As a Universal Hydrogen-branded plane, equipped with the largest hydrogen fuel cell ever to power an aircraft, made its maiden test flight in eastern Washington, co-founder and CEO Paul Eremenko declared the moment the dawn of a “new golden age of aviation.”

The 15-minute test flight of a modified Dash-8 aircraft was short, but it showed that hydrogen could be viable as a fuel for short-hop passenger aircraft. That is, if Universal Hydrogen — and others in the emerging world of hydrogen flight — can make the technical and regulatory progress needed to make it a mainstream product.

Dash-8s, a staple at regional airports, usually transport up to 50 passengers on short hops. The Dash-8 used in Thursday’s test flight from the Grant County International Airport in Moses Lake had decidedly different cargo. The Universal Hydrogen test plane, nicknamed Lightning McClean, had just two pilots, an engineer and a lot of tech onboard, including an electric motor and hydrogen fuel cell supplied by two other startups.

The stripped-down interior contained two racks of electronics and sensors, and two large hydrogen tanks with 30 kg of fuel. Beneath the plane’s right wing, an electric motor from magniX was being driven by the new hydrogen fuel cell from Plug Power. This system turns hydrogen into electricity and water — an emission-free powerplant that Eremenko believes represents the future of aviation.

The fuel cell operated throughout the flight, generating up to 800kW of power and producing nothing but water vapor and smiles on the faces of a crowd of Universal Hydrogen engineers and investors.

“We think it’s a pretty monumental accomplishment,” Eremenko said. “It keeps us on track to have probably the first certified hydrogen airplane in passenger service.”

Aviation currently contributes about 2.5% of global carbon emissions, and is forecast to grow by 4% annually.

Still using jet fuel

The test flight, which was a success, doesn’t mean entirely zero-carbon aviation is just around the corner.

Beneath the Dash-8’s other wing ran a standard Pratt and Whitney turboprop engine (notice the difference in the photo above), with about twice as much power as the fuel-cell side. That redundancy helped smooth a path with the FAA, which issued an experimental special airworthiness certificate for the Dash-8 tests in early February.

One of the test pilots, Michael Bockler, told TechCrunch that the aircraft “flew like a normal Dash-8, with just a slight yaw.” He noted that at one point, in level flight, the plane was flying almost entirely on fuel cell power, with the turboprop engine throttled down.

“Until both motors are driven by hydrogen, it’s still just a show,” said a senior engineer consulting to the sustainable aviation industry. “But I don’t want to scoff at it because we need these stepping stones to learn.”

Part of the problem with today’s fuel cells is that they can be tricky to cool. Jet engines run much hotter, but expel most of that heat through their exhausts. Because fuel cells use an electrochemical reaction rather than simply burning hydrogen, the waste heat has to be removed through a system of heat exchangers and vents.

ZeroAvia, another startup developing hydrogen fuel cells for aviation, crashed its first flying prototype in 2021 after turning off its fuel cell mid-air to allow it to cool, and was then unable to restart it. ZeroAvia has since taken to the air again with a hybrid hydrogen/fossil fuel set-up similar to Universal Hydrogen’s, although on a smaller twin-engine aircraft.

Mark Cousin, Universal Hydrogen’s CTO, told TechCrunch that its fuel cell could run all day without overheating, thanks to its large air ducts.

Another issue for fuel cell aircraft is storing the hydrogen needed to fly. Even in its densest, super-cooled liquid form, hydrogen contains only about a quarter the energy of a similar volume of jet fuel. Wing tanks are not large enough for any but the shortest flights, and so the fuel has to be stored within the fuselage. Today’s 15-minute flight used about 16kg of gaseous hydrogen — half the amount stored in two motorbike-sized tanks within the passenger compartment. Universal Hydrogen plans to convert its test aircraft to run on liquid hydrogen later this year.

Making modules

Eremenko co-founded Universal Hydrogen in 2020, and the company raised $20.5 million in a 2021 Series A funding round led by Playground Global. Funding to date is approaching $100 million, including investments from Airbus, General Electric, American Airlines, JetBlue and Toyota. The company is headquartered just up the road from SpaceX in Hawthorne, California, with an engineering facility in Toulouse, France.

Universal Hydrogen will now conduct further tests at Moses Lake. The company will work on additional software development, and eventually convert the plane to use liquid hydrogen. Early next year, the aircraft will likely be retired — with the fuel cell heading to the Smithsonian Air and Space Museum in Washington, DC.

Universal Hydrogen hopes to start shipping fuel cell conversion kits for regional aircraft like the Dash-8 as soon as 2025. The company already has nearly 250 retrofit orders valued at more than $1 billion from 16 customers, including Air New Zealand. John Thomas, CEO of Connect Airlines, which plans to be the first U.S. carrier to use Universal Hydrogen’s technology, said the “partnership provides the fastest path to zero-emissions operation for the global airline industry.”

Universal Hydrogen isn’t just producing the razors — it’s also selling the blades.

Almost all the hydrogen used today is produced at the point of consumption. That’s not only because hydrogen leaks easily and can damage traditional steel containers, but mainly because in its most useful form — a compact liquid — it has to be kept at just 20 degrees above absolute zero, usually requiring expensive refrigeration.

The liquid hydrogen used in the Moses Lake test came from a commercial “green hydrogen” gas supplier — meaning it was made using renewable energy. Only a tiny fraction of hydrogen produced today is made this way.

If the hydrogen economy is really going to make a dent in the climate crisis, green hydrogen will have to become a lot easier — and cheaper — to produce, store and transport.

Eremenko originally started Universal Hydrogen to design standardized hydrogen modules that could be hauled by standard semi-trucks and simply slotted into aircraft or other vehicles for immediate use. The current design can keep hydrogen liquid for up to 100 hours, and he has often likened them to the convenience of Nespresso units. Universal Hydrogen says it has over $2 billion in fuel service orders for the decade ahead.

Prototype modules were demonstrated in December, and the company hopes to break ground later this year on a 630,000-square-foot manufacturing facility for them in Albuquerque, New Mexico. That nearly $400 million project is contingent on the success of a previously unreported $200+ million U.S. Department of Energy loan application. Eremenko says the application has passed the first phase of due diligence within the DOE.

A long runway

Some experts are skeptical that hydrogen will ever make a meaningful dent in aviation’s emissions. Bernard van Dijk, an aviation scientist at the Hydrogen Science Coalition, appreciates the simplicity of Universal Hydrogen’s modules, but notes that even NASA has trouble controlling hydrogen leaks with its rockets. “You still have to connect the canisters to the aircraft. How is that all going to be safe? Because if it leaks and somebody lights a match, that is a recipe for disaster,” he says. “I think they’re also underestimating the whole certification process for a new hydrogen powertrain.”

Even when those obstacles are overcome, there is the problem of making enough green hydrogen using renewable electricity, at a price people will be prepared to play. “If you want to get all European flights on hydrogen, you’d need 89,000 large wind turbines to produce enough hydrogen,” says van Dijk. “They would cover an area about twice the size of the Netherlands.”

But Eremenko remains convinced that Universal Hydrogen and its partners can make it work, with the help of a $3 per kilogram subsidy for green hydrogen in Biden’s Inflation Reduction Act. “Of all the things that keep me awake at night,” he says, “the cost and availability of green hydrogens is not one of them.”

Universal Hydrogen takes to the air with the largest hydrogen fuel cell ever to fly by Kirsten Korosec originally published on TechCrunch

https://techcrunch.com/2023/03/02/universal-hydrogen-takes-to-the-air-with-the-largest-hydrogen-fuel-cell-ever-to-fly/

A government watchdog has found that the Secret Service and ICE’s Homeland Security Investigations (HSI) unit repeatedly failed to obtain the correct legal paperwork when carrying out invasive cell phone surveillance.

The findings were published last week by Homeland Security’s inspector general, tasked with oversight of the U.S. federal department and its many law enforcement units, which said that the agencies often used cell-site simulators without obtaining the appropriate search warrants.

Cell-site simulators — commonly known as “stingrays” — are surveillance gear used by law enforcement that impersonate cell towers to trick nearby cell phones into connecting to them, allowing police to track their real-time location. Some newer stingrays are believed to be capable of capturing the calls and SMS text messages of nearby phones.

But stingrays are controversial because they also ensnare every other device within their range, including devices owned by people with no connection to crime. Stingrays are also developed under strict nondisclosure agreements, which broadly restricts what is publicly known about stingrays, or even what police can disclose about them. Prosecutors have dropped court cases rather than risk revealing proprietary technical details about how cell-site simulators work.

Because of how invasive cell-site simulators are, the inspector general said that federal agencies must first obtain a search warrant, authorized by a judge, before a cell-site simulator can be used. The inspector general said only exigent or emergency circumstances allow for warrantless use of cell-site simulators, which can range from having to act quickly to prevent the destruction of evidence, through to an immediate risk or danger to life, a national security threat or a cyberattack. In those cases, the authorities have to apply for a court order within 48 hours of deploying the cell site simulator — or run the risk of falling foul of the law for carrying out illegal surveillance.

In its redacted report, the inspector general said that the Secret Service and ICE HSI “did not always obtain court orders” as required by their own agency’s policies or federal law.

The watchdog’s report described two sets of problems. The first is that the Secret Service and ICE HSI “did not correctly interpret” the internal policies governing the use of cell-site simulators in emergency situations. In one case, ICE HSI said it did not believe it needed a warrant because a party had “provided consent.”

The other problem was how the Secret Service and ICE HSI used cell-site simulators to support requests from local law enforcement agencies. In one case highlighted by the inspector general, a county judge “did not understand” why prosecutors sought an emergency surveillance order because, not understanding the statute, the judge “believed it to be unnecessary,” leading to a raft of warrantless deployments. In at least one other case, the inspector general rebuked ICE HSI as it was “unable to provide evidence” it ever applied for an emergency court order in one case deemed an exigent circumstance.

Both the Secret Service and ICE HSI accepted the watchdog’s six recommendations, which included shoring up its internal policies and procedures.

The redacted report did not reveal the number of times cell-site simulators were deployed in recent years. U.S. Immigration and Customs Enforcement, or ICE, which enforces immigration law and carries out deportations, is known to have used stingrays hundreds of times between 2017 and 2019.

In a blog post, digital rights group the Electronic Frontier Foundation criticized the report’s redactions. “The OIG should release this information to the public: knowing the aggregate totals would not harm any active investigation, but rather inform public debate over the agencies’ reliance on this invasive technology,” wrote EFF policy analyst Matthew Guariglia.

Secret Service and ICE conducted warrantless stingray surveillance, says watchdog by Zack Whittaker originally published on TechCrunch

https://techcrunch.com/2023/03/02/secret-service-ice-warrantless-stingray/

Scott Belsky, chief product officer and executive vice president for Adobe’s Creative Cloud, believes there’s a big difference between the hype cycle around web3 last year and what we’re seeing this year with ChatGPT and other generative AI models.

Belsky, who was interviewed by Forbes reporter Alex Konrad at the Upfront Summit in Los Angeles today, says web3 did not promise to reduce the time it takes to complete tasks, and generative AI most definitely does in his view.

“Web3 did not promise to reduce the workflow, the work that has to be done across any idea or action in the organization. And in fact, it added more friction and more work,” he said.

He believes that the real value of generative AI is speeding up tasks from hours or minutes to seconds, and that’s powerful. In fact, he sees it as being more like the value that collaborative products have brought to the enterprise — products like Figma perhaps, a collaborative design product that Adobe is in the midst of trying to acquire for $20 billion. The transaction is facing several regulatory roadblocks, which could explain why he didn’t mention it in the interview.

“I think it will be more akin to the sort of trend of collaborative products replacing every sort of function in an organization — you know, there’s a whole suite of startups that have actually been quite successful reimagining every function of the enterprise to be more collaborative and web-based, as opposed to like old clunky on-premise software,” he said. “I think that AI will do the same thing to reduce the workflow around all these job functions, and we’re starting to see a lot of examples of that, and I think we’re in the early days of that.”

Further, he believes that in the hands of creative individuals, generative AI could enhance their skills, rather than replacing them.

“If you ask any great creative what makes them great, it’s having more surface area for discovery and having more time to see more possible solutions so that they can have more choices of which path to follow,” Belsky said. “What an amazing opportunity for AI to actually suggest — instead of like a whole room of interns, here’s some amazing new possibilities.”

Belsky speculates that as AI becomes more deeply embedded in the creative process, there may be an audit trail built into the work’s metadata to help users determine what parts were created by AI and what role humans had in the work’s creation.

He says that it’s a bit early for enterprise users to trust it because of the need to understand that audit trail, as well as that proper permissions were given by the work’s original creator and any adjacent people such as the models used or other people involved in the content’s creation.

“A lot of our very big enterprise customers are very concerned about using generative AI without understanding how it was trained. They don’t see it as viable for commercial use in a similar way to using a stock image and making sure that if you’re going to use it in a campaign you better have the rights for it — and model releases and everything else. There’s that level of scrutiny and concern around the viability for commercial use,” he said.

Ultimately, Belsky thinks the stuff that wows us today probably won’t be where companies are attacking generative AI. Instead, they’ll be exploring much more practical business use cases that reduce manual work and speed up processes.

“Some of the use cases of generative AI that we see on social media [wow us], but actually the more practical ones that may end up really being a business opportunity are things that just enable content velocity and personalization.”

Adobe’s Scott Belsky talks generative AI — and why it’s not going to end up like web3 by Ron Miller originally published on TechCrunch

https://techcrunch.com/2023/03/02/adobes-scott-belsky-talks-generative-ai-and-why-its-not-going-to-end-up-like-web3/

The company has plenty of competitors; other marketing tools also let merchants showcase their reviews on their websites. To stay ahead of its rivals, Yoon says the company is about to launch a new service designed for B2C users called

According to a recent report, 92.4% of consumers use reviews when deciding to purchase products. It’s understandable. They want confidence in their purchase choice before buying a product online.

A Seoul-based startup called Indent wants to help these shoppers through access to video reviews and product ratings, while also helping to drive sales for e-commerce merchants by providing them with this same precious customer feedback. Its video review marketing tool is called VREVIEW, and it works by sending a chatbot to a merchant’s customers to collect their video reviews and product ratings. The merchants can then upload these customer-generated video reviews and product ratings to their websites to attract more potential shoppers and consumers.

Investors like Indent’s conversion rates, seemingly. The startup just raised $8.1 million (10.5 billion won) in a Series A led by SV Investment, with participation from strategic investor LG Uplus (a telco unit of LG), Korea Investment Partners and Crit Ventures, among others. The round brings the outfit’s total funding to date to $13.7 million and it will be used to continue developing its video platform. 

According to Indent CEO Morgan Yoon, Indent already has about 40 clients in the U.S., China and Japan and intends to enter the U.S. market in the second quarter of this year to attract more users overseas. Within those geographies, it works with 3,800 online merchants, mainly direct-to-consumer brands. In addition, South Korean corporations, including LG Uplus, Korean cosmetics brands maker AmorePacific and MLB Korea use Indent’s services, Yoon said. 

Indent claims it has more than 12 million registered users, accounting for about 60% of total online shoppers in South Korea. According to tests conducted by Indent, consumers using the video review platform showed, on average, a 6x higher order conversion rate.

The company has plenty of competitors; other marketing tools also let merchants showcase their reviews on their websites. To stay ahead of its rivals, Yoon says the company is about to launch a new service designed for B2C users called Spray, which the startup unveiled in beta version last April. Spray is a TikTok-like video review service that allows individual consumers to upload their customer-generated video reviews and share feedback, Yoon said, adding that spray users can get monetized if new shoppers purchase products via the Spray service.

In addition, Spray analyzes users’ shopping behavior and preferences and recommends to merchants (or influencers) which products fit best for them to sell, Yoon explained.

Indent, which has been gathering data on the actions of buyers, can recommend about 66 million products, Yoon added. 

The company currently generates revenue via VREVIEW subscription fees from its merchant clients. It also charges licensing fees from D2C brand merchants when they buy user-generated content (UGC) from their consumers.

The company employs 27 people in South Korea.

Image Credits: indent

Indent raises $8.1M funding for its AI-powered customer video review tool by Kate Park originally published on TechCrunch

https://techcrunch.com/2023/03/02/indent-raises-8-1m-funding-to-continue-developing-its-ai-powered-customer-video-review-tool/

Developing cutting-edge AI systems like ChatGPT requires massive technical resources, in part because they’re costly to develop and run. While several open source efforts have attempted to reverse-engineer proprietary, closed source systems created by commercial labs such as Alphabet’s DeepMind and OpenAI, they’ve often run into roadblocks — mainly due to a lack of capital and domain expertise.

Hoping to avoid this fate, one community research group, EleutherAI, is forming a nonprofit foundation. The organization today announced it’ll found a not-for-profit research institute, the EleutherAI Institute, funded by donations and grants from backers, including AI startups Hugging Face and Stability AI, former GitHub CEO Nat Friedman, Lambda Labs and Canva.

“Formalizing as an organization allows us to build a full time staff and engage in longer and more involved projects than would be feasible as a volunteer group,” Stella Biderman, an AI researcher at Booz Allen Hamilton who will co-run the EleutherAI Institute, told TechCrunch in an email interview. “In terms of a nonprofit specifically, I think it’s a no-brainer given our focus on research and the open source space.”

EleutherAI started several years ago as a grassroots collection of developers working to open source AI research. Its founding members — Connor Leahy, Leo Gao and Sid Black — wrote the code and collected the data needed to create a machine learning model close to OpenAI’s text-generating GPT-3, which at the time was getting a lot of press.

The company curated and open sourced The Pile, a collection of datasets designed to be used to train GPT-3-like models to complete text, write code and more. And it released several models under the Apache 2.0 license, including GPT-J and GPT-NeoX, language models that for a while fueled an entirely new wave of startups.

To train its models, EleutherAI relied mostly on the TPU Research Cloud, a Google Cloud program that supports projects with the expectation that the results will be shared publicly. CoreWeave, a U.S.-based cryptocurrency miner that provides cloud services for AI workloads, also supplied compute resources to EleutherAI in exchange for models its customers can use and serve.

EleutherAI grew quickly. Today, over 20 of the community’s regular contributors are working full-time, focusing mainly on research. And over the past 18 months, EleutherAI members have co-authored 28 academic papers, trained dozens of models and released ten codebases.

But the fickle nature of its cloud providers sometimes forced EleutherAI to scuttle its plans. Originally, the group had intended to release a model roughly the size of GPT-3 in terms of the number of parameters, but ended up shelving that roadmap for technical and funding reasons. (In AI, parameters are the parts of the model learned from historical training data and essentially define the skill of the model on a problem, such as generating text.)

In late 2022, EleutherAI became well-acquainted with Stability AI, the now-well-financed startup behind the image-generating AI system Stable Diffusion. Along with other collaborators, it helped to create the initial version of Stable Diffusion. And since then, Stability AI has donated a portion of compute from its AWS cluster for EleutherAI’s ongoing language model research.

After another big patron — Hugging Face — approached EleutherAI and nonprofit discussions kicked off, Biderman says. (Many EleutherAI staff were involved with the company’s BigScience effort, which sought to train and open source a model akin to GPT-3 over the course of a year.)

“EleutherAI has largely focused on large language models that are architecturally similar to ChatGPT in the past, and will likely continue to do so,” Biderman said. “Beyond training large language models, we are excited to devote more resources to ethics, interpretability and alignment work.”

One might wonder whether the involvement of commercially motivated ventures like Stability AI and Hugging Face — both of which are backed by substantial venture capital — might influence EleutherAI’s research. It’s a natural assumption — and it’s even evidence-backed. At least one study shows a direct correlation between donations and the likelihood that nonprofits speak up about a proposed government rule.

Biderman asserts that the EleutherAI Foundation will remain independent and says she doesn’t see a problem with the donor pool so far.

“We don’t develop models at the behest of commercial entities,” Biderman said. “If anything, I think that having a diverse sponsorship improves our independence. If we were fully funded by one tech company, that seems like a much bigger potential issue from our end.”

Another challenge the EleutherAI Foundation will have to overcome is ensuring its coffers don’t run dry. OpenAI is a cautionary tale; after being founded as a nonprofit in 2015, the company later transitioned to a “capped-profit” structure in order to fund its ongoing research.

Broadly speaking, nonprofit initiatives to fund AI research have been a mixed bag.

Among the success stories is the Allen Institute for AI (AI2), founded by the late Microsoft co-founder Paul Allen, which aims to achieve scientific breakthroughs in AI and machine learning. There’s also the Alan Turing Institute, the U.K.-based, government-funded national institute for data science and machine learning. Smaller promising efforts include AI startup Cohere’s Cohere For AI (despite its corporate ties) and Timnit Gebru’s Distributed AI Research, a global distributed research organization.

But for every AI2, there’s former Google chairman Eric Schmidt’s fund for AI research. Over $125 million in size, it attracted fresh controversy after Politico reported that Schmidt wields an unusually heavy sway over the White House Office of Science and Technology Policy.

Time will tell which direction the EleutherAI Foundation ultimately takes. Likely, the mission will evolve and change over time — in positive ways, we can only hope.

Stability AI, Hugging Face and Canva back new AI research nonprofit by Kyle Wiggers originally published on TechCrunch

https://techcrunch.com/2023/03/02/stability-ai-hugging-face-and-canva-back-new-ai-research-nonprofit/

I want to talk about weird tech for a minute. A lot of strange and wonderful pieces of technology have come across our desks over the years. More often that not, they’re the purview of the startup. Small companies are more nimble and willing to take risks. That’s a big part of the reason the entire smartphone industry converged on the smartphone factor. It’s also why a lot of laptops seem to look like MacBooks these days.

With shareholders and boards holding their feet to the fire, corporations are less willing to take big risks. Such ideas also tend to get gummed up in layers of bureaucratic red tape. It’s a bit like attempting an evasive maneuver with a giant ship. It’s difficult, ill-advised and can end very badly.

That’s why it’s always nice to see a company get a little weird with it. Try something new and not particularly practical. Earlier this week at MWC, I spent a bit of time with the Motorola Rizr. It’s that phone with the rolling screen that extends from five to 6.5 inches, depending on orientation, apps and other inputs. Will it ever come to market? Probably not. But Moto parent Lenovo is the king of mass-producing someone’s strange idea, from the foldable laptop it demonstrated at this year’s show to all of the weird and not especially practical deployments of e ink.

A concept device primarily serves to gauge consumer interest and demonstrate that a company is still innovating. Scaling for manufacturing and ensuring your device can stand up to reasonable consumer wear and tear is another thing entirely.

And, of course, there were those who had their doubts that foldable screens would ever come to market, and look at where we are now. Samsung claims to have sold north of 10 million devices, and this MWC saw even more players getting into the act. I recognize that MWC attendees are an extremely skewed and very small sample size, but it’s kind of wild seeing so many people using them in the wild in Barcelona this week.

Image Credits: Brian Heater

OnePlus offered their own concept device. The 11 Concept is intended as a gaming handset. As of yet, I’m not convinced there’s all that much market for gaming-first phones. Given boosts in processing power, it seems likely that consumers will be interested in mainstream flagships with gaming capability, rather than those entirely centered on the functionality.

That said, it was a fun reveal, tiny pumps moving cooling fluid up and down the device. That comes with its own light show, illuminating the fluid along its path. Early teases of the concept understandably drew comparison to Nothing’s first phone (the OnePlus connection certainly didn’t hurt), but ultimately it was clear that the 11 Concept represents a wholly different direction.

The Phone (1) does, however, shine a light on an interesting phenomenon in its own right. A light-up back isn’t some earth-shaking revelation, it’s an interesting design choice. But in a world that’s gone from breakthrough to stale and samey over the course of a decade, it doesn’t take a lot to shake things up.

Image Credits: Brian Heater

Before departing the Barcelona Fira for a final time this year, I made a point to check out Huawei Watch Buds. I’m a sucker for weird gadgets — and this is nothing if not that. Practical? No. Logical? Not really. Destined to be a huge hit? No way. Weird? Definitely yeah. It’s a contrast with the Watch D. The wearable has a band that tightens to give blood pressure readings. Again, it seems a ways off from mainstream adoption, but it’s easy to imagine blood pressure being the next major wave for smartwatch health, along with glucose measuring.

The Watch Buds, on the other hand, are just straight-up fun. I would never buy them and wouldn’t recommend others do, but the idea of a smartwatch face that flips up to reveal a charging case underneath is nothing if not fun. I appreciate anything that gives me one fewer piece of technology to carry around, but man is that a thick watch. If you really want to consolidate, something like the Power 1 AirPods charging iPhone case makes a lot more sense on the face of it.

The truth is that being a giant company is a double-edged sword when it comes to experimenting. On one side is the earlier giant ship analogy. On the other are tremendous resources that allow for such experimentation. Though, as we’ve seen with the recent round of mass layoffs, it’s precisely those projects that are often the first to get the ax.

An ode to big tech cos putting earbuds in smartwatches and other weird crap by Brian Heater originally published on TechCrunch

https://techcrunch.com/2023/03/02/an-ode-to-big-tech-cos-putting-earbuds-in-smartwatches-and-other-weird-crap/

Why, hello, Mr. Drone. Yes, you look very, very scary. Who’s a good drone? Who’s a good drone? Image Credits: Brinc

Drone developer Brinc today showed off its newest drone, designed to help cops and others who care about public safety do their jobs from a safe distance. The new quadcopter is called Lemur 2 and is tooled up and ready to go into areas that otherwise would be too dangerous for humans.

The new drone builds on the learnings from the previous version and adds a stack of new features. The new drone has increased autonomous abilities, with onboard lidar sensors that can help create 3D maps, and a handy standby mode where the drone just sort of hangs out, ready to take action when the drone operator wants to take over. The autonomous hover feature doesn’t need GPS systems to hang out and can evade obstacles and objects.

The drone relies on a broad sensor array, including cameras, night-vision, thermal imaging, lidar sensors, a spotlight, night-vision lights, and microphones and speakers that afford operators the ability to do two-way comms. The data is transmitted to a custom controller using heavy encryption, and the new drone can use mesh networking between drones, effectively extending the range of each drone. The company claims that this extends the product’s ability to operate over larger areas, indoors, and underground.

“Today marks the next step on BRINC’s journey to advance drone technology in the service of public safety. Our mission at Brinc is to revolutionize public safety by leveraging technology to de-escalate dangerous situations,” Blake Resnick, CEO of Brinc, said in a statement to TechCrunch. “Each drone deployed to a dangerous situation is one less individual in harm’s way, and a potential life saved. The LEMUR 2 is the next era of first response technology that will undoubtedly make law enforcement and emergency services in our country more efficient and safer for all involved.”

The company says it is heavily focused on manufacturing and sourcing within the U.S., with R&D and a lot of its manufacturing happening at its Seattle headquarters. This has given Brinc a competitive edge: It says it is National Defense Authorization Act (NDAA) compliant and approved by the U.S. government for its products to be used by federal agencies and contractors.

“Throughout our journey, we have worked with past and present law enforcement and emergency services professionals to understand their unique challenges and enhance their ability to do difficult jobs safely with best-in-class technology,” Resnick says. “We look forward to building upon our success and continually pushing the boundaries of what BRINC can offer to benefit public safety.”

The company produced a “Cops”-style dramatization video that, among other things, shows it smashing its way through windows and open doors to illustrate its vision for how it hopes law enforcement will take the new drone to heart:

Brinc’s Lemur 2 straps on blue strobes and is ready for action by Haje Jan Kamps originally published on TechCrunch

https://techcrunch.com/2023/03/02/brinc-lemur-2/

It is raining AI-powered features all across search engines. Today, Brave Search launched a new “Summarizer” feature, which is powered by different large langue models (LLMs) — OpenAI’s GPT tech isn’t one of them. Just like the name suggests, its job is to provide a synopsis of a search query using different sources.

The summary feature is available to all Brave Search users on desktop and mobile. In the examples given by the company, it can summarize results for queries like “are acetaminophen and ibuprofen the same” through medical resources or “what happened in East Palestine Ohio” through news links.

Besides the summary, the improved Brave Search will also highlight relevant sentences in listed results as news articles. Previously, it just highlighted search keywords from the page description.

Image Credits: Brave

“With 22 million queries per day, Brave Search is the fastest-growing search engine since Bing. Unlike AI chat tools which can provide fabricated responses, the Summarizer generates a plain-written summary at the top of the search results page, aggregating the latest sources on the Web and providing source attribution for transparency and accountability. This open system is available to all Brave Search users today to help them better navigate search results,” said Josep M. Pujol, chief of Search at Brave.

The company said that its LLM is trained to fight “unsubstantiated assertions,” referring to AI chat of other search engines like Bing going awry and sprouting out misinformation because of prompt engineering. Just like other offerings, Brave Search offers citations and links so that people can look at sources to double-check the information. So people have to decide for themselves if the mentioned sources are reliable — but there is a chance that people won’t even look at these links.

Brave warns users that they should not trust everything produced by AI-powered search results in its announcement. But it’s not clear if a similar warning will be shown in search results when people try out this feature.

“Given the current advancements in AI, it’s crucial to remind users that one should not believe everything an AI system produces, in much the same way one should not believe everything that is published on the Web. At the risk of stating the obvious, we should not suspend critical thinking for anything we consume, no matter how impressive the results of AI models can be,” the company said.

In the announcement, Brave noted that this new release won’t generate a summary for all queries. Currently, it is applicable for only 17% of the queries on the search engine, but the company expects this percentage to increase over time.

The company explained that the Summarizer feature is reliant on its own LLMs, instead of the popular GPT tech by OpenAI. It said that it uses a mix of three models: the first one is a question-answering model to get answers from text across pages; the second model is a classifier to weed out hate speech and spam; the final model rewrites the sentences to present a concise result.

Brave is competing with a lot of search engines going into AI-powered search. Last month, Bing made headlines with the announcement of a GPT-powered search where you can chat with a bot. In response, Google announced Bard in a closed beta. Relatively smaller players in the space like Neeva and You.com have also announced AI-aided search features. Given a lot of them are making errors in the early versions of these features, the search engine with the least glitches will have an advantage in AI-powered search.

Brave Search launches an AI-powered summarization feature by Ivan Mehta originally published on TechCrunch

https://techcrunch.com/2023/03/02/brave-search-launches-an-ai-powered-summarization-feature/

The best-known unicorns in the world are getting fit, showing that it is possible to trim losses while still posting growth. They are living experiments when it comes to corporations looking to get lean without cutting muscle.

European fintech giant Klarna is working through a valuation reset and a change in investor priority ahead of an eventual public offering. It’s not the only private tech company adjusting its valuation and working to scale its revenue and profitability to meet the new valuation reality the startup world is still digesting.

The Exchange explores startups, markets and money.

Read it every morning on TechCrunch+ or get The Exchange newsletter every Saturday.

Here in the United States, Instacart is undergoing a similar reforging. Much like Klarna, Instacart soared amid the pandemic, watching its valuation explode as it caught a business updraft during COVID-triggered economic upheaval. And, like Klarna, it has had to slash its valuation and clip staffing so that it can, at last, go public.

Earlier this week, we dug into Klarna’s 2022 results, paying special attention to its Q4 data; the company’s full-year results obfuscated the fact that Klarna made material progress toward profitability as the end of the year neared. That mattered more than its historical losses from earlier in the year.

Instacart’s Q4 results impressed. Are they good enough to push it toward an IPO? by Alex Wilhelm originally published on TechCrunch

https://techcrunch.com/2023/03/02/instacart-q4-results-ipo/

To get a roundup of TechCrunch’s biggest and most important crypto stories delivered to your inbox every Thursday at 12 p.m. PT, subscribe here.

Welcome back to Chain Reaction.

PSA: I’ll be at ETH Denver this week, so if you see me, say hi! I’ll have a few Chain Reaction pins on me and the first few people to find me will get one. Think of it like a free NFT, but instead of it going in your crypto wallet, it’ll go in your wallet, wallet. Woah!

Anyways, let’s get into the news; oh, and Happy March!

February wrapped up as a big month for the NFT market as non-fungible tokens on the Ethereum blockchain surpassed $1.5 billion in volume for the first time since May 2022.

NFT marketplace Blur hit an all-time high for monthly volume at $1.12 billion for February, making up 74.6% of the total volume across all Ethereum NFT marketplaces, according to data from The Block. (The Block’s data aggregation filters out wash trading — when traders buy and sell items between themselves to artificially raise volumes and prices.)

By comparison, OpenSea, now the second-largest Ethereum NFT marketplace, had $270.11 million in volume for February, the data showed. At its peak, OpenSea had about $4.8 billion in monthly volume in January 2022 but has since seen its overall transaction volume deflate.

Amid the recent rebound, Blur has bested the once-largest NFT marketplace OpenSea in monthly volume for the third month in a row as the crypto market debates the issue of NFT creator royalties.

“If you look at what’s been happening recently with OpenSea and Blur, obviously that’s a concern broadly speaking in terms of the market and royalty fights,” Yat Siu, chairman of Animoca Brands, said to TechCrunch. “However, the volume as a result of that has increased tremendously, which means it has brought back another kind of excitement into the space.”

More below.

This week in web3

Ethereum NFT marketplace passes $1B in volume for first time since May as the creator royalties war heats up (TC+)

As mentioned above, the NFT market is getting hot again and the rise of Blur in the NFT market has helped reignite a debate concerning royalties. In previous quarters, OpenSea tried to balance creator royalties as it held the top position for NFT marketplaces, but Blur’s aggressive stance is causing OpenSea to change its tack. But as massive NFT marketplaces drop fees, this could be a “slippery slope” that hurts creators in the long term, Siu said.

Does web3 need a venture bailout now that AI has all the hype? (TC+)

Shifting investor priorities, more expensive cash and a dearth of the large deals that were so common during the last startup boom could leave many late-stage web3 companies short on cash. And the clock is ticking. For startups stuck in a now passé category, watching venture dollars flow elsewhere cannot feel great, even if such evolutions in capital flows are normal.

Chainlink’s new platform lets web3 projects connect to Web 2.0 systems like AWS and Meta

Chainlink, a web3 services platform, is launching a self-service, serverless platform to help developers connect their decentralized applications (dApps) or smart contracts to any Web 2.0 API, the company exclusively told TechCrunch. This new platform also supports more widely used programming languages like JavaScript so that developers who are new to web3 can get into the space. It will also provide integrations to Amazon Web Services (AWS), Meta and others.

The latest pod

For last week’s episode, Jacquelyn interviewed Alex Adelman, the co-founder and CEO of Lolli. Founded in 2018, Lolli is a bitcoin rewards app that lets people earn bitcoin or cash back when they shop online or in-person at over 10,000 stores like McDonald’s, Starbucks, Dunkin’, CVS, Costco and so on.

Adelman previously was on the team that built a commerce gateway, Cosmic, that was acquired by PopSugar in 2015 then Ebates and Rakuten in 2017. And similar to Jacquelyn, Adelman also went to UNC-Chapel Hill — go Tar Heels!

Lolli has grown significantly over the past few years, from partnering with less than 1,000 stores to over 10,000 stores, to date. Adelman dived into the rewards system in the crypto ecosystem and how it has evolved over the years — and what the future holds for Lolli.

We also dived deep into the topic of Bitcoin NFTs and Ordinals, which is the latest craze for the community. We discussed whether Bitcoin NFTs are good for the ecosystem, how the technology can grow long term and possibilities for these digital inscriptions to potentially fit into Lolli’s business model.

Subscribe to Chain Reaction on Apple Podcasts, Spotify or your favorite pod platform to keep up with the latest episodes, and please leave us a review if you like what you hear!

Follow the money

1. China’s regulatory compliant blockchain Conflux raised $10 million in a private token sale
2. Decentralized crypto exchange Mangrove raised $7.4 million in a Series A round
3. Singapore-based digital asset exchange DigiFT raised $10.5 million in a pre-Series A round
4. Institutional DeFi-focused asset management platform Hashnote raised $5 million
5. Term Labs raised $2.5 million in a seed round to build safer crypto lending for institutions

This list was compiled with information from Messari as well as TechCrunch’s own reporting.

Are NFT marketplaces becoming an open sea for creator royalties? by Jacquelyn Melinek originally published on TechCrunch

https://techcrunch.com/2023/03/02/are-nft-marketplaces-becoming-an-open-sea-for-creator-royalties/