Arinna secures $4M in seed funding to address the space power challenge

Arinna secures $4M in seed funding to address the space power challenge

The aspirations of countries and wealthy individuals regarding space call for improved power sources, and a new startup established by two PhDs from Stanford might have the solution. 

Arinna, launched by CEO Koosha Nazif and CTO Alex Shearer, announced on Wednesday that it has secured a $4 million seed funding round to create ultrathin solar panels from a novel material developed during their doctoral studies. 

The funding round was spearheaded by Spacecadet Ventures, with involvement from Anorak Capital and Breakthrough Energy Foundation; the company chose not to disclose its valuation.

Arinna, named after the Hittite deity of the sun and pronounced similarly to arena, anticipates having its initial products tested in orbit by the end of this year. Following the successful qualification of their photovoltaics in space, the company aims to construct a facility capable of producing the material at megawatt scale by 2028.

“We are developing qualification panels to send to our initial clients that will illustrate that these two-dimensional photovoltaics possess the efficiency and resilience to withstand space conditions,” Shearer stated. “We plan to substantiate this at a larger scale over the coming year, while simultaneously refining the necessary processes to manufacture each layer of our photovoltaic cells in a roll-to-roll format.”

Arinna specializes in solar cells tailored for spacecraft. In the era before SpaceX, when most satellites were custom-made, spacecraft utilized expensive yet durable solar panels made from rare earth materials. With the advent of mass-produced satellites, lower-cost silicon panels are now being employed, albeit they deteriorate faster due to cosmic radiation.

Instead, Arinna’s innovation is grounded in a new material — transition metal dichalcogenides, or TMDs, which are atomically thin semiconductors developed only in recent decades. The ultrathin solar technology from Arinna enables extremely flexible cells that the company asserts are both less expensive and more resilient than traditional space solar panels.

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

“Over the years, much of the solar advancement has been about squeezing out marginal percentage gains on familiar, existing technologies,” Ben Gaddy, a materials scientist and senior director at Breakthrough Energy, remarked to TechCrunch. “This represents a completely distinct category of materials.”

A visualization of a space data center utilizing Arinna’s Solar Panels.

Nazif and Shearer crossed paths at Stanford while engaged in their doctoral studies. Nazif focused on materials that could be repurposed to fabricate photovoltaic cells akin to traditional semiconductors, while Shearer worked on methods for large-scale production of those cells. “Koosha was the visionary, and I’m the builder,” Shearer humorously noted.

The company anticipates its photovoltaics will be significantly more flexible than traditional panels and 32% more efficient. Furthermore, according to Shearer, Arinna’s technology will not necessitate protective coverings, can endure for 15 years in orbit, and can be produced in a matter of weeks.

These enhancements would be substantial advancements over current technology, provided the company successfully navigates its orbital testing campaign this year without unexpected challenges and can fulfill its mass production objectives. 

“From my experience with all the space companies we’ve backed, power poses a significant barrier, a bottleneck,” remarked Wiz Khuzai, a general partner at Spacecadet Ventures who led the funding round, to TechCrunch. “[Arinna] is set to unlock the next generation of power solutions in space.”

Harbinger's upcoming product will be hybrid emergency vehicles

Harbinger’s upcoming product will be hybrid emergency vehicles

Trucking startup Harbinger remains a relatively recent player in the market, yet the adaptability of its electric vehicle platform has enabled it to secure another client in a distinct sector. This time, Harbinger’s chassis will serve in emergency vehicles for the 70-year-old company Frazer.

The two firms disclosed on Wednesday that Frazer will manufacture ambulances based on the hybrid variant of Harbinger’s platform, as well as larger mobile healthcare units. Frazer will also engage with Harbinger’s newly launched energy storage division, introduced earlier this year in collaboration with Airstream.

This agreement exemplifies how companies like Harbinger are achieving success with electric and hybrid vehicles, despite challenges in the passenger vehicle market within the United States. Grounded, another Detroit-based startup, announced this week that it partnered with Colgate to create a small fleet of mobile dental care units.

The secret to Harbinger’s achievements lies in its versatile platform, according to co-founder and CEO John Harris in an exclusive conversation with TechCrunch. The straightforward truck chassis can be adjusted in length based on customer specifications, and Harbinger can integrate a range-extending combustion engine if needed. Although Harbinger has only been around for a few years, this singular platform now supports RVs (built with THOR Industries), FedEx delivery vans, a smaller box truck model, and ambulances, assisting the company in raising over $300 million to date.

“When you consider the step van and RV applications, we have three wheelbases, four different GVWR [gross vehicle weight ratings], and about four powertrain options, with four, five, [or] six battery packs, along with the hybrid across all of it. We maintain 99.5% parts commonality,” said Harris. “That’s the game changer.”

Frazer CEO Laura Griffin remarked to TechCrunch that transitioning to Harbinger’s hybrid powertrain — mainly electric but utilizing the gas engine to recharge the battery — was an obvious choice as it reduces her customers’ overall ownership costs and boosts their operational availability.

“We’re always on the lookout for innovations that can enhance the experience for our end users, typically municipalities, 911 services, and hospitals,” she stated. “They’re doing it in comparison to other medium-duty chassis, so it fulfills all of our requirements.”

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

Griffin indicated that Frazer will procure the battery-powered auxiliary power units from Harbinger, deploying them on both the new hybrid emergency vehicles and existing combustion models. These will substitute standard generators and enable first responders (or users of the mobile healthcare units) to power medical equipment in the field without draining a vehicle’s battery or combustion engine.

“In the rear of an emergency vehicle, like an ambulance, you can visualize there’s a significant amount of equipment, and all of the latest sophisticated tools tend to rely on power,” Griffin explained. “So we’re in search of abundant clean power sources that don’t have to be attached to the chassis.”

Harris anticipates this will develop into a lucrative business regardless of how many hybrid vehicles Frazer acquires, as the auxiliary power units remain valuable across various powertrains.

“It will lead to a quicker growth trajectory, since there are thousands of ambulances,” he stated. He is also exploring other sectors, particularly in Harbinger’s home state of California, where there are stricter regulations against gas generator usage.

“There’s considerable interest from individuals expressing that they don’t want a generator six feet from an operator for 12 hours a day; they’d prefer to save money with batteries and are eager to reduce emissions,” he noted.

Following its pivot, Y Combinator graduate Glimpse secures $35 million in a funding round spearheaded by a16z.

Following its pivot, Y Combinator graduate Glimpse secures $35 million in a funding round spearheaded by a16z.

Dispute-tracking fintech Glimpse revealed on Wednesday that it has secured a $35 million Series A funding round, spearheaded by Andreessen Horowitz, with contributions from 8VC and Y Combinator.

Founders Akash Raju, Anuj Mehta, and Kushal Negi, who attended Purdue together, initially focused on a startup that specialized in Airbnb product placements. This venture was launched in 2020, but by 2024, the team shifted to a completely new concept: Glimpse, a platform designed to assist retailers in automating their financial deduction procedures. 

Last year, it completed a $10 million funding round, led by 8VC following the business pivot, which at that time was referred to as a Series A round. Now, they are designating the recent $35 million as a Series A and reclassifying their previous Series A as a seed round. To date, the company has garnered $52 million, including funds raised prior to their pivot.

“We ultimately recognized that we were missing product-market fit and opted for a significant pivot,” Raju stated regarding the earlier, unsuccessful initiative. “During this journey, we gained insight into the back-office operations of brands and the complexities of retail sales, which ultimately inspired us to create Glimpse as it stands today.”

They connected with their lead a16z investor through a shared founder acquaintance. “We developed a robust partnership as we expanded the business. We’re truly thrilled to collaborate with them in this next phase of growth,” he added.

Deductions represent the amounts a retailer detracts from what they owe a brand when processing an invoice. This is a standard practice and generally operates as follows: a brand bills the retailer, and the retailer compensates the brand. If the amount paid is less than the billed total, a justification is provided, such as if the goods were found to be damaged. 

Some deductions are legitimate, while others are deemed invalid; these invalid deductions can be burdensome to track and manage behind the scenes. “These inaccuracies are surprisingly prevalent,” Raju noted as the company’s CEO, adding that “a brand might send inventory correctly but still encounter charges for a short shipment.” 

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

“Teams typically log into various retailer systems, gather scattered documents, assess line items, reconcile with internal records, and manage disputes from start to finish. This challenge arises due to fragmented, unstructured data and siloed workflows across diverse systems and teams,” he described how the process generally unfolds.

If the brand fails to settle every invalid deduction, it may result in “consistent revenue leakage,” he remarked. 

Glimpse asserts it streamlines this process by examining deductions, identifying invalid ones, and filing disputes, thus aiding businesses in reclaiming funds they might have overlooked or lost. The platform’s AI agents access a retailer’s portal, retrieve and centralize all essential documents, and then categorize each deduction, Raju explained. Subsequently, the AI agents validate each adjustment against internal data (such as supply chain records and marketing calendars) to ascertain which deductions are legitimate and which are not. 

The company reported that it collaborates with over 200 retail brands, including Suave and its lip balm product Chapstick. 

“When discrepancies are detected, Glimpse automatically initiates disputes, monitors the process, applies recovered funds, and synchronizes everything back to the brand’s ERP,” Raju stated, emphasizing that the product connects across multiple systems. Besides the core enterprise resource planning financial software, it integrates with promotional calendars and retail platforms, significantly shortening the process to just days, he noted.

Despite Glimpse’s automation, Raju mentioned that his company maintains human oversight, “primarily to ensure outcomes,” he explained, like “following up on disputes to facilitate resolution and cash recovery, as well as ensuring quality control on critical activities such as classification and data extraction.”

The system becomes more adept each time a deduction is handled, continuously optimizing its classification, validation, and resolution processes. “In the long run, this fosters a compounding data advantage, where each new integration and client enhances the system’s intelligence and overall efficacy throughout the network,” he claimed. 

Others are also addressing invalid deductions with software solutions, like Revya and Confido.

“Our ambition is to serve as the AI infrastructure for CPG and retail brands, and this funding will assist us in advancing that ambition,” he concluded. 

Iranians Do Not Have Missile Alert System, Thus Volunteers Develop Their Own Warning Map

Iranians Do Not Have Missile Alert System, Thus Volunteers Develop Their Own Warning Map

Since the beginning of Donald Trump’s confrontation with Iran more than three weeks ago, U.S. military forces have allegedly targeted upwards of 9,000 locations, creating an atmosphere of fear and uncertainty among Iranians in Tehran and across the country. With no governmental warning system in place and amidst Iran’s longest internet blackout, Iranians are confronted with a lack of information.

Even before the airstrikes by Israel and the U.S., the lack of a public emergency alert system and strict state-controlled digital censorship adversely affected millions. Following last year’s 12-day conflict between Israel and Iran, Iranian digital rights advocates launched ‘Mahsa Alert,’ an innovative platform that delivers push notifications regarding warnings of Israeli assaults, confirmed target areas, and offline mapping capabilities. While it does not serve as a substitute for a coordinated emergency service, this tool aids citizens in critical moments.

“There is no emergency alert system in Iran,” asserts Ahmad Ahmadian, CEO of Holistic Resilience, the U.S.-based organization supporting Mahsa Alert. Established last summer, the platform addresses a vital need by charting Iran’s landscape of repression and surveillance. Lightweight applications for Android and iOS have been developed for offline functionality, essential due to Iran’s internet restrictions. Updates are minimal; a recent one was only 60 kilobytes.

Mahsa Alert features overlays of verified “confirmed attacks” through videos or images provided via a Telegram bot or social networks. Alerts regarding evacuation zones, “danger areas,” and potential hazards to nuclear or military sites keep the public informed. Ahmadian notes that most confirmed attacks correspond with pre-identified map locations.

The platform also catalogs CCTV, government checkpoints, medical facilities, religious locations, protest sites, and more. Mahsa Alert’s visibility internationally has increased on social media, encouraging users to disseminate its findings, resulting in over 100,000 daily active users in a brief period. Roughly 335,000 individuals have utilized it this year, with 28% reportedly from within Iran, particularly during January’s crackdown on demonstrators.

A startup from a former Thiel fellow has just unveiled a drone that claims to be able to substitute police helicopters.

A startup from a former Thiel fellow has just unveiled a drone that claims to be able to substitute police helicopters.

As I converse with Blake Resnick, he navigates his drone startup’s latest office in Seattle—a spacious 50,000-square-foot facility that Resnick anticipates won’t be completely operational until later this year—possibly November. Nevertheless, the large (and currently somewhat vacant) space signifies the potential of a rapidly expanding company focused on dominating its specific sector.

The sector in focus is public safety and the startup at hand is Brinc, which markets drones to law enforcement and public agencies throughout the U.S. The company aims to be the “DJI of the West,” as Resnick describes it—a reference to the Chinese drone producer and an indication that Resnick aspires for Brinc to be equally associated with the technology it provides.

A former Thiel Fellow—a prestigious initiative that supports young entrepreneurs in bypassing or delaying college—Resnick established Brinc in 2017 and soon after attracted interest from then-OpenAI founder Sam Altman, who eventually became one of Brinc’s initial seed investors. Since that time, Brinc has participated in several funding rounds and, according to Resnick, was valued at nearly half a billion dollars during its latest round.

On Tuesday, Brinc introduced its latest creation, a public safety drone named Guardian that Resnick claims is “the closest thing to a police helicopter replacement that the drone industry has ever produced.” Brinc asserts it is the world’s “most capable 9-11 response drone” to date.

Guardian certainly boasts impressive specifications and features. The drone can achieve speeds of up to 60 mph and offers a flight duration of 62 minutes, according to its developer. It is also equipped with thermal imaging cameras and two additional 4K cameras—all featuring zoom functions. “Even from a considerable height, a police department could capture details like license plate information,” Resnick shares. Moreover, it is fitted with a spotlight and a loudspeaker that surpasses the volume of a police siren.

The drone’s landing station, which Brinc refers to as a “charging nest,” enables fully automated battery replacement and can be stocked with essential safety equipment such as defibrillators, flotation devices, and Narcan, all without human assistance.

Guardian also includes a Starlink panel integrated directly into its structure, marking it, as per Brinc, the first public safety drone with such a feature. Starlink, SpaceX’s satellite internet service, provides the drone with connectivity anywhere in the world. “Starlink has never been incorporated into a commercially available quadcopter before, so [it] grants this airframe limitless range globally,” Resnick explains.

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

Resnick evidently identifies public safety as a significant opportunity. “There are approximately 20,000 police departments in the United States, 30,000 fire departments, and 80,000 police and fire stations—and we believe the top half of that market will likely have a 911 response drone stationed in a recharging nest on the roof in the future,” he states. “It indeed seems like we are looking at a $6 billion to $8 billion market opportunity,” he adds, evaluating markets in both the U.S. and globally.

In this vein, Brinc recently collaborated with the National League of Cities on an initiative to expand “drone as first responder” programs in communities nationwide—a strategy that will undoubtedly help cultivate relationships between the startup and potential customer communities.

Moreover, Resnick believes that recent geopolitical shifts have benefitted his company. Until recently, DJI maintained an unofficial monopoly over the global drone market—including the U.S., where safety agencies have traditionally depended on the products of the Chinese firm. However, the Trump administration recently prohibited the entry of foreign-made drone models into the country, thereby unlocking a vast potential market.

“There exists a significant need for a DJI of the West, or a premier drone manufacturer for the free world, and ultimately, that is what we aspire to be,” Resnick states.

Amazon has recently acquired a startup that creates child-sized humanoid robots.

Amazon has recently acquired a startup that creates child-sized humanoid robots.

Amazon has announced its acquisition of Fauna Robotics, a startup launched two years ago by former engineers from Meta and Google who are creating child-sized humanoid robots for use at home.

Bloomberg was the first to report the acquisition. The specifics of the agreement were not shared. What is known is that Fauna’s team, including its co-founders, will be integrated into Amazon’s operations in New York City.

“We are thrilled about Fauna’s vision to create capable, safe, and enjoyable robots for all,” an Amazon representative stated in a written email. “With Amazon’s robotics knowledge and years of establishing customer trust in home environments through our retail and devices ventures, we are eager to explore innovative ways to enhance and simplify our customers’ lives.”

Earlier this year, Fauna commenced the shipment of its initial product, a 59-pound bipedal robot named Sprout, to select research and development collaborators.

This marks Amazon’s second robotics acquisition this month—at least known to us. Earlier in the month, Amazon confirmed to TechCrunch that it has also acquired Rivr, a Zurich-based startup recognized for its robot capable of climbing stairs for deliveries. The details of that agreement were not disclosed either.

With $3.5B in new funding, Kleiner Perkins is fully committing to AI

With $3.5B in new funding, Kleiner Perkins is fully committing to AI

Kleiner Perkins, the notable U.S. venture capital firm, declared on Tuesday that it has secured $3.5 billion in new funding across two distinct funds, a notable rise from the firm’s $2 billion fundraising effort less than two years prior.

Established in 1972, the firm reported raising $1 billion for its 22nd early-stage venture fund and $2.5 billion for a separate entity intended to finance late-stage growth companies.

This significantly larger capital acquisition was anticipated. In recent years, Kleiner Perkins has successfully obtained early investments in several rapidly expanding AI startups, including Together AI, Harvey, and OpenEvidence. The firm is also involved with Anthropic and SpaceX, two companies that are projected to go public this year.

In a climate where exits are rare, Kleiner Perkins achieved notable returns from last year’s IPO of Figma, a design software firm in which it led a $25 million Series B round in 2018. The firm also reportedly gained a good return when its portfolio company Windsurf was acqui-hired by Google last summer.   

Known for its historic early investments in Amazon and Google, Kleiner Perkins now operates with a streamlined team of just five partners. The firm has experienced some shifts in leadership recently: Ev Randle has moved to rival firm Benchmark, while Annie Case has shifted from a partnership role to advisory, a spokesperson for Kleiner Perkins confirmed.

Kleiner Perkins is part of a trend of significant fundraising by other VC firms. Thrive Capital recently obtained $10 billion in new commitments, while General Catalyst is reportedly aiming for a similar amount. Additionally, an SEC filing validates TechCrunch’s previous report that Founders Fund has completed a $6 billion closure for its fourth growth fund.

New Mexico has delivered Meta its initial legal loss regarding child safety, with the rest of the nation observing.

New Mexico has delivered Meta its initial legal loss regarding child safety, with the rest of the nation observing.

On Tuesday, a jury in Santa Fe mandated that Meta pay $375 million in civil fines after determining the company misrepresented the safety of its platforms and put children at risk.

The office of New Mexico attorney general Raúl Torrez described the ruling as a “pivotal moment for every parent worried about what may occur to their children when they are online,” based on a press release issued immediately after the verdict.

The ruling, which came after a trial lasting six weeks, found Meta culpable on both counts presented by the state under its Unfair Practices Act. At $5,000 for each violation — the highest permissible by law — the penalty might appear minor for a corporation valued at $1.5 trillion by public market investors. However, the monetary amount is less significant than the fact that this is the first jury verdict of its kind against Meta concerning harm to minors.

“Meta leaders were aware that their products harmed children, ignored alerts from their own staff, and misled the public about their knowledge,” Torrez stated after the verdict. “Today, the jury sided with families, educators, and child safety advocates in declaring that enough is enough.”

The lawsuit from New Mexico against the company stemmed from a 2023 undercover probe in which state investigators created fake accounts on Facebook and Instagram pretending to be users under 14 years old. These accounts received sexually explicit content and were solicited for sex by several men from New Mexico who were arrested in May 2024, with two captured in a motel where they believed they would meet a 12-year-old girl, as indicated by their interactions with the accounts.

The operation was fundamental to the state’s argument. The evidence it generated — in conjunction with internal Meta documents and testimonies from ex-employees — indicated that staff members and external child safety experts consistently expressed concerns about hazards on the platforms and were largely disregarded.

Some of the most damaging evidence emerged from individuals who worked within the company.

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

Arturo Béjar, who served as an engineering and product leader at Meta for six years starting in 2009, recounted to the court (after testifying before the Senate years ago) about his attempts to alert Meta officials after his own 14-year-old daughter encountered unwanted sexual proposals on Instagram. He also testified that the same personalized algorithms that make Meta’s platforms effective at targeting advertisements could also be beneficial to predators.

“The product excels at connecting individuals with similar interests,” Béjar stated, “and if your interest is young girls, it will effectively connect you with young girls.” 

Brian Boland, a former vice president of partnerships product marketing at Meta who dedicated nearly twelve years to the company, testified that when he departed in 2020, he “absolutely did not perceive safety as a priority” to CEO Mark Zuckerberg and then-COO Sheryl Sandberg.

Zuckerberg was deposed as part of the lawsuit, and a recording of that deposition, which occurred a year ago but was presented to jurors earlier this month, provided some of the trial’s more notable moments. Zuckerberg labeled research on whether the platforms are addictive as “inconclusive,” a remark that the state contested, pointing out that Meta’s researchers identified that several features were intentionally designed to trigger dopamine responses and prolong time spent on the apps. 

When questioned if, as a parent, he had the right to be informed whether a product his own child utilized was addictive, Zuckerberg noted that there was a lot to “unpack in that.” He then mentioned that he and his spouse personally assess whether products are “appropriate for use” before allowing their children to use them and that they “also monitor how they’re utilized.” He indicated that his children are “younger.”

Predictably, Meta announced plans to appeal. “We respectfully disagree with the verdict,” a spokesperson conveyed to media sources, asserting that the company “strives to ensure safety” on its platforms. 

The New Mexico lawsuit is far from Meta’s only legal trouble. Meta and YouTube are also featured in an ongoing trial in Los Angeles concerning allegations that their platforms are addictive and have caused harm to young users. 

A verdict in that second case may arrive soon. A jury is currently deliberating, in a case initiated by a plaintiff known only as K.G.M., a 20-year-old woman from California who asserts she became addicted to social media during childhood, resulting in anxiety, depression, and body-image issues. (TikTok and Snap were also included as defendants but settled prior to trial.) 

On Monday, the judge presiding over the Los Angeles case instructed jurors to continue deliberating after the panel indicated it was struggling to reach a verdict on one of the defendants — suggesting the potential for at least a partial retrial. 

Simultaneously, a second phase of the New Mexico case — a bench trial (meaning there is no jury) on public nuisance claims set to commence on May 4 — could lead to additional penalties, alongside court-ordered adjustments to Meta’s platforms, including age verification measures and enhanced protections for minors. 

Instead of asserting that Meta violated a specific consumer protection law, the state argues that the company’s platforms have broadly harmed the health and safety of residents in New Mexico.

Lululemon wagers that Epoch Biodesign can surpass its own limits, quite literally.

Lululemon wagers that Epoch Biodesign can surpass its own limits, quite literally.

As the world transitions to electricity, the oil and gas sector is relying on plastics to enhance future profits. However, Jacob Nathan is determined to change that narrative.

Nathan began exploring methods to decompose plastics during his high school years. Now, as the founder and CEO of Epoch Biodesign, he has developed a technique that utilizes an array of enzymes to “convert this synthetic waste” into a form suitable for creating more plastic, he mentioned to TechCrunch.

“For us, a bale of textile is the same as a barrel of oil,” Nathan stated, indicating that discarded fabric, rather than crude oil, serves as the foundational material for Epoch’s operations. Unlike oil, the cost of this raw material is not influenced by the unpredictable decisions of global leaders.

Epoch’s strategy focuses on decomposing both pre-consumer and post-consumer plastic waste into monomers — the essential components used to create plastic. The company depends on enzymes, the cellular machinery at a molecular level, to achieve this. However, due to the unpredictable nature of biology, Epoch employs only the enzymes and avoids using the microorganisms that produce them. To obtain these compounds, Epoch collaborates with industrial suppliers that already produce enzymes in large quantities.

Through a series of enzyme treatments, Epoch is able to reclaim over 90% of the target monomers. “The only remnants from our process are dyes, which are collected and can be handled separately,” Nathan explained.

The method is initially focused on nylon 6,6, a durable synthetic material prevalent in products ranging from apparel to airbags, carpets, and climbing ropes. 

“It’s the initial synthetic fiber. It’s what was developed by the innovators at DuPont. Its effectiveness ensures its continued use across multiple applications,” Nathan stated.

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

The timing is ideal, Nathan noted. “Recently, the prices of precursors for nylon 6,6 and other materials have surged as much as 150% based on spot pricing,” Nathan remarked. By utilizing waste textiles instead of petroleum, Epoch can completely avoid such fluctuations. “When we separate the material production from the processes of extraction, refinement, and the unpredictability tied to fossil carbon, we can establish far greater consistency.”

This proposal has struck a chord with investors, including the apparel powerhouse Lululemon, which generates significant amounts of clothing derived from plastics. Lululemon recently took part in a $12 million funding round that also featured Exantia, Happiness Capital, Kompas VC, and Leitmotif.

The fundraising will support the establishment of a demonstration-scale facility close to Imperial College London; the company aims to follow this with a commercial-scale facility expected to commence operations in 2028, which should have the capability to generate 20,000 metric tons annually of monomer.

Once fully operational, Nathan mentioned that Epoch may expand its efforts to recycle additional plastics. “The technology can be adapted for various materials and types of plastics,” he said. “Nylon 6,6 will achieve maturity first, but we have some thrilling developments in the pipeline.”

OpenAI’s Sora was the most eerie application on your device — now it’s closing down

OpenAI’s Sora was the most eerie application on your device — now it’s closing down

On Tuesday, OpenAI declared that it will be discontinuing Sora, a TikTok-like social application that debuted six months prior. No explanations were offered for the closure, nor was there any information provided on when it will be formally terminated.

Upon its launch as an invite-only social platform, Sora appeared to generate considerable demand for invitations. However, similar to Meta’s Horizon Worlds — a troubled virtual reality social platform that was once pivotal to the company’s notorious metaverse — Sora failed to maintain lasting appeal. The Sora 2 video and audio generation model is impressively advanced, yet interest in a solely AI-driven social feed was fleeting.

Sora was designed to operate as an AI-centric version of TikTok, replicating the familiar vertical video interface. Its key feature, “cameos,” enabled users to scan their faces and produce highly realistic deepfakes of themselves. These “cameos” could be public, allowing anyone to craft videos featuring them. (Cameo took legal action against OpenAI over the feature’s name and won, compelling the company to rename it to “characters.”)

In an outcome that surprised absolutely no one, this overhyped deepfake app turned out to be quite bizarre.

At its inception, Sora resembled an underregulated maze of unsettling Sam Altman videos. I was forever altered after viewing a lifelike clone of the OpenAI CEO meandering through a slaughterhouse of fattened pigs and inquiring, “Are my piggies enjoying their slop?”

[embedded content]

Sora was not intended to permit users to create videos of public individuals who had not explicitly opted in, yet bypassing OpenAI’s regulations proved to be remarkably simple. Soon enough, deepfakes of actual figures like civil rights leader Martin Luther King, Jr. and actor Robin Williams surfaced, prompting both of their daughters to take to Instagram, urging users to halt the creation of videos featuring their late fathers.

After producing numerous videos in which Sam Altman pilfered Nvidia chips from a Target, users changed tactics. Instead, they purposely generated content with copyrighted characters, courting legal issues for the individual they loved to deepfake — we witnessed Mario using cannabis, Naruto purchasing Krabby Patties, and Pikachu engaging in ASMR.

Techcrunch event

San Francisco, CA
|
October 13-15, 2026

This didn’t go as planned. Instead of suing, Disney, known for its litigious nature, invested $1 billion in OpenAI along with a licensing agreement that would have allowed Sora to produce videos with characters from Disney, Marvel, Pixar, and Star Wars.

This appeared to be a pivotal moment for the AI sector. Yet, with Sora’s shutdown, the arrangement is no more — although notably, it seems that no actual financial exchange occurred before its dissolution. (Disney offered some courteous comments regarding the situation on Tuesday, stating to the Hollywood Reporter that it will “continue to engage with AI platforms” moving forward.)

The initial excitement surrounding Sora was palpable. According to data from the mobile analytics firm Appfigures, the app reached approximately 3,332,200 downloads in November across the iOS App Store and Google Play. Had the app maintained its growth, OpenAI might have continued its operation, but that was not the case. By February, downloads plummeted to 1,128,700. This figure may seem substantial, but it pales in comparison to the 900 million weekly active users of ChatGPT.

During its existence, Appfigures estimates that Sora generated roughly $2.1 million from in-app purchases, which allowed users to acquire additional video generation credits. It’s difficult to believe that the computing demands of the Sora app had a significant impact on a company that is already incurring major losses, yet the app may have been too much of a risk to retain if it wasn’t experiencing growth.

When OpenAI launched the Sora app, I braced myself for a reality where we could easily create deepfakes of one another. Although I seldom create TikToks, I felt compelled to post a public service announcement that this alarming technology was rapidly approaching. It ultimately amassed over 300,000 views, which is unusual for my typically inactive TikTok account, but this announcement elicited a genuine reaction from people. I never anticipated that it would only endure for six months.

However, the disappearance of Sora doesn’t signify the end of the threat. The Sora 2 model remains accessible — it’s just secured behind the ChatGPT paywall. Moreover, OpenAI is far from the only entity making this technology widely available. It’s just a matter of time before another social AI video application enters the market, inundating us with another wave of clips featuring Snow White storming the Capitol.