Tech News

Google is shutting down Duplex on the Web, its AI-powered set of services that navigated sites to simplify the process of ordering food, purchasing movie tickets and more. According to a note on a Google support page, Google on the Web and any automation features enabled by it will no longer be supported as of this month.

“As we continue to improve the Duplex experience, we’re responding to the feedback we’ve heard from users and developers about how to make it even better,” a Google spokesperson told TechCrunch via email, adding that Duplex on the Web partners have been notified to help them prepare for the shutdown. “By the end of this year, we’ll turn down Duplex on the Web and fully focus on making AI advancements to the Duplex voice technology that helps people most every day.”

Google introduced Duplex on the Web, an outgrowth of its call-automating Duplex technology, during its 2019 Google I/O developer conference. To start, it was focused on a couple of narrow use cases, including opening a movie theater chain’s website to fill out all of the necessary information on a user’s behalf — pausing to prompt for choices like seats. But Duplex on the Web later expanded to passwords, helping users automatically change passwords exposed in a data breach, as well as assisted checkout for ecommerce retailers, flight check-in for airline sites and automatic discount finding.

The promise of Duplex on the Web was that you’d be able to issue Google Assistant a command like “Book me a car from Hertz” and have Duplex pull up the relevant web page and automatically fill in details like your name, car preferences, trip dates and payment information (using information from Gmail and Chrome autofill). But the rollout was slow to begin with, with only a limited number of sites and partners supported for specific use cases. Android was the only platform from which Duplex on the Web could be used, with the service coming to Chrome for Android as “Assistant in Chrome” in late 2019.

Was the technical lift too much in the end for Google to justify maintaining Duplex on the Web? Maybe. As the Duplex on the Web support page outlines, Duplex used a special user agent that crawled sites as often as several hours a day to “train” periodically against them, fine-tuning AI models to understand how the sites were laid out and functioned from a user’s perspectives. It was surely resource-intensive, and could be tripped up easily if site owners chose to block the crawler from indexing their content.

Some brands were no doubt uncomfortable with the idea of Google inserting itself between them and their customers, as well. But perhaps the straw that broke the camel’s back was cuts on the Assistant side of Google’s business. According to a recent report in The Information, Google plans to invest less in developing Google Assistant for devices not made by Google, spurred by the idea that other areas of the enterprise, such as hardware, will prove to be more profitable over the long term.

Time will tell whether that’s the case. But what’s for sure is, Duplex on the Web has joined the infamous hall of much-ballyhooed-then-abandoned Google products.

Google shuts down Duplex on the Web, its attempt to bring AI smarts to retail sites and more by Kyle Wiggers originally published on TechCrunch

Shared micromobility company Bird has lost nearly all of its value since going public through a special purpose acquisition merger last year, falling from a 52-week high of $9.05 per share to around 23 cents per share this afternoon. In its short life on the public markets, Bird has garnered a reputation for burning through cash as it tries to be everywhere at once.

Bird’s free fall has investors and industry watchers questioning the company’s future and the state of the industry overall.

The upshot? Bird CEO and president Shane Torchiana predicts a major consolidation in the industry, with two or three companies coming out on top. Bird, he said, has a chance to be one of those companies.

That bullishness might prompt the rise of a few investor eyebrows considering the last year.

Bird has gone through a major restructuring, an executive shakeup, a round of layoffs,an exodus from multiple markets, a delisting warning from the New York Stock Exchange, a confession that it had overstated revenue for the past two years and a warning to investors that Bird may not have enough funds to continue operating for the next 12 months.

Torchiana contends the turmoil has forced Bird to take action and develop a strategy that drives down costs, improves efficiency and eventually even leads to profitability.

His plan includes increasing battery-swappable scooters, taking more control over asset allocation and making nice with cities. The company aims to be free cash flow positive by next year and to become adjusted EBITDA positive on a full-year basis, even if it needs to sacrifice some growth to achieve that.

Money, vehicles, ridership and staying lean

First things first: Bird needs to raise some more money so it can become a self-sufficient company. It closed out the third quarter with $38.5 million in free cash flow and operating expenses at $29.4 million.

Torchiana said he thinks around 3% or 4% of what Bird has raised historically should get the company out of its hole and into self-sustaining territory. Bird wouldn’t disclose its total funding amount, but per Crunchbase, the company has raised $883 million to date. That means it’ll need to scrounge together another $26 million to $35 million.

The problem is, given Bird’s shaky track record, investors are understandably dubious of claims that it can succeed. Tom White, an analyst at D.A. Davidson investment bank, said he isn’t sure which investors would throw Bird a bone at this point.

“Given Bird’s market cap, raising any significant amount of money would most likely mean substantial dilution for existing equity holders,” White told TechCrunch. “The white knight scenario here could be a strategic investment, where someone invests a lot of money for a decent-sized stake in the business.”

Bird’s plan to stay in the shared scooter game by Rebecca Bellan originally published on TechCrunch

Opendoor co-founder Eric Wu announced today that he is stepping down from his role as CEO of the real estate technology company and that Carrie Wheeler will be taking over as chief executive.

In a blog post, Wu said that he will transition to serving as president of marketplace. Wheeler has served as the chief financial officer of Opendoor since September of 2020.

In announcing the shift, Wu wrote: “First, I believe that an important driver of shareholder value is the discovery and refinement of product market fit, and I want to spend my days, nights, and weekends delivering products that address consumer needs. Second, and more importantly, I’ve spent countless hours with Carrie, and I know she is the leader and executive Opendoor needs.”

Wheeler will join Opendoor’s board, of which Wu will continue to serve as a member. John Rice, who has served on the company’s board since 2021, has been appointed lead independent director.

In early November, Opendoor launched a new marketplace called Opendoor Exclusives in its Dallas-Fort Worth and Austin markets. At the time, Wu said: “We’ve designed Opendoor Exclusives to be a new marketplace where you can directly buy and sell a home, without any of the hassle of the traditional real estate model. We are doing this by leveraging our nine years as a leading buyer and seller of homes, opening our platform and services, and aiming to bring together hundreds of thousands of our home sellers and home buyers. Today marks a founding moment as we expand our vision to service all homebuyers and homeowners and aim to redefine the real estate transaction forever.”

The moves come after a tumultuous year for Opendoor. On November 2, Opendoor announced it was letting go of about 550 people, or 18% of the company, across all functions.

At that time, Wu said his company was navigating “one of the most challenging real estate markets in 40 years.”

Meanwhile, the company’s stock has taken a massive hit. Opendoor went public in late December 2020 after completing its planned merger with the SPAC Social Capital Hedosophia Holdings II, headed by investor Chamath Palihapitiya. The eight-year-old company closed its first day of trading on the Nasdaq stock exchange at $31.25, “well above the $10 share price at which Social Capital sold shares in an April [2020] initial public offering,” per Crunchbase and Reuters reporting. At the time of writing today, shares were trading at $1.76, only slightly higher than the company’s 52-week low of $1.46. This means that the company is valued at just $1.11 billion, down from a valuation of $8 billion in 2021.

In August, Opendoor agreed to pay $62 million to settle charges by the Federal Trade Commission, which said the company’s claims that it helps people make more money by selling their house to the company rather than listing it on the open market were deceptive.

When it comes to venture capital, Opendoor last raised $300 million at a $3.5 billion pre-money valuation in March of 2019. Over time, it has raised about $1.3 billion in equity funding and nearly $3 billion in debt financing to finance its home purchases. Investors in the company include General Atlantic, the SoftBank Vision Fund, NEA, Norwest Venture Partners, GV, GGV Capital, Access Technology Ventures, SV Angel and Fifth Wall Ventures, along with others.

Founders include Wu and Founders Fund general partner Keith Rabois.

When one door closes … Opendoor CEO Eric Wu steps down, CFO Carrie Wheeler steps in as new chief executive by Mary Ann Azevedo originally published on TechCrunch

Activision Blizzard QA testers who work on “Diablo” in Albany, New York have formed the gaming giant’s second union in an unanimous 14-0 vote. Just seven months ago, QA testers at Activision Blizzard division Raven Software won a historic vote to form the first ever union at a major U.S. gaming company.

Activision Blizzard initially tried to block the union vote, arguing that developers at the Albany studio should be included in the unit. But the National Labor Relations Board (NLRB) ruled that the vote could go ahead as planned.

“With this victory, we’re advocating for ourselves and each other because we care deeply about our work and the games we make. Organizing has empowered all of us to fight hard for the dignity and respect every worker deserves on the job,” said Amanda Deep, an Associate Test Analyst at Blizzard Albany in a statement.

The “Diablo” QA testers will be unionized under the Communication Workers of America, which also represents the Raven Software union.

“Our colleagues at Raven inspired us when they announced the formation of the Game Workers Alliance/CWA. We can only hope that our win will continue to grow the labor movement at other video game studios across the country,” Deep said.

Labor organizers in the gaming industry are organizing to “to reject burn out culture, crunch, wage discrepancies and more,” the CWA said. But union activity at Activision Blizzard has been met with much pushback from management. In May, the NLRB found merit to a complaint that Activision Blizzard allegedly threatened employees for discussing their working conditions and wages.

Activision Blizzard workers in Albany vote to form the company’s second union by Amanda Silberling originally published on TechCrunch

On Wednesday, Erika Brodnock and Johannes Lenhard published their book, “Better Venture: Improving Diversity, Innovation, and Profitability in Venture Capital and Startups,” which serves as a guide for those looking to increase diversity, equity, and inclusion in the venture capital market.

The authors conducted more than 80 interviews to gauge how much progress there has — and hasn’t — been in the American and European startup ecosystem, suggesting that the current venture funding model is archaic. They point out that, in its current structure, it resembles the economic model used by European slave traffickers. This perspective puts into context how, centuries later, Black founders raise less than 2% of all venture capital funding.

Brodnock and Lenhard sat down with TechCrunch to discuss their book and what impact they hope it has on the industry.

This interview has been condensed and edited for clarity.

TC: How did you meet, and why did you decide to write this book?

JL: I was starting to do some more digging around diversity, equity, and inclusion topics in the VC world. I did a couple of pieces for Crunchbase, mostly around women, and Erika was one of the people I spoke to. It turns out that there was a much bigger project that it felt like we immediately needed to start attacking.

EB: I think the first piece we wrote together literally just jumped out at us and onto the page. It was really easy. He wanted to explore writing this book, and he pulled together a proposal. I quickly jumped into it and started to add bits and said, “You know what, how about we look at the history of venture, a look at where we are going, and a look at where we’ve been?”

Personally, I’ve always had these bold ideas, and I’ve always wanted to really change the system that I kind of came up with, whether in edtech for kids or venture capital. I want to disrupt the system. [Johannes has] opened the doors for us to be able to get, quite frankly, some of the most incredible people to speak us, who are incredibly senior in the venture capital world. It’s been an absolute delight to do this book.

Q&A: ‘Better Venture’ authors on why VC has failed to reinvent itself by Dominic-Madori Davis originally published on TechCrunch

Google said today it is testing end-to-end encryption for RCS-based group chats on its Messages app — RCS stands for Rich Communication Services. The company noted that it will be rolling out this feature to select users that are part of the app’s open beta program in the coming weeks.

This comes after a bunch of Redditors noticed that Google was testing end-to-end encryption for group chats in October. The company’s latest announcement makes it official.

The search giant first started testing end-to-end encryption for individual RCS chats in 2020. In June, it rolled out the feature to all users of its Messages app.

Google has been pushing manufacturers and carriers to adopt RCS — a hypercharged version of SMS with features like typing indicators, delivery, and read receipts. It has been also running campaigns to convince Apple to adopt this standard for its own Messages app. This way, folks using Android would be able to send rich multimedia messages with better photos and videos to their friends using iPhones.

While Apple hasn’t budged yet, Google didn’t miss a chance to nudge the Cupertino-based tech giant once again with the new announcement.

“Today, all of the major mobile carriers and manufacturers have adopted RCS as the standard – except for Apple. Apple refuses to adopt RCS and continues to rely on SMS when people with iPhones message people with Android phones, which means their texting is stuck in the 1990s,” Neena Budhiraja, group product manager for the Messages app, said in a blog post.

“Hopefully, Apple can #GetTheMessage so we don’t have to keep waiting to remove the whole ‘green-versus-blue bubble’ thing. Happy birthday, SMS – you were a great start, and you had a good run, but everyone is ready for an upgrade.”

Google has also tried to bridge the experience across two operating systems by adding features like reaction support for iPhone texts.

Google is testing end-to-end encryption for group chats in the Messages app by Ivan Mehta originally published on TechCrunch

Payment giant Stripe is going to offer a new product that makes it easier to hold cryptocurrencies without signing up to a cryptocurrency exchange. The company’s new fiat-to-crypto widget can be embedded in any crypto product so that users can enter their card information and acquire crypto that can be used in another web3 product.

There are a few use cases where widgets like this one make sense. The most common one is probably non-custodial wallets. Once you have crypto assets in your wallet, you can interact with various web3 products using WalletConnect. You can buy NFTs, use DeFi products, etc.

But how do you upload crypto assets to these software wallets? The most logical use case is that you first sign up to a centralized crypto exchange like Coinbase or Kraken. Once you have verified your identity, you can buy cryptocurrencies and then transfer those assets to your non-custodial wallet.

That’s not a seamless experience. That’s why non-custodial wallet developers like Argent or ZenGo have been using fiat-to-crypto widgets. Stripe isn’t the first company coming up with this product. Crypto companies like MoonPay and Ramp are already available in many crypto wallets.

In that case, MoonPay and Ramp take care of KYC requirements, fraud and compliance. They support dozens of fiat currencies and customers from more than 150 countries. Essentially, you create an account with MoonPay or Ramp when you want to buy crypto assets and top up your wallet — no exchange-to-wallet transfer required.

That’s a much smoother experience as you don’t have to jump between multiple products. Of course, convenience isn’t cheap. MoonPay charges 4.5% on card payments while Ramp charges 2.9%. They also both offer bank transfers for lower fees.

Stripe’s fiat-to-crypto onramp would work more or less similarly. It is currently only available to U.S. customers who are invited to test the product.

When a user wants to buy crypto, Stripe’s widget asks you how much you want to spend in your fiat currency — the company adds fees on top of that. You can then select the cryptocurrency you want and see how much you will receive once the trade is confirmed.

In Stripe’s example screenshot, the company charged $4.99 to buy $100 worth of USDC. But that fee may vary depending on the amount and payment method. As it is a Stripe product, it should work with different payment methods from day one, such as card payments, bank transfers, Apple Pay, Google Pay, etc. Behind the scenes, Stripe partners with Zero Hash to obtain cryptocurrencies.

Stripe expects that its widget will appeal to developers working on DEX, NFT platforms, wallets and dApps. For instance, the company says it has been testing the product with Audius (a blockchain-based music streaming platform), Magic Eden (an NFT marketplace) and Argent (an Ethereum wallet that I’ve already covered).

The payment company handles KYC, payments, fraud and compliance so that crypto companies can focus on crypto instead of competing with Stripe. There are also some side benefits with Stripe’s fiat-to-crypto onramp product.

For instance, it integrates with Link. If you have made a Stripe-powered purchase in the past and saved your payment information, Stripe can help you retrieve this information and pay more quickly.

After the FTX debacle, products like Ramp, MoonPay and Stripe’s fiat-to-crypto widget are going to be essential. They can make it easier to avoid centralized exchanges. And it’s good to see that there are multiple companies working on that issue.

Stripe announces fiat-to-crypto onramp widget by Romain Dillet originally published on TechCrunch

GoStudent — the Austria-based late-stage Tutor marketplace which has raised $686.3 million (and attained a €3BN valuation) to date — has acquired Germany-based Studienkreis, a traditional tutoring company, founded in 1974, which is focused on the German-speaking ‘DACH’ region. Terms of the deal were not disclosed.

Studienkreis had previously been owned, since 2017, by the London-based private equity firm IK Partners. It is best known as a tutoring provider in Germany with more than 1,000 learning centres nationwide, serving 125,000 families annually.

GoStudent said Studienkreis would continue to operate under its existing brand, while the the two companies looked for synergies

The acquisition gives GoStudent a substantial offline presence, ready-made market for its online offering and the ability to cater to families that prefer their children participate in group or real-world, centre-based classes.

The startup has rolled-up a number of other companies in the last 12 months, including UK-based Seneca Learning, Tus Media Group in Spain, and Fox Education from Austria.

The global online tutoring market, valued at USD 150 billion in 2020, is projected to reach USD 278 billion by 20261.
1 Global Online Tutoring Market Size, Status and Forecast 2021-2027.

According to data from Status and Forecast the global online tutoring market is was valued at $150 billion in 2020, and is projected to reach $278 billion by 2026.

Felix Ohswald, CEO and co-founder of GoStudent said in a statement: “Over 1.5 million online tutoring sessions are booked each month at GoStudent, but we believe the future of learning is hybrid. Combining online and offline creates an omnichannel model which brings maximum value to families and builds a barrier for competitors.”

Earlier this year GoStudent raised a $340M Series D funding round to push into international markets.

Meanwhile, in the more vertical language learning tutor space, Preply closed a round of $50 million this year.

GoStudent uses its warchest to acquire large network of traditional tutoring centres in Europe by Mike Butcher originally published on TechCrunch

Prior to launching as a startup in 2005, Box began as an idea that co-founder and CEO Aaron Levie had for a marketing class — to bring the power of the internet to file-sharing. The concept may not feel revolutionary today, but back then, you could email a file if it was small enough, or you could put it on a thumb drive and physically deliver it to the recipient. Other options were limited.

It’s hard to believe now, but the original academic idea grew into a startup, and later a way to take on the entrenched enterprise content management industry.

Box, which began so modestly, reported an even $250 million in revenue for the most recent quarter, the third quarter of its fiscal 2023, putting it on a $1 billion run rate for the first time. (Notably, we were told back in 2014 or so by venture capitalist Jason Lemkin that Box would reach the $1 billion run rate figure one day; he also predicted that it wouldn’t be easy. Two points, Lemkin.)

“We’re really proud of the fact that this is our first billion-dollar revenue run rate quarter, so we can now say that we’ve crossed that billion revenue threshold, which is super exciting,” Levie told TechCrunch.

Revenue was up 12% in the quarter compared to last year, more modest growth than Box has posted in recent quarters. What contributed to the slowdown? Levie said growth was affected by the strong U.S. dollar, which is impacting many companies right now. Measured using older currency exchange rates (“constant currency,” in corporate speak), Box’s growth would have been 17%, much more in line with recent reports.

“It’s pretty material actually, and the way we talk about it is that previously if we’d sold the deal for $1, we’re getting 80 cents now on that deal. So that’s a material headwind,” he said.

But in spite of the economic challenges that everyone is facing right now, Box is taking advantage of the need for customers to work remotely, or at least spend significantly less time away from a conventional office, and Levie said his company’s solutions have become mission-critical for customers.

“I think we’re being very strategic. I think the way companies manage content is very strategic, and so I think that puts us in a good position relative to other software companies because of that value proposition,” he said.

While he doesn’t have a crystal ball to see what budgets will look like moving forward, he does believe that Box remains in a strong position. The company seems to be looking at emphasizing profitability over growth, something that should please Wall Street investors right now. How did that translate into this quarter’s numbers? Let’s have a look.

Box reaches $1B run rate in spite of a quarter dogged by currency challenges by Ron Miller originally published on TechCrunch

If you have been on Twitter in the last few days, you likely noticed a deluge of screenshots from a service called ChatGPT. From the OpenAI group, ChatGPT is a conversational tool that allows you to provide the system with prompts that it responds to in written format.

(You can make a free OpenAI account and give the service a shake yourself. Just don’t identify as a journalist during the onboarding process — you’ll get jammed up. Self-describe in a different manner and you can get right in.)

The Exchange explores startups, markets and money.

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

TechCrunch has been busy covering OpenAI lately, with our own Darrell Etherington writing this morning about ChatGPT and how it is “quickly becoming apparent that how a user interfaces with generative models and systems is at least as important as the underlying training and inference technology.” We also have eyes on new generations of the well-known, and well-liked, GPT-3 AI writing tool.

That there is excitement among founders and venture investors in generative AI services like what OpenAI is building is well-known. Hell, it’s been a minute since Copy.AI showed that leveraging AI writing tools could build eight-figure ARR startups. Others are chasing similar magic, and the technology continues to improve. Hence all the screenshots.

Naturally, I had to see if I was in near-term employment trouble. So I ran a little test with the ChatGPT model this morning. The results are below.

Let’s have some fun

First, a reminder that using computers to generate text is not new, and this is not the first time that I have stared down the barrel of a new tech tool that could, in theory, be coming for my job. Back in 2014, responding to news that the AP was planning on using some automated tech to report on corporate earnings, I wrote the following:

ChatGPT isn’t putting me out of a job yet, but it’s very good fun by Alex Wilhelm originally published on TechCrunch