Blog

Meet the Huawei P20. It’s a pretty nice phone. I played around with it, and I can confirm that it is, indeed, a solid flagship with some suitably over-the-top features — what’s that you say? Three rear-facing cameras?! But all of this is kind of a moot point if you live here in the States.

Sure, Huawei’s been having a lot of issues trying to sell its phones in America. In fact, just as I was playing around with the P20, news was breaking that Best Buy was planning to stop selling the company’s phones — it was a bit like finding out your starting pitcher needs Tommy John surgery before opening day. Only with, you know, lots more international espionage and such.

Rather than deal with the rigmarole of getting shot down by carriers and retailers this time out, the company is simply making it clear right off the bat that the new flagship just won’t be available here — not through any sort of official channels. And honestly, it’s probably best for Huawei to just focus on those countries that have long stocked its phones — from the look of the FCC reports earlier this week, this situation is going to get worse long before it gets any better.

For the rest of the world, there’s plenty to like here. The P20 looks like a cross between the iPhone X and HTC’s latest shiny metal phones. It’s got a 5.8-inch display (6.1 on the P20 Pro) and some crazy camera specs, including three rear lenses, including an eight-megapixel telephoto and 40-megapixel (!) RGP, coupled with a built-in color temperature sensor. There’s still a front-mounted fingerprint sensor and some strange choices, like a 2D face unlock function that makes do with the lack of depth sensing.

No pricing or availability at press time, except here in the States, where the latter is just a picture of a big red circle with a line through it.

Late last year, TPCast announced an adapter that cut the HTC Vive cords. The company is back with an enterprise version that delivers 2k content to several HTC Vive units with sub 2ms latency.

Unveiled at Nvidia’s GTC 2019 conference in San Jose, the product is aimed at VR uses cases where multiple people are sharing content across different VR headsets. The company says this includes medical, automotive, real estate, and training.

Like the consumer version, the units strap on the back of HTC Vive headsets and streams the content wirelessly from the connected PC to up to four VR headsets. At launch only the HTC Vive is supported though the company says it intends to support more models by the third quarter of 2018.

The unit, called the TPCAST Business Edition Wireless Adapter, will be available directly from TPCast at first and then available from retail channels. Pricing was not announced but it’s logical that it will cost significantly more than the $220 the company charges for the consumer version.

Marshall makes some of the best sounding and yet affordable headphones on the market and the company today is announcing its first set with active noise cancellation. Called the Marshall Mid ANC, the $269 wireless headphones feature the classic Marshall look but sport Bluetooth aptX wireless tech.

Four microphones measure ambient noise and block out unwanted noise. Sound comes from 40mm drivers which should sound just fine. Battery life with the noise cancellation clocks in at 20 hours (same as Bose’s latest) and 30 hours without.

The market is full of similar noise cancellation headphones. This set of cans from Marshall is interesting, though. This company provides products that carefully walk the line between being hipster without going overboard. Their products feature a retro, yet modern look and now with this set offers the latest technology. I’m in.

Massachusetts-based Soft Robotics has already shown us some pretty impressive capabilities for its rubbery robotic grippers, but the demos have mostly been straightforward picks off of a conveyor belt. The startup’s latest trick is the much sought-after activity of bin picking.

It’s a surprisingly tough task, given the complexity of picking up even several of the same object arranged at different angles. Most solutions require a fair amount of offline training and modeling in order to exhaust all the potential scenarios the arm might encounter when retrieving something from a bin.

The new SuperPick system, on the other hand, leans heavily on the company’s malleable, air-filled grippers, which are able to conform to and pick up a wide array of different shapes.

“It is designed to autonomously retrieve, sort and fulfill orders, with little to no human intervention,” Soft Robotics CEO Carl Vause told TechCrunch. “We do no training. We don’t do the offline learning, we don’t create the 3D models, because all of the things that are driving you to create those models, we can solve with the material science. We have people throw stuff in front of the robot and it can pick it up.”

The system is capable of more than 600 picks per hour, according to Soft Robotics. Like most robotic systems, SuperPick is designed to replace dull, repetitive warehouse tasks that often lead to a high turnover rate — and honestly, warehouse tasks don’t get much more monotonous than picking, sorting and placing objects from bins.

The system is available now. No word on pricing — that’s largely dependent on the job for which it’s being configured, but the company promises it will be available for a fraction of the cost of existing systems, courtesy, in part, to off-the-shelf components like webcams.

While pregnant, I imagined I’d spend maternity leave making elegant baby albums with detailed captions each month. That didn’t happen and now I have thousands of baby photos on my phone and no idea how to even begin organizing them. Created for parents who, like me, are overwhelmed by their digital memories, Precious uses artificial intelligence to help them wrangle and enjoy photos of their kids, automatically creating collages, slideshows, animations and time-lapse videos.

Founded by brothers Chris and Daniel Lau, Precious launched as Baby Art 16 months ago. The app quickly gained traction and has since acquired 54,000 paying users, up from 2,000 at launch (after a free seven-day trial, subscriptions cost $4.99 a month or $39.99 a year). Precious was part of Y Combinator’s latest batch of startups, which held its Demo Day last week.

Before launching Precious, the siblings spent several years working together on artificial intelligence-based products and saw the technology move from being available mainly for researchers to something that was in the cloud and accessible to consumers. While they were brainstorming ideas for a new app, Chris and his wife had a baby.

“Daniel saw me starting at my baby photos all the time and we felt we could make something really awesome for parents with AI,” Chris, Precious’ chief executive officer, told TechCrunch in an interview. The brothers started making trips to Babies R Us to chat with other moms and dads about what they needed.

“I think one of the coolest parts was when we were walking around and there was a dad who said I take 200 photos of my kid every day,” says Daniel, the startup’s chief technology officer. “He showed me his camera roll and it was a complete mess.”

While snapping 200 photos a day of your kid might seem a bit extreme, lots of parents have phones and cloud storage accounts crammed with pictures that aren’t the greatest, aesthetically speaking, but that they can’t bear to delete because their child will never be that little again. But then it becomes overwhelming to find specific photos.

“We drown ourselves because we can’t let them go, but it makes them impossible to use later,” says Chris, adding that one of the things mentioned by the parents he and Daniel spoke to was having “this lurking feeling, like we wish we could do more with the photos, that usually manifests when they get a coupon for a photo book. But all this joy you can get out of these photos and videos is why we take them.”

There are already a ton of baby photo apps out there, each with their own angle. These include Lifecake, which is billed as a “visual timeline,” apps like BlinkPix that make it easier to print and mail photos, and Tinybeans, a photo-sharing app that’s popular among parents who want to keep their kids’ pictures off social media. What makes Precious unique is that it uses artificial intelligence to help parents manage their photo libraries.

The app relies on facial detection to identify photos of kids and determines if photos are noisy, blurry or poorly lit to find the best pictures. For families with more than one child, Precious also considers heuristics like age and gender to differentiate between siblings. Once it decides which photos to use, the app automatically creates photo collages and albums each day. Precious also works like Timehop, Facebook Memories or Momento for baby photos, resurfacing older photos and creating milestone albums.

One of Precious’ coolest features is an automatic time lapse creator that automatically finds and aligns photos of your child’s face so you can watch them grow up in a short video. I’ve never attempted to organize the thousands of photos of my daughter on my phone, so it took Precious several days to sort through all of them and build the time lapse, but I enjoyed watching its progress every evening (and also tweaking it by removing or adding photos). Once it was done, I sent the video, which ended up being about 45 seconds long, to my family, who were all thrilled with it even though I already blast them with multiple kid photos every day. I’m not a very sentimental person, but watching my daughter morph from a newborn blob into an impish, expressive toddler felt magical and made me a bit teary. It also convinced me to subscribe.

“Finding good photos is a solved problem with AI. It’s not hard to do, but the trick now is finding photos that are meaningful,” says Chris.

“Many moms and dads aren’t really techies per say, so saying it is AI-based doesn’t mean they are really going to respond to that. It’s meaningless to them,” he adds. “So we show them things that they will get like the time lapse and tell them that it’s easy. That’s the differentiator for them.”

Other plans include taking advantage of artificial neural networks in newer smartphone software and moving more of Precious’ scanning technology onto phones, instead of sending photos to the cloud for processing, so the app will work faster and use less battery power. The Lau brothers want to make accounts sharable so photos from moms, dads, grandparents and caretakers can be gathered and organized, and add a printing feature that lets users create and order physical photo albums from the app.

Precious, which is currently self-funded aside from the financing it received as part of Y Combinator’s program, will also introduce a free version with premium services, but the startup will continue to rely on its subscription model instead of showing targeted ads or selling data to third parties. One of the reasons for this is privacy, which is especially important for photos of children. The Lau brothers want to add photo storage options eventually but say “we aren’t keeping photos until we can do it right.”

“I’m a parent and obviously I don’t want my kids’ stuff ending up on the Internet outside of my control. That’s another reason we wanted to build this because it’s a key hurdle that keeps parents from sharing photos more easily. Our business is not to sell you ads. We want to be paid for providing a good service and that’s it,” says Chris, adding “you are not the product, our product is the product and we’ve proven that works here. That’s why we are excited to make something that says what it will do and keep things private, but is still easy and convenient.”

For years, the goal of marketers was to understand the customer so well, they could respond to their every need, while creating content specifically geared to their wishes. Adobe Cloud Platform has long acted as a vehicle to collect and understand customer data inside the Adobe toolset, but today Adobe took that a step further.

The company hopes to transform Adobe Cloud Platform into a company’s experience record keeping system, a central place to collect all the data you may have about a customer from both the Adobe Cloud Platform and external data sources.

Suresh Vittal, vice president of platform and product at Adobe Experience Cloud says tools like CRM were intended to provide a record keeping system for the times, and they were fine in a period when entering and retrieving data was state of the art, but he thinks there needs to be something more.

“A lot of investments for past generations of software evolution have been around batch-based operational systems. While they were necessary back then, they are not sufficient where these brands are going today,” he told TechCrunch.

Adobe Systems world headquarters in San Jose, California USA Photo: Getty Images Lisa Werner / Contributor

Over time, as companies gather more and data, Adobe believes they need something that centers around the dynamic interactions brands are having with customers. “We believe every customer needs an experience system of record, a central [place to record] where the brand brings together experience data, content and a unified profile to power the next generation of experience,” he said.

To achieve this goal, the company is doing more than creating a new construct, it has built a new data model along with tools tools for data scientists to build custom data models.

Of course where there is data, there needs to be some machine learning and artificial intelligence to help process it, especially in a case where the goal is to pull disparate data into a central record. Adobe’s particular flavor of AI is called Sensei and the company is giving developers access to the some of the same AI algorithms it uses in-house to build its platform.

Any time you start pulling data together from a variety of sources to create a central record keeping system about a customer, there are huge privacy implications, and even more so with GDPR coming on line at the beginning of May in the EU. Vittal says the company has built in a governance and compliance layer into the toolset to help companies comply with various regulations around sharing data.

“You cannot turn all of this data into something useful without safeguards— semantics and control.” He says this involves creating a data catalogue, labeling data in the record and associating rules with each type. That way, data emanating from the EU will need to be handled a certain way, just as any personally identifiable information needs to be safeguarded.

This is where the machine learning comes in. “When you create data across the experience system of record, the data catalog recognizes [certain types of] data and and recommends labels based on types of data using machine learning.”

All of this is very likely an attempt to compete with Salesforce, which provides sales, marketing and customer service stitched together with their own artificial intelligence layer, Einstein. The recent $6.5 billion MuleSoft purchase will also help in terms of pulling data of disparate enterprise systems and into the various Salesforce tools.

The tools and services announced today give Adobe a fully intelligent, machine learning-driven solution of their own. The whole notion of a customer experience record, while a bit of marketing speak, also serves to help differentiate Adobe from the pack.

When it comes to sales software, Salesforce isn’t the only game in town.

A number of startups have emerged to target the sales industry, with roughly $3 billion per year getting invested into marketing startups. Now one of the largest players is Intercom.

The seven-year-old startup is now announcing a $125 million Series D round led by Kleiner Perkins, with participation from Google Ventures. Veteran investor Mary Meeker will also be joining the board.

The valuation is $1.275 billion, meaning that the company is considered a “unicorn.” The round brings the total amount raised to $241 million. Existing investors include Bessemer Venture Partners, Index Ventures, Jack Dorsey and Mark Zuckerberg.

Intercom is the “next generation customer database that’s specifically built for internet businesses,” co-founder and CEO Eoghan McCabe said in an interview with TechCrunch. “Salesforce is not built for internet businesses.”

“When a business uses the Intercom Messenger, they see conversation rates increase,” McCabe claims. Intercom says that customer sales go up an average of 82%.

The startup touts its “messaging-first approach,” helping its clients with customer acquisition and support. It says it powers 500 million conversations per month across its 25,000 customers. Atlassian, New Relic and Shopify use Intercom.

Intercom says it will use the funding to further develop its customer platform. It plans to double its research and development teams. Says McCabe, Intercom is “putting our foot on the pedal.”

Intercom, which has a strong presence in both the United States and Ireland, hopes to use the capital to expand its workforce beyond 500 employees.

It also would like to further expand its geographic footprint, adding customers across Europe and Asia.

Dandelion, a clean energy startup that was originally incubated inside Google parent Alphabet, has raised $4.5 million in funding to build out its business — a geothermal heating and cooling system for homes that claims it will drastically reduce its customers’ bills — it claims to cut bills in half (notwithstanding the upfront costs, more on that below) — while also being significantly more friendly for the environment compared to conventional systems that use gas and fossil fuels.

The company opened for business first in Upstate New York — a market with extreme cold and hot spells — where it says it has started to install systems in people’s homes, and it’s going to use the funding to help work through what it says is a waitlist of “thousands” of customers nationally.

“We have been overwhelmed with demand and support from homeowners across the country,” said Kathy Hannun, cofounder and CEO of Dandelion, in a statement. “This round will help us ramp up operations to serve these customers and launch our new and improved 2018 offering.”

The funding was led by New Enterprise Associates, with participation also from new investors BoxGroup, Daniel Yates, and Ground Up, and previous investors Borealis Ventures, Collaborative Fund, and ZhenFund, the Chinese-based VC associated with Sequoia in China. It brings the total raised by Dandelion to $6.5 million, including a seed round it announced when first spinning out in July of last year.

The impressive list of backers — and the fact that Dandelion was originally incubated at Alphabet X, the company’s “moonshot” factory — underscores a couple of trends worth pointing out.

The first is the double maxim that everything is now a “tech” challenge, and that the interests of the tech world touch everything. As legacy businesses continue to try to update their systems or become more responsive to some of the challenges of running their legacy operations, a company like Dandelion becomes a direct threat, or a potential, strategic acquisition target.

The second is the ongoing interest among tech investors and tech companies to expand their horizons and explore companies and ideas that might prove to be disruptive in the same way that tech has been, further down the line; or whose solutions could prove to be a helpful boost to their more direct tech interests — for example by becoming acquirers of data systems to run these services better, or by making the cost of electricity to run other services (like internet, or maybe, these days, bitcoin mining) less expensive. (Dandelion is already proving its role in that wider ecosystem: just earlier this month, it acquired Geo-Connections, a geothermal SaaS startup.)

“Over the next decade, homeowners across America will replace their expensive, conventional home heating systems with Dandelion geothermal,” said Yates in a statement. “I’m thrilled to be part of the team that will lead this transition.” Yates — who had founded the energy efficiency startup Opower, which went public and then was acquired by Oracle — is joining the board as an executive director with this round.

In the case of Dandelion, its challenge and opportunity has been in the world of legacy energy services. Built largely on fossil fuel systems and centralised operation models — you have large plants and generators located in one place that distribute their energy to smaller stations, which distribute to individuals — the idea behind Dandelion has been to built a heating and cooling system that is significantly more decentralised: it operates directly from a person’s home — or more specifically, underneath it — leveraging the ground’s natural state of being 50°F, in order to work.

One big issue with scaling up geothermal energy solutions prior to Dandelion has been the up-front installation costs, both from a financial and practical point of view. As Hannun has described it:

The process of installing ground loops in homeowners’ yards has typically been messy and intrusive, using wide drills that are designed to dig water wells at depths of over 1,000 feet. These machines are unnecessarily large and slow for installing a system that needs only a few 4” diameter holes at depths of a few hundred feet. So we decided to try to design a better drill that could reduce the time, mess and hassle of installing these pipes, which could in turn reduce the final cost of a system to homeowners.

The company’s solution has been to build a system that bores a much smaller hole (a few inches is all that’s needed) at a much shallower depth of hundreds of feet — making the installation something that can be done in less than a day.

So far, the company’s upfront costs might prove to be too much of a gating factor for the majority of homeowners. Installations run between $20,000 and $25,000 in upfront costs alone. For those willing to take the plunge — or dig into the challenge, as the case may be — over twenty years, the company has claimed that savings can be about $35,000.

The company also tells me that homeowners are buying a lot of these using financing and will save around 20 percent annually if they finance. (The savings percentage comes after a tax credit, a spokesperson said. “Here is a real example from one of our homeowners. He needed a 4-ton system, which is the size most homes require. He formerly spent $2,621 on fuel oil annually (832 gallons over the year at $3.15/gallon). To run geothermal, he requires $803 in additional electricity costs. With Dandelion pricing, starting at $115/mo, geothermal heating costs for his home are $1,380 + $803 = $2,183. This is about 20% savings annually for heating alone. His air conditioning will also be over twice as efficient with geothermal than it was with conventional a/c.”)

The savings in terms of using clean versus dirty energy, of course, come from the start.

Huawei is about to unveil its brand new flagship smartphone — the P20. While many details have already leaked, this is going to be an interesting launch. The conference will begin at 3 PM in Paris, 9 AM in New York, 6 AM in San Francisco.

The company chose to unveil its new device at the magnificent Grand Palais in Paris. This is the first time Huawei chooses Paris to launch a major new device. Carriers and retailers in the U.S. have stopped selling Huawei devices. That might be the reason why the press conference is happening in Europe.

Huawei is now the second largest phone manufacturer in the world behind Samsung and just slightly above Apple. The P10 was a great device, so it’s clear that many Android fans will pay attention to today’s launch.

So much for ‘We are accountable‘; Facebook founder and CEO Mark Zuckerberg has declined a summons from a UK parliamentary committee that’s investigating how social media data is being used, and — as recent revelations suggest misused — for political ad targeting.

The DCMS committee wrote to Zuckerberg on March 20 — following newspaper reports based on interviews with a former employee of UK political consultancy, Cambridge Analytica, who revealed the company obtained Facebook data on 50 million users — calling for him to give oral evidence.

Facebook’s policy staff, Simon Milner, previously told the committee the consultancy did not have Facebook data. “They may have lots of data, but it will not be Facebook user data,” said Milner on February 8. “It may be data about people who are on Facebook that they have gathered themselves, but it is not data that we have provided.”

In his letter to Zuckerberg, the chair of the committee Damian Collins accuses Facebook officials of having “consistently understated” the risk of user data being taken without users’ consent.

“It is now time that I hear from a senior Facebook executive with the sufficient authority to give an accurate account of this catastrophic failure of process,” Collins continues. “There is a strong public interest test regarding user protection.”

Regardless of rising pressure around what is now a major public scandal, Zuckerberg has declined the committee’s summons.

In a statement a Facebook spokesperson said it will be offering its CTO or chief product officer to answer questions.

“We have responded to Mr Collins and the DCMS and offered for two senior company representatives from our management team to meet with the Committee depending on timings most convenient for them. Mike Schroepfer is Chief Technology Officer and is responsible for Facebook’s technology including the company’s developer platform.  Chris Cox is Facebook’s Chief Product Officer and leads development of Facebook’s core products and features including News Feed.  Both Chris Cox and Mike Schroepfer report directly to Mark Zuckerberg and are among the longest serving senior representatives in Facebook’s 15 year history,” the spokesperson said.

Facebook declined to answer additional questions.

Collins made a statement before today’s evidence session of the DCMS committee, which is hearing from Cambridge Analytica whistleblower Chris Wylie — saying it would still like to hear from Zuckerberg, even if he isn’t able to provide evidence in person.

“We will seek to clarify with Facebook whether he is available to give evidence or not, because that wasn’t clear from our correspondence,” said Collins. “If he is available to give evidence, then we will be happy to do that either in person or by video link if that will be more convenient for him.”