Ask a true car or bike enthusiast to drive an EV and they’ll look at you with disdain and disgust. An EV’s quiet whirr doesn’t capture the rawness you get from the sound of a V6 engine roar, or the feeling of the gears turning as you hit the accelerator, or even the smell of gasoline, engine oil, et al. For coffee-lovers, this analogy applies to the art of coffee-brewing too. The idea of pressing a button and getting instant coffee just diminishes the beauty of a cup of golden brew. That’s where the COFFEEJACK V2 comes in, with its gloriously analog brewing method. Designed for a personal hands-on experience, this portable brewer relies entirely on manual brewing rather than automation. Armed with a rotating hand-crank, it looks like a portable coffee grinder, but the COFFEEJACK V2 actually uses that crank to build pressure (up to 10 bars of it). Rotate the crank and the COFFEEJACK V2 creates the right amount of push needed to force hot water through a portafilter filled with a puck of coffee, giving you luxuriously hand-brewed espresso complete with the crema. The best part? A compact, non-electric design that lets you literally brew your coffee anywhere, whether in the comfort of your kitchen or the idyllic wilderness of your campsite. Designer: COFFEEJACK Click Here to Buy Now: $214 $267.5 (20% off) Hurry, only 14 of 750 left! Raised over $347,400. Visually, the COFFEEJACK V2 embodies that rare intersection of industrial utility and considered aesthetics. Its cylindrical form speaks the language of serious outdoor equipment – reminiscent of high-end flashlights or premium binoculars – while the textured silicone grip bands provide both tactile feedback and visual contrast against the machined aluminum body, with inner components made from food-grade stainless steel. The knurled texture on the crank feels delightfully industrial, but the crown jewel of the V2 remains its exposed planetary gear mechanism, designed to evoke a sense of rawness seen under the hood of a vintage car or a Swiss watch. Remember that ‘joy of analog’ emotion I mentioned earlier? You won’t get that from a Nespresso pod – but the COFFEEJACK V2, ah, it feels divinely rugged. Ritualistic almost. What fascinates me about the COFFEEJACK V2 is how it rejects the automation trend dominating home coffee. This device demands engagement. You control the water temperature. You determine the grind. You generate the pressure. The result feels less like a convenience product and more like a specialized tool for the discerning coffee enthusiast who refuses to compromise, regardless of GPS coordinates. This isn’t COFFEEJACK’s first rodeo – their V1 tried to disrupt the Aeropress but came with its set of limitations, the V2 emerges in stainless steel and aluminum glory, absolutely redesigning the entire device altogether with beefed-up seals, superior thermal properties, and a new rotary input rather than the press-input of V1. The upgraded 54mm portafilter basket now accommodates a proper 18g dose, matching standard home espresso dimensions. The brilliance lies in its gear-driven pressure system. Through a series of interlocking cogs that multiply your hand strength (so it feels effortless as you use it), you can generate a legitimate 10 bars of pressure – the sweet spot for proper extraction. An analog pressure gauge provides real-time feedback, while coffee nerds can opt for the Bluetooth digital sensor to track their extraction curves. Crank the COFFEEJACK V2 and it consistently delivers pressure, passing the water through a professional-grade shower screen onto the large 54mm portafilter tamped with coffee. The process, although analog, replicates the exact procedure followed by a barista-grade espresso machine. The COFFEEJACK’s all-metal interior means no skimping or cost-cutting – the coffee is brewed the way it’s supposed to be. The result is an espresso you’d probably never believe came from simple elbow-grease and a tiny tumbler-sized device. Once you’re done savoring your coffee, cleaning up is easy too, as the dry coffee grounds pop out in the form of a solid dense puck, reducing any messy clean-up. At roughly thermos-sized but with the density of professional camera gear, the COFFEEJACK V2 occupies a practical middle ground in the packability spectrum. The insulated water chamber maintains temperature stability – crucial when you’re brewing at altitude where water boils at lower temperatures and heat dissipates quickly. The double-wall stainless steel construction ensures durability in the field, making the COFFEEJACK V2 about as adventure-friendly as you are. The market has responded enthusiastically: the Kickstarter campaign has already secured over £347,400 ($465,680 USD), demolishing its modest £7,000 ($9,282 USD) goal forty times over. Early backers can still claim one for £159 (approximately $210), below the planned £198 retail price. While that positions it as a premium product, it delivers capabilities previously unavailable at any price point. And if you still think you’re better off drinking that instant coffee or something from a Nespresso pod, you’re honestly beyond saving… Does it challenge your thousand-dollar coffee maker? Oh, absolutely – it’s designed to be as technically on-point as automated countertop machines – but with two distinct differences, first, the manual bespoke interaction, and secondly, unmatched portability that gives you the best coffee anywhere you want. Watching steam rise from a perfect crema-laden espresso while perched on some remote overlook might permanently change how you define luxury in the outdoors. The COFFEEJACK V2 reminds us that sometimes the most satisfying technologies aren’t those that remove human effort, but those that honor it in precisely the right ways. Click Here to Buy Now: $214 $267.5 (20% off) Hurry, only 14 of 750 left! Raised over $347,400.The post This Analog Espresso Maker is the Ultimate EDC Coffee Companion – Meet the COFFEEJACK V2 first appeared on Yanko Design.

Scientists have known for decades that certain neurons in the hippocampus are dedicated to remembering specific locations where an animal has been. More useful, though, is remembering where places are relative to each other, and it hasn’t been clear how those mental maps are formed. A study by MIT neuroscientist Matthew Wilson and colleagues sheds light on that question.  The researchers let mice explore mazes freely for about 30 minutes a day for several days. While the animals were wandering and while they were sleeping, the team monitored hundreds of neurons that they had engineered to flash when electrically active. Wilson’s lab has shown that animals essentially refine their memories by dreaming about their experiences. The recordings showed that the “place cells” were equally active for days. But activity in another group of cells, which were only weakly attuned to individual places, gradually changed so that it correlated not with locations, but with activity patterns among other neurons in the network. As this happened, an increasingly accurate cognitive map of the maze took shape. Sleep played a crucial role in this process: When mice explored a new maze twice with a siesta in between, the mental maps of those allowed to sleep during the break showed significant refinement, while those of mice that stayed awake did not.  “On day 1, the brain doesn’t represent the space very well,” says research scientist Wei Guo, the study’s lead author. “Neurons represent individual locations, but together they don’t form a map. But on day 5 they form a map. If you want a map, you need all these neurons to work together.”

Tiny flying robots could perform such useful tasks as pollinating crops inside multilevel warehouses, boosting yields while mitigating some of agriculture’s harmful impacts on the environment. The latest robo-bug from an MIT lab, inspired by the anatomy of the bee, comes closer to matching nature’s performance than ever before.  Led by Kevin Chen, an associate professor in the Department of Electrical Engineering and Computer Science and the senior author of a paper on the work, the team adapted an earlier flying robot composed of four identical two-winged units, combined into a rectangular device about the size of a microcassette. The wings managed to flap like an insect’s, but the bot couldn’t fly for long. One problem is that the wings would blow air into each other when flapping, reducing the lift forces they could generate. In the new design, each of the four units has a single flapping wing pointing away from the robot’s center, stabilizing the wings and boosting their lift forces. The researchers also improved the way the wings are connected to the actuators, or artificial muscles, that flap them. In previous designs, when the actuators’ movements reached the extremely high frequencies needed for flight, the devices often started buckling. That reduced the power and efficiency of the robot. Thanks in part to a new, longer wing hinge, the actuators now experience less mechanical strain and can apply more force, so the bots can fly faster, longer, and in more precise paths. The robots can precisely track a trajectory enough to spell M-I-T.COURTESY OF THE RESEARCHERS Weighing less than a paper clip, the new robotic insect can hover for more than 1,000 seconds—almost 17 minutes—without any degradation of flight precision. “When my student Yi-Hsuan Hsiao was performing that flight, he said it was the slowest 1,000 seconds he had spent in his entire life. The experiment was extremely nerve-racking,” Chen says. The new robot also reached an average speed of 35 centimeters per second, the fastest flight researchers have reported, and was able to perform body rolls and double flips. It can even precisely track a trajectory that spells M-I-T. “At the end of the day, we’ve shown flight that is 100 times longer than anyone else in the field has been able to do, so this is an extremely exciting result,” Chen says. COURTESY OF THE RESEARCHERS From here, he and his students want to see how far they can push this new design, with the goal of achieving flight for longer than 10,000 seconds. They also want to improve the precision of the robots so they could land in and take off from the center of a flower. In the long run, the researchers hope to install tiny batteries and sensors so the robots could fly and navigate outside the lab. The design has more room for those electronics now that they’ve halved the number of wings. The bots still can’t achieve the fine-tuned behavior of a real bee, Chen acknowledges. Still, he says, “with the improved lifespan and precision of this robot, we are getting closer to some very exciting applications, like assisted pollination.” 

The UK's Architects Registration Board (ARB) has advanced to the accreditation phase three new degree programs at the University of Leeds that, for the first time, offer qualifications that meet the body's new Competency Outcomes. They are said to represent the first alternative to the standing Parts 1-3 system and requirement for a minimum of two years of professional practice. Chair Alan Kershaw said he hopes their approval "represents real progress in our work to modernise architectural education and training, and shows how innovative learning providers can make the most of the flexibility opened up by ARB’s education reforms."  Our recent guide to the long road of licensure reform in the UK can be found here.

Researchers from MIT, Yale, McGill University and others found that adapting the Sequential Monte Carlo algorithm can make AI-generated code better.Read More

Training LLMs on trajectories of reasoning and tool use makes them superior at multi-step reasoning tasks.Read More

When Instagram was acquired for $1 billion in 2012, co-founder Kevin Systrom believed that joining Facebook would help Instagram’s “skyrocketing growth” reach even greater heights. In some ways, it did. Instagram now has billions of users and has since “generated many multiples of that price and then some,” Systrom said on Tuesday from a Washington, DC courtroom. But according to him, that success often came in spite of, not because of, Facebook’s help.While testifying in the Federal Trade Commission’s lawsuit to force the spin-off of Instagram and WhatsApp from Meta, Systrom said that CEO Mark Zuckerberg repeatedly withheld critical resources from Instagram and constrained its growth to avoid harming Facebook’s engagement. To the chagrin of Meta’s attorneys, Systrom also made predictions about how, in hindsight, Instagram would have probably still succeeded on its own.Over the course of about six hours, Systrom remained steady and confident on the witness stand. Zuckerberg himself sat in the same seat last week, describing how Instagram would likely not have become the social media powerhouse it is today without his help. In contrast, Systrom’s testimony portrayed Zuckerberg as a withholding and jealous boss. He described how he and Instagram’s other co-founder, Mike Krieger, quit in 2018 after growing increasingly frustrated with Zuckerberg’s meddling in Instagram’s operations. In court, Systrom was presented with an internal chart from that same year detailing the feature integrations Facebook had made with Instagram. With the help of features like notifications promoting Instagram within Facebook and cross-posting between the apps, Instagram experienced growth, while Facebook saw a neutral effect. Systrom said that, shortly before he and Krieger quit, Zuckerberg decided to end the feature integrations because, in Systrom’s view, he didn’t want Instagram to grow at the expense of Facebook. “We were a threat to their growth,” Systrom testified.“If Instagram didn’t grow as quickly, Facebook wouldn’t shrink as quickly, or plateau as quickly,” Systrom said in court. “I don’t think he [Zuckerberg] ever said it out loud that way, but that was the only reason we were having this discussion.”At the time, Instagram had just reached one billion users, which was about half of Facebook’s user base, with a fraction of the employees. Systrom felt that Zuckerberg was “underinvesting” in Instagram and giving it “zero resources,” which Systrom thought was “in stark contrast to the effort I was putting in.”According to Systrom’s telling, ego played a role. Zuckerberg was “very happy to have Instagram in the family,” he testified. “But also, I think as the founder of Facebook, he felt a lot of emotion around which one was better, meaning Instagram or Facebook, and I think there were real human emotional things going on.” “I think there were real human emotional things going on”Systrom recalled other instances where Instagram was denied the resources it needed. When Mark Zuckerberg declared that video would be the next big shift in social networking, Facebook started allocating internal resources towards the push. The company initially allocated 300 employees to making video a prominent part of Facebook, while Instagram received no additional headcount. Following the Cambridge Analytica data scandal that embroiled Facebook in controversy over its privacy practices, Systrom stated that his organization received “zero” of the billions of dollars in trust and safety resources that Zuckerberg had publicly committed to spending. Instead, he said Instagram was given access to a centralized team that was more focused on Facebook. He also described how, years earlier, Zuckerberg suddenly yanked members of the Facebook growth team who had been deployed to help Instagram. During cross-examination, Meta attorney Kevin Huff attempted to discredit Systrom’s testimony. He hardly gave an inch by maintaining that Instagram would have likely been successful as an independent company. “You deal in a world of probabilities,” he said. “You can never be sure. Some things you can be more sure of.”Huff’s questioning of Systrom got tense on several occasions. His stone-faced, one-liner responses prompted rounds of laughter in the courthouse media room, though Judge James Boasberg rarely cracked a smile. When Huff brought up an early email Systrom sent to Zuckerberg crediting an integration with Facebook for much of Instagram’s early growth, Systrom said he was only emphasizing the benefit to appease Zuckerberg. Huff then asked Systrom if he was lying to Zuckerberg in the email. Seemingly irritated, Systrom stared back and simply said, “Sir.”See More:

Solar cells are joined together during production at the SunSpark Technology Inc. manufacturing facility in Riverside, California, U.S., on Tuesday, April 3, 2018. | Photo: Getty Images Solar cells from four Southeast Asian countries that have been major suppliers to the US are facing newly increased tariffs, hiked up as high as 3,521 percent.  The tariffs affect Cambodia, Malaysia, Thailand, and Vietnam, which together accounted for more than three quarters of total module imports last year, according to Bloomberg. The tariffs essentially make the products “unmarketable” in the US, the Wall Street Journal reports. The tariffs essentially make the products “unmarketable” The move comes after long-running Commerce Department investigations into whether Chinese companies were funneling products through Southeast Asia to avoid tariffs and lower prices. It also heightens Donald Trump’s trade war with China after roiling global markets with drastic tariff proposals this month. The president called a 90-day pause on tariffs, excluding China.  US solar companies have been split on what to do about cheap solar cells from Southeast Asia. Domestic manufacturers petitioned the Commerce Department to investigate, while renewable energy project developers are worried about the tariffs raising costs for construction and the manufacturing of panels in the US using imported cells. Cambodia refused to comply with the investigation, and got hit with the highest tariffs at 3,521 percent. Duties for companies in Vietnam reach as high as 395.9 percent, 375.2 in Thailand, and 34.4 for Malaysia.  The US International Trade Commission still has to weigh in on the proposed tariffs in June to finalize them.

Diffusion Transformers have demonstrated outstanding performance in image generation tasks, surpassing traditional models, including GANs and autoregressive architectures. They operate by gradually adding noise to images during a forward diffusion process and then learning to reverse this process through denoising, which helps the model approximate the underlying data distribution. Unlike the commonly used UNet-based diffusion models, Diffusion Transformers apply the transformer architecture, which has proven effective after sufficient training. However, their training process is slow and computationally intensive. A key limitation lies in their architecture: during each denoising step, the model must balance encoding low-frequency semantic information while simultaneously decoding high-frequency details using the same modules—this creates an optimization conflict between the two tasks. To address the slow training and performance bottlenecks, recent work has focused on improving the efficiency of Diffusion Transformers through various strategies. These include utilizing optimized attention mechanisms, such as linear and sparse attention, to reduce computational costs, and introducing more effective sampling techniques, including log-normal resampling and loss reweighting, to stabilize the learning process. Additionally, methods like REPA, RCG, and DoD incorporate domain-specific inductive biases, while masked modeling enforces structured feature learning, boosting the model’s reasoning capabilities. Models like DiT, SiT, SD3, Lumina, and PixArt have extended the diffusion transformer framework to advanced areas such as text-to-image and text-to-video generation.  Researchers from Nanjing University and ByteDance Seed Vision introduce the Decoupled Diffusion Transformer (DDT), which separates the model into a dedicated condition encoder for semantic extraction and a velocity decoder for detailed generation. This decoupled design enables faster convergence and improved sample quality. On the ImageNet 256×256 and 512×512 benchmarks, their DDT-XL/2 model achieves state-of-the-art FID scores of 1.31 and 1.28, respectively, with up to 4× faster training. To further accelerate inference, they propose a statistical dynamic programming method that optimally shares encoder outputs across denoising steps with minimal impact on performance. The DDT introduces a condition encoder and a velocity decoder to handle low- and high-frequency components in image generation separately. The encoder extracts semantic features (zt) from noisy inputs, timesteps, and class labels, which are then used by the decoder to estimate the velocity field. To ensure consistency of zt across steps, representation alignment and decoder supervision are applied. During inference, a shared self-condition mechanism reduces computation by reusing zt at certain timesteps. A dynamic programming approach identifies the optimal timesteps for recomputing zt, minimizing performance loss while accelerating the sampling process. The researchers trained their models on 256×256 ImageNet using a batch size of 256 without gradient clipping or warm-up. Using VAE-ft-EMA and Euler sampling, they evaluated performance using FID, sFID, IS, Precision, and Recall. They built improved baselines with SwiGLU, RoPE, RMSNorm, and lognorm sampling. Their DDT models consistently outperformed prior baselines, particularly in larger sizes, and converged significantly faster than REPA. Further gains were achieved through encoder sharing strategies and careful tuning of the encoder-decoder ratio, resulting in state-of-the-art FID scores on both 256×256 and 512×512 ImageNet.In conclusion, the study presents the DDT, which addresses the optimization challenge in traditional diffusion transformers by separating semantic encoding and high-frequency decoding into distinct modules. By scaling encoder capacity relative to the decoder, DDT achieves notable performance gains, especially in larger models. The DDT-XL/2 model sets new benchmarks on ImageNet, achieving faster training convergence and lower FID scores for both 256×256 and 512×512 resolutions. Additionally, the decoupled design enables encoder sharing across denoising steps, significantly improving inference efficiency. A dynamic programming strategy further enhances this by determining optimal sharing points, maintaining image quality while reducing computational load. Check out the Paper. Also, don’t forget to follow us on Twitter and join our Telegram Channel and LinkedIn Group. Don’t Forget to join our 90k+ ML SubReddit. Sana HassanSana Hassan, a consulting intern at Marktechpost and dual-degree student at IIT Madras, is passionate about applying technology and AI to address real-world challenges. With a keen interest in solving practical problems, he brings a fresh perspective to the intersection of AI and real-life solutions.Sana Hassanhttps://www.marktechpost.com/author/sana-hassan/Meet VoltAgent: A TypeScript AI Framework for Building and Orchestrating Scalable AI AgentsSana Hassanhttps://www.marktechpost.com/author/sana-hassan/A Code Implementation of a Real‑Time In‑Memory Sensor Alert Pipeline in Google Colab with FastStream, RabbitMQ, TestRabbitBroker, PydanticSana Hassanhttps://www.marktechpost.com/author/sana-hassan/LLMs Still Struggle to Cite Medical Sources Reliably: Stanford Researchers Introduce SourceCheckup to Audit Factual Support in AI-Generated ResponsesSana Hassanhttps://www.marktechpost.com/author/sana-hassan/Stanford Researchers Propose FramePack: A Compression-based AI Framework to Tackle Drifting and Forgetting in Long-Sequence Video Generation Using Efficient Context Management and Sampling

Season 2 of Andor’s first of four chapters, episodes 1 through 3, picks up a year after the events of season 1 and starts sprinting. It’s a funnier-than-I-expected opening to season 2 as well. Like the first season did so well, Andor is still focusing on the small things. While the big picture of the rebellion is never far out of mind, this first chapter takes great pains to highlight the personal costs of taking on the Empire with great characters, some truly impressive filmmaking, and one of the best dance parties Star Wars has ever seen.There’s a scene almost right off the bat between Andor and an Imperial technician who’s helping him steal an experimental new TIE fighter. She’s nervous, scared, wondering if the risk she’s taking is even worth it, but Cassian is there to help talk her through it. He says all the right things, clearly drawing on experience he’s had in the year since we last saw him. It’s a perfect way to immediately show his growth.But the most important part of this conversation is when the tech talks about how she’d had an okay life at this facility. She had fun. It’s like she feels guilty about ever being happy as a part of the Empire and she’s not sure how to deal with it. This is a small conversation about a very small thing, but it says so much about this first chapter of Andor season 2, and it’s really quite something that they were able to do so much so efficiently in these opening moments. There’s a feeling throughout these first 3 episodes that any sense of normalcy or comfort is going away. The way the team behind Andor sets out to accomplish that is really savvy. This season is written and edited so well, and particularly in these first 3 episodes, Andor is built in such a way that very directly contrasts the realities of managing a rebellion with continuing to live in the Empire while you do it. It’s a fascinating little tight rope act, and one that the writers and directors almost flawlessly pull off. Even in the opening sequence, we see him very capably tending to this would-be rebel helping him, but immediately he’s back in over his head when he narrowly escapes in a ship he can barely fly.Ranking the Star Wars Disney+ Live-Action TV ShowsAnother great example of this arrives in episode 1. While Cassian is stuck on a forest planet with a group of rebels that don’t trust him and can’t even get on the same page with each other, we’re taken to a secret meeting of top Empire brass in a snowy, mountain top fortress. The way the episode jumps back and forth between these two scenes connects them in a really sneaky way. It’s edited quickly, almost like a montage, but these are two seemingly unrelated scenes being drawn together. On the one hand, you see the uphill battle that the rebellion is facing, with factions fighting for the same cause but killing each other because of distrust. This crew holding Andor hostage is just a bunch of selfish idiots – but, meanwhile, the Empire is quite casually plotting the destruction of an entire planet over coffee and canapes at a corporate retreat. These moments get lumped together as effectively one scene, and that one scene isn’t about what the rebellion or the Empire are up to separately. It’s a scene about how far apart the two sides are in their plans and how they get executed. While the Empire can get the ISB and the smear-campaign-pitching Ministry of Enlightenment douchebags marching in lockstep, the rebellion is literally starving in the mud, fighting each other while stranded on a planet full of beasts. They are not playing on the same level.What we said about season 2 before the premierePlay"Season 2 of Andor builds on nearly everything that worked so well about season 1, and continues fleshing out the prequel era of Star Wars. Ultimately the tale of the unsung heroes of the rebellion, Andor creates very personal stories at the heart of a much larger struggle. Tony Gilroy and company manage to weave the dramatic irony inherent in a prequel series into the storytelling itself, making Andor season 2 the most engaging the Star Wars franchise has been in a long time." – Clint GageRead the complete Andor season 2 spoiler-free reviewIt’s also a scene that proves just how well Andor is put together. Smart and efficient filmmaking like this pops up throughout this series to subtly drive home the themes while you’re not even looking.I will say though, episode 2 gets a little long where Cassian’s part of the plot is concerned. This band of misfits in the jungle are just too meatheaded to be interesting for as much time as we spend with them. They did their job very well in episode 1, but I don’t know that I needed the Star Wars version of rock, paper, scissors to play such a big role in Andor’s escape. It’s another vote in favor of the release schedule for season 2, though. I would’ve been a little more critical of episode 2 if I had to wait a whole week to see episode 3.Every Upcoming Star Wars Movie and TV ShowThankfully this first batch of episodes also features the galaxy’s most intriguing character, Mon Mothma. The fact that a major plot point of season 1 involved the senator agreeing to arrange the marriage of her daughter to a mobster’s son in order to finance the rebellion was one of the reasons I fell so hard for this show. Season 2 uses that wedding on Chandrila to do a couple of very cool things. It’s already incredible that Mon sending her daughter to marry into a shady financier’s family plays such an enormous role in the organization of this rebellion. But now her childhood friend and banking guru Tay has lost some money and gotten divorced, which in turn is causing him to make uncomfortable waves for Mon and Luthen Rael’s plans.Season 2 doubles down on butterfly-effect stuff in all the right ways.“Just to say that again, because it’s wild: A dude’s wife leaving him has enormous implications for the future of the rebellion. Forget teaching a kid in the desert how to use a light saber, this is the kind of fascinating butterfly-effect stuff that the second season of Andor is doubling down on in all the right ways. That something as tiny and personal as a marriage falling apart is a real threat to these early days of the rebellion is such a fun thread to pull on.More than Chandrillan divorce – and there’s always one guy who just got divorced at a wedding – the four-day extravaganza at the Mothma estate highlights that contrast that’s painted so well early in this season. Mon Mothma has always been obliged to play nice in the Senate and at home, while secretly funding the rebellion. Placing all of her anxiety around the rebellion, including the unexpected arrival of Stellan Skarsgaard’s Luthen Rael, against the backdrop of such a traditional event shows how determined the rest of her world is to carry on as though nothing is amiss and the Empire isn’t capable of blowing up a whole planet for a mineral.By the time the wedding is over, Tay’s been dealt with in the only way Luthen knows how to deal with loose ends, and Mon is doing shots – shots that she has very much earned, and dance-partying her way through some pretty boss EDM. Her tragedy is juxtaposed with a fresh and terrible loss for Cassian. Not only does Mon have to grin and bear it through an upper-class tradition, that contrast is used to invade Cassian’s life as well. It’s another sequence like cross-cutting the jungle meatheads and the Empire planning to strip-mine Ghorman. It creates a single story out of two otherwise unconnected threads. The real craft of Andor season 2 is in these moments, because they corral disparate parts of the rebellion into the whole. Mon Mothma and Cassian Andor have not met. Their only connection is that they both know Luthen Rael, but quite independently of each other. Here at the end of episode 3 though, they’re in the same place thanks to the absurdly clever editing of that house music.Everything Revealed at Hasbro's Star Wars Celebration 2025 PanelMeanwhile, on the antagonist’s side of the spectrum, the two main villains from season 1 are embroiled in a domestic chamber piece. Dedra Meero and Syril Karn squaring off against Syril’s mother, who, by the way, I believe to be the second or third best character in this entire show, is an incredible scene. It’s written with some of the most passive-aggressive dialogue humankind has ever seen. The three performances have at least a few layers of awkward subtext at play and, beyond any of that, we get more of what makes this first chapter great – the attempt at normalcy in the midst of rebellion. Taking a relationship to the “meet the parent” phase is a big and stupidly normal step for two ambitious, true-believer imperial agents like Dedra and Syril. There’s an incongruity to it that’s both really funny and really creepy. Seeing Syril more or less trade one mom for another is a wild place for this series to go, but at the same time, it’s exactly the kind of thing that made the first season so intriguing – and that season 2 is highlighting even more effectively.And while I was happy to see Bix again, her peaceful bit of respite harvesting grain with Brasso and Wilmon interrupted by an imperial audit is the most on-the-nose part of chapter 1. The officers are so blatantly slimy and their tactics so familiar, the storyline just didn’t have nearly as much to offer as the others. I certainly preferred the farmers on that planet to the meatheads in the jungle, but to be fair that’s a pretty low bar to clear.