté kinishee ni (language be damned). Think, especially, of the success of Japanese television and movies like Shogun or Tokyo Vice or Godzilla Minus One. Of Japanese Pop, and anime, and food. It's an endless list.But Troublemakers is more than just a cultural document. It is proof of something shared, a commonality of human experience that exists everywhere. Speaking to Yuto and Manami, you sense a desire—and an invitation—to connect. With everyone. And that's, ultimately, what Troublemakers tries to do. ©2024 The Full-Bleed Podcast is a production of Magazeum LLC. Visit magazeum.co for more information.

Godzilla goes to spaaaace! In a movie that's way more alien invasion space opera than kaiju, the hilariously-dressed Xiliens from Planet X introduce us to Monster Zero (aka Ghidorah) and get Earth wrapped in a convoluted plot involving trapping sleepy Godzilla and Rodan in bubbles. We talk the first American-Japanese co-production for Toho -- which includes a major role for Rebel Without a Cause's Nick Adams and a complex English/Japanese language shoot -- along with space lady clones, a backslide in character dynamics, and of course, Godzilla's famous little dance and its surprisingly important ramifications. ---- Help us nominate The Mixed Reviews for a Film & TV Podcast Award! Go to ⁠podcastawards.com⁠. ---- ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Patreon⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ |⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠Discord⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Part of⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠The Glitterjaw Queer Podcast Collective⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Email: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠skreeonkpodcast@gmail.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Theme song:⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠"BIO WARS - Synth Cover" by Kweer Kaiju⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Sources include: Ishirō Honda: A Life in Film, from Godzilla to Kurosawa by Steve Ryfle and Ed Godziszewski A Critical History and Filmography of Toho's Godzilla Series by David Kalat Godzilla FAQ: All That's Left to Know About the King of the Monsters by Brian Solomon Wikizilla

As the 15 July entry deadline for the 3rd edition of the annual Citeline Japan Awards approaches, join us in this bilingual English/Japanese podcast to learn more about the categories and event itself. (Japanese starts at 2' 25".) English event site: https://www.citeline.com/en/awards/citelinejapanawards Japanese event site: https://www.citeline.com/ja-jp/awards/citelinejapanawards

Have you wondered how to expand your business' reach across different languages and cultures? Do you know what role multilingual and localized content can play in your business? Naoko Takano, Localization and Community Program Manager for WordPress, joined me on the SEJShow to explore the significance of localization and internationalization in WordPress' mission.  Naoko has been involved with WordPress localization since its infancy and has seen firsthand how multilingual localization has built larger communities around businesses.  Learn the power of multilingual content and what it means for effectively broadening your reach online. Discover tips and opportunities for your business to collaborate across different cultures and how to leverage this power to improve your ROI.   WordPress, being an open-source solution from WordPress.org, your mission is to empower the publisher. –Loren Baker, 06:45 One of the advantages WordPress has is that we have so many different types of languages, and it's possible to add more languages as long as there are translators. –Naoko Takano, 13:49 The mission of WordPress is democratizing publishing. Our mission is to reach all the people using the internet and want to publish, not only in English-speaking countries. So yes, we want to expand beyond. –Naoko Takano, 23:28   [00:23] - About Naoko. [02:18] - WordPress localization and global growth. [06:45] - Open source as WordPress.org's growth driver. [08:49] - The role of translation in WordPress business expansion. [10:58] - Volunteer-driven operations at WordPress.org. [12:55] - WordPress plugins vs. enterprise CMS for expanding businesses. [13:49] - WordPress' multilingual capabilities. [14:55] - Anticipation of multilingual support in WordPress core software. [16:07] - Balancing WordPress' multilingual offerings with business interests. [16:54] - Gutenberg's fourth phase: Multilingual support? [17:07] - Adapting translation plugins for WordPress' block editor. [19:54] - WordPress' growth in the Spanish market. [21:52] - The most active countries in WordCamps events. [23:12] - The impact of translation on WordPress.org's international usage. [24:42] - Democratization through cultural collaboration in WordPress. [26:12] - Localization in WordPress: More than translation. [30:15] - The future of multilingual WordPress. [32:48] - The international WordPress community's size. [34:07] - Site translation and localization for various audiences. [36:24] - Connecting with the WordPress community online.   Resources Mentioned: https://wordpress.org/    I would like to see a more technology-assisted way for us to read and reach the audience in different parts of the world. That helps us find more engaging content and allows the content provider to reach further than the limitation of their language area. –Naoko Takano, 30:55 We benefit as a community by using the version we are familiar with and providing that to the client. The backend is also translated, which helps people make the plugins usable without the language barrier. –Naoko Takano, 08:49 We, the translation community, always say over 20,000 active people have been translating something within a year or so.–Naoko Takano, 32:52   For more content like this, subscribe to our YouTube channel: https://www.youtube.com/user/searchenginejournal Are you looking to keep up with current and effective digital marketing today? Check out https://www.searchenginejournal.com for everything you need to enhance your knowledge and skills in search marketing.   Connect With Naoko Takano: Naoko Takano is the Localization & Community Program Manager at Automattic with over ten years of experience in the US. She specializes in software localization and English-Japanese translation and applies her expertise to WordPress support, technical writing, and community organization.  A full-time contributor to the WordPress Polyglots and Community Teams for almost two decades, Naoko works as a Polyglots Global Mentor and General Translation Editor (GTE) for Japanese, aiding translation contributors and enhancing WordPress accessibility across different languages. Connect with Naoko on LinkedIn: https://www.linkedin.com/in/naoko/ Follow her on Twitter: https://twitter.com/naokomc   Connect with Loren Baker, Founder of Search Engine Journal: Follow him on Twitter: https://www.twitter.com/lorenbaker Connect with him on LinkedIn: https://www.linkedin.com/in/lorenbaker  

In today's episode we talk about a lot of new products being released like the Team Rocket apparel and various boxes. Also, the English and Japanese Pokemon 151 products are being scalped heavily, so beware of that. Check out this fun interview we did with Gary King Pokemon: https://www.youtube.com/watch?v=e4ek2b7-eEo&t=14s ---------------------------------------------------------------------------------------------------------------------------------------- News, Analysts, Breakdowns, and Personal thoughts, come join in the fun discussing it all with us. ---------------------------------------------------------------------------------------------------------------------------------------- Check us out on the following: Twitter: https://twitter.com/CBA_Podcast Instagram: https://www.instagram.com/cardboard_addicts_podcast ---------------------------------------------------------------------------------------------------------------------------------------- Zenturion XYZ's Content: https://linktr.ee/zenturionxyz Grumpy's Content: https://linktr.ee/agrumpycharizard Gonzo's Content: https://linktr.ee/026gonzo Ren's Content: https://linktr.ee/renscollectibles

今日はTwitterで仲良くさせていただいている、ヨーコさんをお迎えして、英語を教えること、日本語を教えることやこれからの目標についてお話ししています。

★Join my patreon to get transcripts and learn quick & easy native phrases★ ★You can read the Japanese subtitles on Podcast YouTube!★ ★I have another Podcast show, Life in Japan ★ いつもありがとう

This is English Japanese bilingual reading of John Irving's novel translated by Haruki Murakami.

This is English Japanese bilingual reading of Auggie Wren's Christmas story. Written by Paul Auster. Japanese translation is Haruki Murakami.

This is English Japanese bilingual reading of “Cathedral”. Japanese translation is Murakami's version.

The goal is to be in line with the vision and then the way in which you align with the vision that's where you get to creative.Brittany Arthur———We are excited to announce our EP60 featuring Brittany Arthur, Co-founder & Director of Design Thinking Japan & Business Karaoke Podcast.With Brittany, we dive into the connection between culture and innovation.We learn how design as facilitation can foster business innovation and growth and where Brittany saw people's life change to create problem-solving.Further, we dig into how design thinking can break company structure hierarchies and bridge cultural differences.She also shares how to measure the impact of design facilitation and how to leverage the digital layer on a hybrid world of digital-physical facilitation. During the episode we explore:How do you evaluate if a workshop or design thinking session was a success?How do you ensure your projects have an impact and not get "lost in the transition to implementation".How do you ensure facilitation has a long-term impact on your work/workshops.The design and innovation culture in Japan and how it differs to other regions in the world?How the right facilitation can bridges cultures and bring people together.Best practices in design facilitationHow to keep the alignment with the business vision and strategy, but still foster positive transformationHow to navigate design facilitation in a hybrid world of physical and digital sessionsThanks a lot for your time and for learning Brittany!———ABOUT THE GUESTBrittany Arthur supports companies in Japan to bring their business aspirations to life and to ignite creative confidence at the Japanese workplace by equipping people with tools for innovation.She specializes in innovation, Design Thinking and Design Sprints in Japan and in Japanese.She is the Co-founder Director of Design Thinking Japan & Business Karaoke Podcast, the only English-Japanese,

This is English Japanese bilingual reading of Winter Dreams. Japanese translation is Murakami's version.

This is English Japanese bilingual reading of The Ling Goodbye. Japanese translation is Murakami's version.

This is English Japanese bilingual reading of The Catcher in the Rye. Japanese translation is Murakami's version.

This is English Japanese bilingual reading of The Headless Hawk. Japanese part is translated by Murakami.

This is English Japanese bilingual reading of The Great Gatsby. Japanese translation is Murakami's version.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

Fukuoka is a city famous for beautiful women. Why do you think this is? In Nara, where Yoko san is from, are many beautiful women too? You can listen in English, Japanese and English/Japanese. 02:00 English, 04:30 Japanese, 08:23 English/Japanese. Transcription : https://www.lostinnihongo.com/ --- Send in a voice message: https://podcasters.spotify.com/pod/show/yoko67/message

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

ようこさんは、ともだちがいません（Yoko has no friends). Do you have any friends? Of course you do. Lucky you. Please treasure them. Find out how I cope with having no friends in Japan... You can listen in English, Japanese and English/Japanese. 01:10 English, 02:07 Japanese, 03:35 English/Japanese. Transcription : https://www.lostinnihongo.com/ --- Send in a voice message: https://podcasters.spotify.com/pod/show/yoko67/message

This is English Japanese bilingual reading of Haruki Murakami.

This is English Japanese bilingual reading of Haruki Murakami.

“モーニング mo-nin gu” is Japaglish for “morning special”. Yes, it's another weird Japanese phrase. Listen and find out what “モーニング mo-nin gu” is! You can listen in English, Japanese and English/Japanese. 00:39 English, 3:44 Japanese, 7:49 English/Japanese. Transcription : https://www.lostinnihongo.com/ --- Send in a voice message: https://podcasters.spotify.com/pod/show/yoko67/message

Do you like sweet stuff （あまいもの）? Find out what Japanese sweets I had in Muji Cafe the other day and what I couldn't have! You can listen in English, Japanese and English/Japanese. 00:40 English, 1:45 Japanese, 3:27 English/Japanese. Transcription : https://www.lostinnihongo.com/ --- Send in a voice message: https://podcasters.spotify.com/pod/show/yoko67/message

When Nina Cataldo moved back to Japan after finishing college she was surprised to find that even though she considered it her “motherland”, having grown up in Tokyo, she was seen as an outsider. This set her on the path to creating a community for women who identify as Hafu (mixed-race Japanese person) called “Hafu Women”. A proud Hafu herself,  Nina advocates and educates for recognition and the fair treatment of marginalized communities in Japan. Nina is someone making waves and following her passions, so I just knew I had to interview her to find out more about her ikigai. If you enjoyed this episode and it inspired you in some way, we'd love to hear about it and know your biggest takeaway. Come and find https://www.linkedin.com/in/jennifershinkai/ (Jennifer) and Nina on LinkedIn. In this episode you'll hear: How Nina's experience of being shy at a young age has helped her in her life and career as an adult What Nina did when she noticed a lack of support for Hafu Women like herself Nina shares her Brave and Bold Mastermind for Asian Women and how it helps entrepreneurs to accelerate their businesses The importance of Asian women being even more visible in spite of racial tension and discrimination Nina's challenges with her own energy when her work is also her passion About Nina Nina Cataldo is a DEI Facilitator and Multicultural Communications Specialist based in Tokyo, Japan. Originally from Tokyo, Nina grew up in the Pacific Northwest of the United States before returning to Japan at the age of 23. She is the founder of Hafu Ladies and co-founder of the Brave & Bold Mastermind program for Asian women entrepreneurs. In 2019, Nina quit her full-time job in order to sail around the world on Peace Boat for 4 months as an English-Japanese interpreter. Upon returning to Tokyo, she began her career in the fields she's most passionate about: communication and fostering human connections through the DEI lense.  Connect with Nina Website - http://www.ninamcataldo.com/ (www.ninamcataldo.com) Brave & Bold Mastermind - https://kristymariko.wixsite.com/bravebold2021 (https://kristymariko.wixsite.com/bravebold2021) Hafu Ladies - http://www.facebook.com/hafuladies (www.facebook.com/hafuladies) Connect with Jennifer Linked In: https://www.linkedin.com/in/jennifershinkai/ (https://www.linkedin.com/in/jennifershinkai/) Facebook: https://www.facebook.com/jennifershinkaicoach (https://www.facebook.com/jennifershinkaicoach) Website: https://jennifershinkai.com/ (https://jennifershinkai.com/ )

I talk about some tips on learning Japanese, explain about Japanese grammar and vocabulary. Apart from that, I talk about my personal life, what it's like to live in Japan. Enjoy!

Hey guys! It is Aiko with Schwagirl. I am an American English pronunciation coach. Welcome to my podcast "The Voice of English" Season 2. Season 2 focuses more on communication. In each episode, I bring a guest and he/she will give us tips to become a better communicator as a person who speaks English as a second language. In Episode 25, I invited Maki Takashima who is a British English accent coach based in Japan. Maki Takashima is a British English accent coach, diction coach and translator (English/Japanese, written and spoken). Maki was raised in Tokyo until the age of 15 when she moved to Dusseldorf, Germany, due to her father's work. She then attended a British school for three years where she acquired British English. She has two degrees; one from Sophia University (BA in German Language with Linguistics) and the other from Royal Holloway College, University of London (BS in Physics with Music). After her return to Japan in 1992, Maki worked in Classical Music and many other industries, where she always acted as a translator/interpreter on top of her main role in PR and marketing. In 2004 she became independent and started her own business of English language services; teaching and translating/interpreting. She was also a PR representative for Maestro Seiji Ozawa in the 2000's. In this capacity she dealt with national and international media and travelled extensively with Maestro's projects. Maki is currently based in Tokyo and is known as one of the very few Japanese who can speak and teach British accent, namely RP (Received Pronunciation). She runs her own English language school where she teaches individuals and groups. Her background of music education and singing experience has led her to teach diction to Japan's renowned professional singers, whilst she also enjoys teaching amateur singers and chorus groups. website: https://www.maki-takashima.com/ FB: https://www.facebook.com/maki.takashima blog: https://ameblo.jp/reginacecilia/ twitter: @makiuk Here is the IPA of the words we talked in the episode Box : West coast American (WA) [ɑ] / Received Pronunciation (RP) [ɒ] Talk : WA [ɑ] / RP [ɔː] Door : American (A) [ɔɹ] / RP [ɔː] Can I have a glass of water? : intonation goes up in A, down in RP Glass, have : A [æ] / RP [ɑː] *RP of glass sounds like "gloss" to American Water : WA [ɑ] / RP [ɔː] Cold/called, hole/hall : WA cold [oʊ] called [ɑ] / RP cold [əʊ] called [ɔ] "ou" sound : A [oʊ] / RP [əʊ] / in General British English [əʊ] (but it sounds slightly different from RP.) Enjoy the episode! If you have any questions regarding English learning or pronunciation, living in the US, working in the US, or if you would like to be a guest on this show, please contact me through http://schwagirl.com/contact You can download each episode from this link: http://thevoiceofenglish.libsyn.com/ Support me financially to be able to provide my podcast, youtube videos and FB live for free. Join my patrons on Patreon. You can donate from 1 dollar a month. https://www.patreon.com/schwagirl Don't forget to subscribe to my newsletter If you are interested in learning pronunciation from me, visit http://hatsuonkyosei.com/ for more information. Group courses and private coaching are available as well as online self-study materials.