Uber CEO Dara Khosrowshahi spoke briefly of what’s been going in the world of Uber’s self-driving cars in light of the fatal accident in Tempe, Arizona in March. At Uber Elevate today in Los Angeles, Khosrowshahi said Uber will bring back its self-driving cars “within the next few months, I don’t know.”

To be fair, Khosrowshahi didn’t sound the 100 percent confident in his answer. That’s just to say I wouldn’t be surprised if it takes Uber longer than a few months.

He went on to say that when Uber does bring its cars back on the road, “it’s going to be in as safe of a way as possible.”

Uber pulled its self-driving cars off the roads following the March fatal crash. Later that month, Uber decided not to reapply for a self-driving car testing permit in California. Uber’s previous permit expired March 31.

If Uber wants to continue its tests in California, it will need to apply for a new permit, as well as “address any follow-up analysis or investigations from the recent crash in Arizona,” DMV Deputy Director/Chief Counsel Brian Soublet wrote in a letter to Uber in March. Uber may also need to set up a meeting with the DMV.

In response to a question about how the investigation is going with the National Transportation Safety Board, Khosrowshahi said it’s ongoing. Unlike Tesla being quite vocal about the fatal crash involving its Autopilot system, and damaging its relationship with the NTSB, Khosrowshahi said, “We will not be tweeting ahead of their findings.”

But that doesn’t mean there won’t be leaks. Earlier this week, The Information reported Uber’s software was at fault. Specifically, it was reportedly the fault of the software that determines which objects to ignore and which to attend to.

Following the report, Uber said it’s actively cooperating with the NTSB and can’t comment on the specifics of the accident.

“In the meantime, we have initiated a top-to-bottom safety review of our self-driving vehicles program, and we have brought on former NTSB Chair Christopher Hart to advise us on our overall safety culture,” an Uber spokesperson said in a statement at the time. “Our review is looking at everything from the safety of our system to our training processes for vehicle operators, and we hope to have more to say soon.”

Uber’s flying taxis are taking off, as the transportation upstart looks for new ways to shorten trips made long because of distance or traffic congestion. Flying cars were once nearly the exclusive domain of tech aphorisms (“You promised us flying cars, but instead we have x.”), but now they are actually being put into gear in the form of electric vertical take-off and landing (eVTOL) vehicles. Over the last couple of days at the Uber Elevate summit in Los Angeles, the company further laid out its own ambitious plans to develop and commercially deploy air taxis by 2023.

Uber CEO Dara Khosrowshahi, who has been at the helm for less than one year, admitted he wasn’t initially 100 percent on board for Elevate, he said at the Uber Elevate Summit in Los Angeles. It took a couple of sessions and some reviews of the math for him to be sold on it, he said.

“For me the aha moment came when I started understanding that Uber isn’t just about cars,” Khosrowshahi said. “Ultimately, where we want to go is about urban mobility and urban transport, and being a solution for the cities in which we operate.”

Uber Elevate is Uber’s all-encompassing term for its initiative to launch uberAIR, which is the its aerial electric ride-hailing service, as well as any other initiatives (think food delivery) that may benefit from air transport. Elevate is also the name of the two-day conference Uber held in Los Angeles. Once Uber’s vision is fully implemented, Uber says the service will be cheaper than the cost of owning a car, on a per-passenger, per-mile basis, and autonomous. At launch, however, pilots will be required.

In the U.S., Uber is aiming to launch first in the Dallas-Fort Worth and Frisco, Texas, areas and Los Angeles. Last year, Uber said it would also aim to start testing in Dubai by 2020, but that’s no longer the case. Instead, Uber now has an open call out to interested international cities to describe the clear need for aerial transit, the enabling conditions of the city and local government commitment.

In order to launch uberAIR, Uber needs the actual vehicles, skyports for them to land on, as well as batteries. The company won’t be developing and producing its own vehicles. Instead, it’s relying entirely on its aerospace partners — some of which have been developing aircrafts for decades.

“There’s a lot that has to come together,” Khosrowshahi said about partnerships. “We absolutely know that we cannot make this happen ourselves.”

At the summit, Uber announced a new partnership with Karem to develop eVTOLs. Karem Aircraft, which has patented Optimum Speed Tiltroter technology for military and commercial applications, has been working with Uber for about a year to create the Butterfly concept. This type of vehicle is supposed to be a passenger-friendly adaptation of Karem’s core technology. And some of Uber’s previously announced partners also showed off what they’ve been working on over the past year. Embraer, for example, unveiled its first eVTOL concept.

At this point there are more than 70 companies working on eVTOLs for deployment in Uber’s air taxi network.

The company also needs skyports to enable people to board and exit these eVTOLs, which is where partners like Gannett Fleming and Corgan come in. On day two, these partners showed off their skyport designs as part of a skyport competition Uber ran.

Since Uber wants its offering to be all-electric from the start, it’s working with a number of battery partners. One of them is E-One Moli, a newly announced partner that will put its battery technology in the first eVTOL prototypes from Uber’s Elevate vehicle partners. With these batteries, uberAIR vehicles could travel up to 84 miles on a single charge, compared to just 60 miles. That also means Uber needs a way to charge these vehicles, which is where partners like ChargePoint come in.

Air traffic control

On day one of the summit, Uber Head of Aviation Eric Allison spoke about how certain skyports could handle hundreds or even thousands of landings per hour. In order to manage the skies and ensure uberAIR doesn’t simply replicate the horrendous traffic patterns we already have on the roads, Uber is working to develop systems that enable the ecosystem to function in what will be a more complex version of standard air traffic control, Allison told TechCrunch earlier in the day.

There are many ways to conceive of this, but one way is something Uber Director of Engineering for Airspace Systems Tom Prevot calls Dynamic Skylane Networks. He said you can think of them as a virtual network of lanes, overpasses, on-ramps and off-ramps in the sky that dynamically adjust to where the air traffic needs to flow.

But that’s a bit down the road. At the beginning of this process, Prevot said, Uber wants to work in parallel with what exists today and be “extremely cooperative, interoperable and transparent for safety and efficiency reasons. But we also need to protect, obviously, privacy information.”

He added that cybersecurity is a “first-class citizen and we need to bake that in from the beginning.”

Although Uber could theoretically create, own and control its own air traffic control system for eVTOLs, Uber says the intention is not to own it. Instead, the idea is to make it an open standard that other companies can work with, and therefore, enable interoperability.

“We don’t own airspace,” Holden said. “We’re just trying to make sure airspace is managed in an extremely safe and efficient way.”

To try to achieve this goal, Uber is working closely with the FAA and NASA. At the Summit, Uber announced it has signed a second space act agreement with NASA to model and simulate airspace requirements for urban air mobility applications. As part of the agreement, Uber will share its plans for implementing its air-based rideshare network.

Using data from Uber, NASA plans to simulate a small passenger-carrying aircraft flying through the Dallas-Fort Worth area. The idea is to identify potential safety issues in an already-crowded air traffic control system.

“We’re designing our flight paths essentially to stay out of the scheduled air carriers’ flight paths initially,” Prevot said at Elevate. “We do want to test some of these concepts of maybe flying in lanes and flying close to each other but in a very safe environment, initially.”

Regulating air taxis

All of these eVTOLs, of course, must comply with regulation from aviation authorities. At Elevate, FAA Acting Administrator Dan Elwell said he’s excited about everything that’s happening. But while Uber aims to start testing in 2020 and deploy commercially in 2023, all Elwell would say is, “We’ll see.”

“Everything is changing, but remember it’s changing within the construct that we’ve built, that we know,” Elwell said in a conversation with Uber Chief Product Officer Jeff Holden. “We have to adapt.” He added, whatever happens, the FAA is going to do this right. If you were to ask him what his reaction is to the idea that “it has to happen and this date is certain, my answer is, ‘well, we’ll see.'”

But Khosrowshahi said he’s confident in the 2020 testing timeframe. That’s because of the partners Uber has in place and the team on board, Khosrowshahi said.

“I think that’s something we can get to,” he said.

In tandem with ensuring uberAIR operates in ways that are safe and consistent with current FAA standards, sound regulation is key to community acceptance. For the purpose of uberAIR, Uber has tapped David Josephson, a noise and acoustics consultant at Josephson Engineering. At the Elevate summit, Josephson explained how urban air mobility noise will be different from the noise of a conventional aircraft. Part of that is due to the fact that airports are not located within city centers. With uberAIR, however, these skyports are going to be within city limits.

“We’ve decided to develop an entirely different set of metrics for this purpose,” Josephson said.

Those measurements entail looking at how many people are going to be affected by any given flight. More specifically, that means looking at how many people are going to be able to hear the noise emissions from the eVTOLs.

Aerial equity

Uber is ultimately presenting AIR as a way to increase access to transportation. In an ideal world, uberAIR would be able to reach neighborhoods that are traditionally underserved by transit agencies, Uber Head of Policy of Autonomous Vehicles and Urban Aviation Justin Erlich told me back in February. But in order to do that, Uber needs to remain conscious of the fact that it’s a goal it’s trying to achieve. That means ensuring the right policy infrastructure is in place and that’s where Erlich comes in.

“We’re thinking about what this looks like for making things wheelchair accessible and so we’re having ongoing conversations with folks in that community,” Erlich said. “We’ll really need to be thoughtful long-term about where the routings are to make sure that we’re serving underserved communities in transit, and to make sure that this technology is made available to everybody.”

In addition to reducing the cost of aerial transit, it’s important to note where the skyports will be located and what areas they will serve, World Economic Forum Head of Drones and Tomorrows Airspace Timothy Reuter said at Elevate.

“One of the causes of inequality in this country is, it’s very difficult for people to live in low cost areas but get access to high wages,” Reuter said.

And as Uber continues to scale and move further into autonomy, it will be worth paying attention to who is going to be tapped to handle the more monotonous, low-wage jobs.

“What we bring to the table here,” Khosrowshahi said, “is building this to not be a service for the few but a service that is ultimately available for mass market.”

If you’re using Signal for secure messaging, here’s something to be aware of. The app is one of the best-regarded encrypted messaging tools out there, but Mac owners who use Signal might inadvertently be putting their privacy at risk.

As Motherboard reports, security researcher Alec Muffett discovered that Signal messages sent to a Mac can persist in the notifications center, even if you have the app’s settings tuned to delete them.

That fact suggests that otherwise private messages live on in the operating system, which is something other researchers are looking into at the moment.

It’s a serious concern for anyone who relies on the Signal Mac app, but remember: to take advantage of this flaw, a hacker would need to compromise or obtain your Mac, and by then it’s probably game over.

To turn off the setting — and we recommend that you do — go to the Settings menu within the Signal for Mac app and select either “Neither name nor message” or “Disable notifications” to make sure that your private messages don’t stray beyond Signal.

Tailor Brands, a startup that automates parts of the branding and marketing process for small businesses, announced this morning that it has raised $15.5 million in Series B funding.

CEO Yali Saar has said the company sits at the intersection of design and machine learning. The idea is to create technology that understands the best practices of logo design, copywriting and social media strategy.

It’s the automated design that’s most immediately eye-catching, and that’s the big feature highlighted on the Tailor Brands website. You’ll need to pay to get access to high-quality image files, but before that, you can actually try creating a logo for free, just by entering some basic information about your company and identifying the designs you prefer.

Related: What do you guys think of the new TechCrunch logo?

Tailor Brands, which launched at TechCrunch’s Startup Battlefield in 2014, said the technology has already been used to create 45 million logos. The company says it had 3.86 million customers last year, and is adding half a million new businesses to the platform each month.

The new funding was led by Pitango Venture Capital Growth Fund and British Armat Group, with participation from Disruptive Technologies and Mangrove Capital Partners. The company has now raised a total of $20.6 million and says it will use the money to expand globally, add more languages and introduce more tools to its full branding suite.

It took some time, but Tiger Global Management is starting to wringing some serious paydays out of its tech investments.

In one of its biggest wins to date, the secretive, 16-year-old, New York-based hedge fund is reportedly set to make roughly $3 billion off Flipkart, the India-based e-commerce juggernaut that’s selling a majority stake in its business to Walmart for a whopping $16 billion.

Part of that stake includes three-quarters of the roughly 20 percent of Flipkart that Tiger had come to own since writing its first, $9 million, check to the company back in 2009.

In a lesser but apparent win, Tiger should also see a return on its investment in Glassdoor, the jobs and salary website that announced yesterday it is being acquired by the Japanese human resources company Recruit Holdings for $1.2 billion in cash.

Tiger had led the company’s $50 million Series E back in late 2013 when Glassdoor was still very much expected to go public. Earlier venture backers like Sutter Hill Ventures and Benchmark will see bigger returns as they bought in at the A round when Glassdoor’s valuation was just beginning to ramp up. Still, it’s probably safe to assume that Tiger made a little something, too. One clue is that in 2016, during its Series H round, Glassdoor was assigned a post-money valuation of $1 billion, presumably more than the company was worth when Tiger bought in more than two years earlier.

It’s a lot of good news, and it’s saying nothing of Spotify’s direct listing on the stock market last month, another deal that involves Tiger Global. In fact, Tiger had become one of streaming company’s biggest shareholders in recent years, and like Flipkart and Glassdoor, that position is also poised to pay off. Consider that Tiger owned 7.2 percent of Spotify as of its first day of trading in April, and the value of that stake was $1.9 billion. It may still be. Spotify’s shares are trading roughly where they started, and if Tiger has sold any of its stake, it hasn’t revealed as much. (Some of Spotify’s major record label shareholders have sold theirs and reported as much on their earnings calls.)

Either way, you can bet these wins will go a long way with investors, who are probably in the process of committing a lot of money to Tiger Global’s newest megafund. Why we think there’s one in the works: the outfit registered its last giant fund — a $2.5 billion vehicle — with the SEC a little less than 2.5 years ago. Not only has it become standard for venture units to raise new funds every couple of years, but Tiger puts its capital to work. Indeed, though Tiger Global has seemingly slowed down slightly from some furious deal-making in 2015 when it invested $1 billion across more than 50 companies  — it went on to report negative returns in 2016 — the outfit remains active.

Just last month, Green Bits, a four-year-old, San Jose, Ca.-based maker of point-of-sale software for cannabis retailers announced $17 million in Series A funding led by Tiger. It also recently wrote a check to Chargebee, a seven-year-old, Walnut, Ca.-based maker of SaaS subscription management and recurring billing software that closed on $18 million in Series C funding in March. In keeping with its longtime tradition, Tiger has also continued to invest in companies in India and other developing countries, participating earlier this year in a $51 million Series D funding for NestAway, a nearly four-year-old, Bengaluru, India-based home rental startup.

Altogether, Tiger Global has made more than 200 investments in 30 countries, according to Bloomberg. As for its Flipkart deal, the firm is presumably celebrating it today. The return on that investment will now be among the largest that the group’s $11 billion venture unit has seen, reports the outlet.

Listen, it’s late on day two of Google I/O and we’re all getting a bit punchy here. During a late-afternoon panel on design and Assistant, Google Principal Designer Ryan Germick explained that the company used improv skills to figure out how to build out the AI’s personality — and answer some of life’s more difficult questions.

One question Assistant gets “more often than you’d expect”: “did you fart?” For one thing, farts are always funny. For another, what’s the point of having a smart assistant if you can’t blame it for your various bodily odors?

Germick explained that the company went through various iterations of answers to the fart question, starting with something along the lines of “of course I didn’t, I don’t have a body.” That, it turns out, is not a particularly satisfying answer. Instead, the company embraced the “artful dodge,” using what anyone who’s taken an introductory improv class will tell you is known as “yes and-ing.”

So go ahead and ask your Assistant if it farted and you’ll probably hear something along the lines of “you can blame me, if you want,” along with somewhere in the neighborhood of 25 additional answers.

And always remember Isaac Asimov’s unspoken fourth rule of robotics: he who smelt it, dealt it. 

Given Google’s recent rebranding mode and a few pieces of news trickling out over the past week, it seemed safe to expect some key updates for the operating system formerly known as Android Wear. But the company only really mentioned Wear OS in passing at yesterday’s keynotes.

Google’s push to offer an open, Android-like experience for wearable devices has stagnated a bit in recent years, along with the category itself — but the company is pushing out some key updates for devs this week at I/O. Over on the Android Developer blog, the company is highlighting some key features for developer preview 2, which launches this week.

The biggest news here is the addition of an “enhanced battery saver mode.” Battery life is certainly one of the chief concerns on these devices, thanks to their relatively small size. While in the new mode, the device will sport a “power efficient watch face,” while shutting off radios, the touchscreen and tilt to wake — essentially all of those features that make your smartwatch smart. A quick press of the side button will show the time and a longer press will restore functionality.

Also new here is support for Actions on Google, which should make Wear devices more useful when it comes to working with third-party smart objects. That’s a nice addition that brings the products up to speed with the rest of the Google Assistant ecosystem.

This all comes a couple of days after an LG Wear OS watch appears to have hit the FCC. None of this is earthshaking, but at the very least, it shows that Google’s not entirely given up the wearables ghost.

ZTE wasn’t kidding around when it suggested that a U.S. Department of Commerce order would “severely impact” its survival. It’s hard to image a successful path around the seven-year ban on the sale of U.S. products to the company imposed after it reportedly failed to sufficiently reprimand staff for flouting Iranian sanctions.

Earlier today, in fact, the Chinese smartphone/telecom manufacturer announced that it had ceased its main business operations as it attempts to figure out the best way forward.

“As a result of the Denial Order, the major operating activities of the company have ceased,” the company wrote in an exchange filing spotted by Reuters. “As of now, the company maintains sufficient cash and strictly adheres to its commercial obligations subject in compliance with laws and regulations.”

The company has been scrambling since the ban, reportedly meeting with Google to determine if there’s some workaround to what might effectively end its use of Android and Play services. Non-U.S. players like Taiwanese chip maker MediaTek have been stepping up in the meantime, but even with that support, it’s tough to imagine ZTE efficiently rallying.

The Chinese government has also been said to be negotiating with the U.S. on the company’s behalf, in hopes of stemming what is looking more and more like a looming trade war between the two countries. The D.O.C. has reportedly opened up a similar inquiry into fellow Chinese smartphone maker, Huawei. 

ZTE also told Reuters that it’s been in touch with the U.S. Government “in order to facilitate the modification or reversal of the Denial Order by the U.S. Government and forge a positive outcome in the development of matters.” 

We’ve reached out to ZTE for further comment on the matter. We’ll update accordingly. 

There’s been another bomb at the box office, and it isn’t a movie.

MoviePass parent Helios & Matheson lost nearly half of its remaining value today as investors continued to flee the cash-burning movie service. That drop followed a 31 percent dive yesterday, after the company filed a statement with the SEC warning that it would have to sell equity in the coming weeks for it to remain solvent. Since Thursday’s opening bell last week, the stock has moved from $2.13 to $0.79, a drop of 63 percent. The company’s market cap is now $51.44 million.

MoviePass CEO Mitch Lowe said in a written statement that “Our burn rate has been slashed by 35-40% by the implementations and abuse prevention measures we have put in place over the last few weeks. We have always known, from when MoviePass took off in August, that it was going to be a high cash burn business model. We are not changing our guidance on 5 million subscribers by the end of this year – which should make us profitable/cash flow positive according to our business model. We have access in capital markets to over $300 million. So there is plenty of cash available to sustain the subscriber growth and movie-going habits of our users.”

Those are the facts as we know them, but let’s consider some of the options the company has now.

Even if you believe the market demand for Helios’ stock (I, for one, find them incredulous), there is an enormous challenge of converting that money into equity now. The envelope math looks like this: A month ago when the stock closed at $4.21, buying 20 percent of the company would have cost roughly $55 million. At the company’s current average burn rate of $21.7 million per month, that cash would have lasted approximately 2.5 months.

Now though, with the stock price so low, getting cash on the balance sheet today is a much harder proposition. That same $55 million that bought an investor a fifth of the company last month would be a complete buyout today. Buying 20 percent only costs a bit more than $10 million now, or roughly two weeks of burn.

So what’s the trick here that will save the company?

The obvious option is to radically control burn. The company could offer pricier tiers for heavy users of MoviePass, and could put a ceiling on the number of films a customer can watch per month as it did temporarily a few weeks ago. Lowe seems deeply committed to overall subscriber growth though, and that makes any sort of constraints on the product unlikely. The reason is that subscribers are the leverage Lowe needs to negotiate better partnership arrangements with theater chains, so he has to keep trying to grow users rapidly.

One theory is that the company could be negotiating equity deals with theater chains like AMC, which could be enticed by the low price of the stock to “buy in” to MoviePass’ popularity. Such media equity partnerships are not unusual — Sony, for instance, was a major shareholder in Spotify, as was Warner Music group, although both have since sold off large percentages of their holdings. Given the reliance of MoviePass on theater chains, building an equity partnership could prove to be the service’s savior.

A well-publicized partnership — including discounted movie tickets for MoviePass — could boost the stock significantly since the cost savings would improve the company’s burn rate. That could be an enticing proposition for the chains, since they could realize an almost immediate gain on their investment, plus the ongoing proceeds of a partnership going forward.

The other tactic would be to sign up more MoviePass subscribers who watch limited films. This is what might be called the “gym membership model” of trying to identify customers who want to buy a membership as an aspirational purchase, but who won’t actually use the facilities often. The challenge, beyond the incredibly short time period to try to build that marketing funnel, is that MoviePass appears to lose money on the very first ticket a customer purchases. The question isn’t how much revenue each customer generates, but how much the losses can be minimized.

The situation is a high-wire act, and the company will either hit the ground in the next few weeks, or it will right the ship, limit expenses and get enough equity investors to give it some cash to burn and keep on growing. I’d say use your MoviePass while you have it, but then again, that’s exactly why the company is faltering to begin with.

A team of Microsoft interns have thought up a new way to put A.I. technology to work – in a screenshot snipping tool. Microsoft today is launching their project, Snip Insights, a Windows desktop app that lets you retrieve intelligent insights – or even turn a scan of a textbook or report into an editable document – when you take a screenshot on your PC.

The team’s manager challenged the interns to think up a way to integrate A.I. into a widely used tool, used by millions.

They decided to try a screenshotting tool, like the Windows Snipping Tool or Snip, a previous project from Microsoft’s internal incubator, Microsoft Garage. The team went with the latter, because it would be easier to release as an independent app.

Their new tool leverages Cloud AI services in order to do more with screenshots – like convert images to translated text, automatically detect and tag image content, and more.

For example, you could screenshot a photo of a great pair of shoes you saw on a friend’s Facebook page, and the tool could search the web to help you find where to buy them. (This part of its functionality is similar to what’s already offered today by Pinterest). 

The tool can also take a scanned image of a document, and turn a screenshot of that into editable text.

And it can identify famous people, places or landmarks in the images you capture with a screenshot.

Although it’s a relatively narrow use case for A.I., the Snip Insights tool is an interesting example of how A.I. technology can be integrated into everyday productivity tools – and the potential that lies ahead as A.I. becomes a part of even simple pieces of software.

The tool is being released as Microsoft Garage project, but it’s open-sourced.

The Snip Insights GitHub repository will be maintained by the Cloud AI team going forward.