Connect with us


This Week in Apps: Clubhouse clones, WWDC21, apps have their best-ever quarter



Welcome back to This Week in Apps, the weekly TechCrunch series that recaps the latest in mobile OS news, mobile applications and the overall app economy. The app industry is as hot as ever, with a record 218 billion downloads and $143 billion in global consumer spend in 2020.

Consumers last year also spent 3.5 trillion minutes using apps on Android devices alone. And in the U.S., app usage surged ahead of the time spent watching live TV. Currently, the average American watches 3.7 hours of live TV per day, but now spends four hours per day on their mobile devices.

Apps aren’t just a way to pass idle hours — they’re also a big business. In 2019, mobile-first companies had a combined $544 billion valuation, 6.5x higher than those without a mobile focus. In 2020, investors poured $73 billion in capital into mobile companies — a figure that’s up 27% year-over-year.

This Week in Apps will soon be a newsletter! Sign up here:

This week we’re looking into app store trends, Apple’s upcoming WWDC, new App Store rejections and what they mean for ATT (App Tracking Transparency), and whatever happened to that Arizona App Store bill, among other stories.

Top Stories

Apps just had the biggest quarter on record

Big news for the app economy this week, as App Annie reported that consumer spending on mobile apps broke a new record. According to new data, worldwide consumers in Q1 2021 spent $32 billion on apps across both iOS and Google Play, up 40% year-over-year from Q1 2020. It’s the largest-ever quarter on record, with mobile consumers spending roughly $9 billion more in Q1 2021 compared with Q1 2020, in part due to the pandemic’s continued impacts.

Image Credits: App Annie

Although iOS saw larger consumer spend than Android in the quarter — $21 billion versus $11 billion, respectively — both stores grew by the same percentage, 40%. Gaming drove a majority of the quarter’s consumer spending, as usual, accounting for $22 billion of the spend — $13 billion on iOS (up 30% year-over-year) and $9 billion on Android (up 35%).

Apple begins rejecting apps with fingerprinting tech

With the launch of iOS 14.5 looming, reports circulated this week that Apple has begun rejecting apps that use third-party SDKs that track users via a method called device fingerprinting. The Adjust SDK was one that didn’t yet comply with Apple’s new App Tracking Transparency guidelines, causing apps that included the SDK to be rejected. That could have been a large number of possible rejections, as Adjust’s website claims it’s trusted by more than 50,000 apps. But Adjust soon updated its SDK (open-sourced and here on GitHub), to hopefully re-enable app updates for its customers.

The App Tracking Transparency (ATT) changes have thrown a whole industry into disarray as companies scramble to comply and diversify their revenues. Snap, however, was found to be looking into alternative ways to bypass ATT by gathering data like IP addresses from companies that analyze ad campaigns to see if it could then cross-reference that data with its own, in order to continue tracking its users. Snap told the Financial Times it had tested the technique, called probabilistic matching, to test the impact of Apple’s policies, but claimed it intended to discontinue the program after Apple introduced its changes. (Sure Jan!) The company says it believes that tracking individual users will no longer be allowed going forward, but gathering data on “cohorts” would be. On Thursday, Apple sent letters to developers warning them to remove code that supported probabilistic matching, just in case.

WWDC 2021 announced

WWDC will again be virtual, Apple announced this week. The online event will take place June 7 to 11, as it did last year, allowing developers worldwide to tune in to watch prerecorded keynotes and sessions, and virtually network and learn from Apple engineers and employees. Though April 18, students can submit their Swift playground to this year’s Swift Student Challenge to win exclusive outerwear and Apple pins.

Everyone has already begun to read wayyyy too much into the imagery Apple published, which shows Memoji characters wearing glasses looking at a Mac screen. Is Apple teasing AR glasses?, some wondered. That doesn’t seem likely, though. The glasses are just a way to reflect the computer screen in a picture whose message largely conveys, yep, it’s another virtual event this year.

The Arizona App Store bill is dead

💀The bill was the work of the Coalition for App Fairness, led by Epic Games, Spotify, Tile and other developers who want alternatives to paying app store commissions and alternative distribution channels. Last week, it had been unclear why the AZ Senate skipped the vote on the bill, which had been passed by the AZ House. But according to the bill’s (HB2005) backer, Rep. Regina Cobb, speaking to The Verge, the decision was due to heavy lobbying by Apple and Google, which caused Senate members who had agreed to a vote to waiver. When the votes weren’t there, the Senate decided to skip bringing it up altogether.

Cobb, in an email to TechCrunch, confirmed the same. “We have been working with members on the committee and had the votes a few days prior,” she said. “Just before the committee was set to proceed, the Chairman notified the lobbyist for App Fairness that the votes were not there so he did not want to waste committee time on the issue.”

Weekly News

Platforms: Apple

📱Apple released the iOS 14.5 beta 6 to developers, which includes one major change: the release of new Siri voices. With iOS 14.5, Siri will no longer default to a female voice, but will rather allow customers to choose from a set of voices, presented in random order. Prior to the release, there had been some expectation that beta 5 would be the last before the public release, but that turned out to not be the case. However, now that beta 6 has arrived, the Release Candidate could follow as soon as next week.

Ahead of WWDC 2021, Apple updated its Apple Developer app. The app offers an updated look-and-feel, which now supports a sidebar navigation on iPad, full-screen video on larger Mac displays and a new way to discover content to watch and read in an updated Search section. As with last year’s WWDC, attendees will be able to connect with and tune into announcements, sessions and 1:1 labs via the Developer app.

📱 U.S. users spent an average of $138 on iPhone apps in 2020, and the number is expected to grow to $180 in 2021, reported Sensor Tower. Throughout 2020, consumers turned to iPhone apps for work, school, entertainment, shopping and more, driving per-user spending to a new record and the greatest annual growth since 2016. Per-device spending on mobile games was a large part of that figure, growing 43% year over year from $53.80 in 2019 to $76.80 in 2020. That’s more than 20 points higher than the 22% growth seen between 2018 and 2019, when in-game spending grew from $44 to $53.80.

Image Credits: Sensor Tower

Platforms: Google

🔎 Google will begin to limit Android 11+ apps from being able to see what other apps the user has installed on their devices starting on May 5. The company now considers this “sensitive information,” though for years had allowed the practice. That made it easy for data-gathering apps that catalog to operate. Now, Google will only allow a few exceptions — apps that will be able to use the permission include antivirus, file managers and browsers. Expect to see data-gathering firms quietly release those sorts of apps in the near future.

Augmented Reality

Google expanded the AR virtual art galleries in its Arts & Culture app to include three new Pocket Galleries from  the Jean Pigozzi Collection and J. Paul Getty Museum.

Health & Fitness

A feature in The Cut this week raises questions about the viability and effectiveness of therapy apps like Talkspace, BetterHelp and AbleTo, some which use dating app-like models for matching clients to therapists as well as similar business models, where you can upgrade to live video and more features. The article questions whether the apps are able to deliver on their promises and notes how some have already faced questions over their use of data, sponsorship deals, treatment of workers and healthcare rule violations.

The global airline industry body IATA said it will launch a digital travel pass for COVID-19 test results and vaccine certificates on iOS by mid-April. The travel pass, meant to help speed up check-ins, is being tested already by U.K.-based Virgin Atlantic, which is using the IATA app on its London to Barbados flight on April 16.

Apple Maps adds COVID-19 travel guidance for over 300 airports worldwide in its recent update. The app will now show COVID-19 health measure information for airports when searched via the app, either through a link to the airport’s own COVID-19 advisory page, or directly on the in-app location card itself.

Social Networks

Image Credits: Facebook

Facebook teamed up with the U.S. Department of Health and Human Services (HHS) and Centers for Disease Control and Prevention (CDC) to launch a set of profile frames that will help to encourage vaccinations against COVID-19. The frames, offered in either English or Spanish, will say “Let’s Get Vaccinated” or “I Got My COVID-19 Vaccine.” Research shows that social norms can impact people’s behavior when it comes to their health, so the hope is that, as more adopt the frames, those hesitant about the vaccine will be encouraged to get one. Facebook will also begin showing users how many of their family and friends have been vaccinated, based on the frame usage.

Facebook this week introduced new tools that allow users to more easily switch to the “Most Recent” view or “Favorites” view of their News Feed, as well as tools to limit comments on posts and other changes. The launch came at the same time as the company argued in a long-winded post how its personalization algorithms are not at fault for the world’s ills and the rising spread of misinformation and dangerous health content. Rather, it blames the users who seek out this content and then engage with it. The company is pushing regulators to give it guidelines on misinformation, knowing that it will have an easier time with increased regulations than any new social media upstarts.

Image Credits: Facebook

Instagram launched Remix on Reels, a TikTok Duets-like feature that allows users to record a video alongside someone else’s content in order to react, collaborate and more. Snapchat is developing its own Duets feature with the same name of Remix, too.

Snap is planning to expand further into hardware, The Information reported, with a new, more advanced version of its Spectacles smart glasses that include AR effects. It’s also reviving an older effort to build its own drone.

Instagram appears to be developing a new way to block the trolls. Reverse engineer Alessandro Paluzzi found in the app’s code a reference to a new option, “Hide More Comments,” which will automatically hide comments with frequently reported words.

Instagram announced it’s expanding its IGTV ads internationally for the first time, initially to markets including the U.K. and Australia.

Facebook extended the deadline for iOS apps using its Facebook Audience Network monetization platform to migrate their apps to bidding only — a response to Apple’s iOS 14 changes. The previous deadline was March 31, which has now been pushed to May 31. Android apps have until September 30.

Streaming & Entertainment

Business-focused social network LinkedIn confirmed it’s building a Clubhouse rival that will offer a speaker stage, as well as tools to join and leave the room, react to comments, request to speak and more. The company sees the feature as an extension of its existing offerings for creators, which also includes things like LinkedIn Live, Stories and newsletters. It says that it believes its offering will be able to differentiate from others because its network will be connected with people’s professional identity, not just a social profile.

Image Credits: LinkedIn

Spotify significantly expanded its personalized playlist lineup with the launch of three new categories of playlists, called “Spotify Mixes.” Inspired by Daily Mixes, these new options offer easy-to-understand titles that include artist mixes (e.g. Drake Mix), genre mixes (e.g. Pop Mix) and decade mixes (e.g. 2010s Mix).

Audioburst launches a platform that will allow developers to integrate podcast feeds into mobile apps. The platform supports both partial or complete podcasts and radio shows and other streaming audio.

Discord became yet another tech company to offer its own Clubhouse-like feature with the launch of Stage Channels. The feature allows certain people to broadcast to a group of listeners, who can raise their hand to ask questions, while moderators can bring people onstage, remove them or put them on mute. Stage Channels are available on desktop, web and in Discord’s mobile apps.

Stage Channels on Discord

Image Credits: Discord/TechCrunch


Apple’s subscription-based gaming service, Apple Arcade, was significantly expanded this week with the addition of more than 30 new titles, including new exclusives like NBA 2K21, Star Trek: Legends and a new version of The Oregon Trail from Gameloft. But it also notably adopted a new strategy as well, with the addition of older titles in two new categories called “Timeless Classics” and “App Store Greats.”

Image Credits: Apple

The former will include well-loved games like chess, backgammon, sudoku, crosswords, solitaire and others, while the App Store Greats will bring award-winning titles to Arcade like Threes!, Mini Metro, Fruit Ninja Classic, Monument Valley and Chameleon Run — all of which will be free and unlocked. With the additions, Apple Arcade has now grown to more than 180 total games for $4.99/mo (or included with an Apple One subscription.)

Finnish mobile gaming giant Supercell, developers of Clash of Clans and Clash Royale, announced it’s making three new Clash titles: a turn-based tactical game, Clash Quest; a virtual board game, Clash Mini; and a co-op action roleplaying game, Clash Heroes. The games are in early stages of development and will be killed if they don’t meet Supercell’s high standards.


Following the meme stock frenzy, free stock trading app Robinhood will remove its controversial digital confetti feature, which critics said was a gamification technique — and not something that should be associated with financial investing.

WhatsApp Pay, the messaging app’s P2P payments feature, is now authorized in Brazil. The feature was first announced to roll out in Brazil last year, but had to be delayed after the country’s central bank said it could damage existing payments systems. It also said WhatsApp had failed to obtain proper licensing. On Tuesday, the bank said the service could go forward — now that it’s launched its own payments system called Pix, which has been widely adopted.

Bakkt released a digital wallet app for trading fiat currency for crypto, but also supports other digital assets not on the blockchain like frequent flier miles, loyalty points, gift cards and in-game assets.


Microsoft’s Cortana mobile app for iOS and Android officially shut down on Wednesday, March 31. The AI-based personal assistant was Microsoft’s answer to Siri, Alexa and Google Assistant, but with more integration into Microsoft’s own apps. The mobile version, however, never caught on and Microsoft has since refocused on AI assistance built directly into its Microsoft 365 apps instead.

Google’s in-house incubator Area 120 launches Stack, an Android app that digitizes personal documents, IDs, receipts and more. It can then automatically name the files, extract key info (like the bill’s due date) and organize them, with help from Google’s DocAI technology. The app can also sync your scans to Google Drive for easy access.

Image Credits: Google

Amazon reports rising demand and engagement for productivity apps on Amazon Fire tablets. During 2020, the pandemic-fueled changes to work and home life led to a 62% increase in productivity by app users month over month across the tablet app store, and a 226% growth in app engagement amongst productivity customers in the last year. The company says it expects these trends to continue in 2021.

Government & Policy

Republicans on the House and Senate antitrust committees wrote letters to Apple, Google and Amazon, asking them about Parler’s removal from their respective platforms, framing it as targeting “one small business.” The tech companies had removed Parler for failing to follow their community guidelines over content and moderation after finding that the app had been used to plan, coordinate and facilitate the storming of the U.S. Capitol.

Apple’s App Store has been found to be hosting more than a dozen apps created by a China paramilitary group accused of Uyghur genocide, The Information reported. The organization, which has been sanctioned by the U.S. for human rights abuses, has published apps that offer news, info about government services and provides aid to small businesses with e-commerce orders, and more.

South Korea’s antitrust regulator said Wednesday it’s going to refer Apple’s local unit and one of its executives to the prosecution for allegedly impeding the regulator’s probe into unfair business practices. It also said it will fine Apple Korea 330 million won ($265,000 USD) for hampering the investigation.

Security & Privacy

The Odyssey Team announced the release of its new Taurine jailbreak for all devices running iOS 14-iOS 14.3. The jailbreak is installed using AltServer, similar to the unc0ver jailbreak.

A research report covered by Ars Technica found that Google collects up to 20x more data on Android users than Apple collects from iOS users. The report had analyzed the telemetry data transmitted directly to the companies themselves through pre-installed apps and when idle. Google disagreed with the report’s finding and says there were flaws in the researcher’s methodology that caused findings to be off by “an order of magnitude.”

The Washington Post covered a story about how an App Store scammer stole a user’s life savings, $600,000, in bitcoin. The user believed the app was safe because Apple markets its App Store as trustworthy, but the app itself had been designed to trick users by pretending to be associated with Trezor, the maker of a hardware device used to store cryptocurrencies. The actual Trezor said it’s been notifying Apple and Google for years about fake apps that were scamming its customers, to no avail. The fake app got through App Review then changed itself to a crypto wallet without Apple’s knowledge.

Funding and M&A

Image Credits: Cameo

💰 Celebrity video request app Cameo raised $100 million in Series C funding in a round led by Jonathan Turner with The new funds value the business at just north of $1 billion.

🤝 Spotify acquired Betty Labs, the maker of the live audio app Locker Room with the goal of taking on Clubhouse. The app currently focuses on sports talk but will be expanded to cater to a wider range of fans and creators across “a range of sports, music, and cultural programming.” It will also add “a host of interactive features that enable creators to connect with audiences in real time,” like Clubhouse. The rise of live audio has offered a potential threat to Spotify’s investment in podcasts, as it connects users to audio programming in real time, instead of through recorded and produced shows. Spotify had earlier said it would add interactivity to its Anchor podcast recording app to better connect podcasters and listeners, but with the Locker Room sale it’s taking a further step towards fully embracing live audio.

💰 Holler raised $36 million from CityRock Venture Partners and New General Market Partners to power “conversational media” in your favorite apps. The company works with partners like PayPal-owned Venmo and The Meet Group to bring AI-powered recommendations of GIFs and stickers to use when messaging in the apps. The content is monetized in partnership with companies like HBO Max, IKEA and Starbucks for branded stickers. Holler’s content now reaches 75 million users monthly, up from 19 million last year.

💰 Madison-based Fetch Rewards raised a $210 million Series D for its customer loyalty and retail rewards app, valuing the business at more than $1 billion. The round was led by SoftBank through its Vision Fund 2, and makes Fetch Rewards one of the Midwest’s few unicorns. The app’s 7 million active users earn points by sharing photos of their receipts, which allows them to earn gift cards and participate in sweepstakes. The funding will help the company speed up its ability to process online receipts.

💰Funko, makers of pop culture collectibles, entered the NFT market by acquiring a majority stake in TokenWave, the makers of the TokenHead mobile app. The app allows users to showcase and track their NFT holdings, and today has over 10 million NFTs on display. Funko aims to launch its first NFTs in June, at starting prices of $9.99, with a unique property offered each week. Its products will be sold on the Worldwide Asset Exchange.

💰 Chinese video streaming platform Bilibili invested HK$960 million (around $123 million USD) into X.D. Network, the makers of the game distribution platform TapTap, giving it a 4.72% stake in the company. X.D. is seen as a competitor with traditional game distributors in China, including the Android app stores operated by smartphone makers.

💰 Chinese grocery app Nice Tuan raised $750 million from Alibaba, DST Global, and others to bolster its supply chain and increase its fresh produce offerings. The company today serves 1,598 cities and counties across China, and sees 15 million orders per day.

💰 U.K.-based Nested raised an additional £5 million for its modern real estate agency that connects home buyers and agents using technology across web and mobile.

💰 Expense tracking app Ensemble raised $3 million in seed funding for its app that helps divorced parents track shared expenses, like medical bills, extracurricular activities, transportation and other things that may fall outside of child support, which tends to cover just food, shelter and clothing needs.

📉 China’s Q&A site Zhihu, a Quora rival, dropped 11% to the bottom of its marketed range following its $522.5 million U.S. IPO. The company is the largest Q&A community by average mobile monthly users and revenue, with 75.7 million MAUs in the last quarter.

💰 Stockholm-based sport fishing app Fishbrain raised $31 million in funding for its mobile app that offers both a social network and social commerce platform for fishing with nearly 12 million registered users worldwide.

💰 Indonesian investment app Ajaib added $65 million to its Series A, bringing the round to $90 million, in an extension led by Ribbit Capital — the fintech investor who also led Robinhood’s $3.4 billion round last month. This is its first investment in Southeast Asia.

💰 Gaming startup Lowkey raises $7 million from Andreessen Horowitz for its app that allows gamers to both create and view short gaming clips. The company sees an opportunity in helping gamers cut down their existing content for distribution to short-form video platforms, like Instagram and TikTok.


Smart Photo Widget

Image Credits: Cromulent Labs

Cromulent Labs, the makers of Launcher and other apps, this week introduced a Smart Photo Widget that automatically frames each photo to fit the widget size you selected. You can also configure the widget to skip any bad photos from album (blurry, dark, screenshots, etc.) as it rotates through the best photos throughout the day. In the Premium version, you can add a date and time or other overlays and filter your photos to match your style. The app is free on iOS with in-app purchases.

Frenzic: Overtime (Soon)

Image Credits: Iconfactory

The popular app publisher and Twitterific developer Iconfactory announced on Friday it will soon launch Frenzic: Overtime, a sequel to the game that first launched on the iPhone App Store 13 years ago. The title will bring players over 45 levels with multiple gameplay modes and hundreds of mini-goals to achieve as they unlock the secrets of Frenzic Industries in this arcade style puzzler. The title will be available on Apple Arcade, but subscribers can already visit the game’s listing page and set a notification to be alerted to its launch.

SPIN Safe Browser

Image Credits: SPIN Safe Browser

A parental control app maker Boomerang, which continues to face App Review roadblocks, pivoted this year to develop safe browser technology. A new version of its SPIN Safe Browser designed for schools in need of filtering content was released this week. The app leverages AppConfig for content filter management through MDMs like Jamf School and Jamf Pro. Administrators can use the feature to enable or disable any of the categories as well as specific URLs that should be blocked or allowed.


Lyron Foster is a Hawaii based African American Musician, Author, Actor, Blogger, Filmmaker, Philanthropist and Multinational Serial Tech Entrepreneur.

Continue Reading


Facebook faces ‘mass action’ lawsuit in Europe over 2019 breach



Facebook is to be sued in Europe over the major leak of user data that dates back to 2019 but which only came to light recently after information on 533M+ accounts was found posted for free download on a hacker forum.

Today Digital Rights Ireland (DRI) announced it’s commencing a “mass action” to sue Facebook, citing the right to monetary compensation for breaches of personal data that’s set out in the European Union’s General Data Protection Regulation (GDPR).

Article 82 of the GDPR provides for a ‘right to compensation and liability’ for those affected by violations of the law. Since the regulation came into force, in May 2018, related civil litigation has been on the rise in the region.

The Ireland-based digital rights group is urging Facebook users who live in the European Union or European Economic Area to check whether their data was breach — via the haveibeenpwned website (which lets you check by email address or mobile number) — and sign up to join the case if so.

Information leaked via the breach includes Facebook IDs, location, mobile phone numbers, email address, relationship status and employer.

Facebook has been contacted for comment on the litigation.

The tech giant’s European headquarters is located in Ireland — and earlier this week the national data watchdog opened an investigation, under EU and Irish data protection laws.

A mechanism in the GDPR for simplifying investigation of cross-border cases means Ireland’s Data Protection Commission (DPC) is Facebook’s lead data regulator in the EU. However it has been criticized over its handling of and approach to GDPR complaints and investigations — including the length of time it’s taking to issue decisions on major cross-border cases. And this is particularly true for Facebook.

With the three-year anniversary of the GDPR fast approaching, the DPC has multiple open investigations into various aspects of Facebook’s business but has yet to issue a single decision against the company.

(The closest it’s come is a preliminary suspension order issued last year, in relation to Facebook’s EU to US data transfers. However that complaint long predates GDPR; and Facebook immediately filed to block the order via the courts. A resolution is expected later this year after the litigant filed his own judicial review of the DPC’s processes).

Since May 2018 the EU’s data protection regime has — at least on paper — baked in fines of up to 4% of a company’s global annual turnover for the most serious violations.

Again, though, the sole GDPR fine issued to date by the DPC against a tech giant (Twitter) is very far off that theoretical maximum. Last December the regulator announced a €450k (~$547k) sanction against Twitter — which works out to around just 0.1% of the company’s full-year revenue.

That penalty was also for a data breach — but one which, unlike the Facebook leak, had been publicly disclosed when Twitter found it in 2019. So Facebook’s failure to disclose the vulnerability it discovered and claims it fixed by September 2019, which led to the leak of 533M accounts now, suggests it should face a higher sanction from the DPC than Twitter received.

However even if Facebook ends up with a more substantial GDPR penalty for this breach the watchdog’s caseload backlog and plodding procedural pace makes it hard to envisage a swift resolution to an investigation that’s only a few days old.

Judging by past performance it’ll be years before the DPC decides on this 2019 Facebook leak — which likely explains why the DRI sees value in instigating class-action style litigation in parallel to the regulatory investigation.

“Compensation is not the only thing that makes this mass action worth joining. It is important to send a message to large data controllers that they must comply with the law and that there is a cost to them if they do not,” DRI writes on its website.

It also submitted a complaint about the Facebook breach to the DPC earlier this month, writing then that it was “also consulting with its legal advisors on other options including a mass action for damages in the Irish Courts”.

It’s clear that the GDPR enforcement gap is creating a growing opportunity for litigation funders to step in in Europe and take a punt on suing for data-related compensation damages — with a number of other mass actions announced last year.

In the case of DRI its focus is evidently on seeking to ensure that digital rights are upheld. But it told RTE that it believes compensation claims which force tech giants to pay money to users whose privacy rights have been violated is the best way to make them legally compliant.

Facebook, meanwhile, has sought to play down the breach it failed to disclose in 2019 — claiming it’s ‘old data’ — a deflection that ignores the fact that people’s dates of birth don’t change (nor do most people routinely change their mobile number or email address).

Plenty of the ‘old’ data exposed in this latest massive Facebook leak will be very handy for spammers and fraudsters to target Facebook users — and also now for litigators to target Facebook for data-related damages.

Continue Reading


Geoffrey Hinton has a hunch about what’s next for AI



Back in November, the computer scientist and cognitive psychologist Geoffrey Hinton had a hunch. After a half-century’s worth of attempts—some wildly successful—he’d arrived at another promising insight into how the brain works and how to replicate its circuitry in a computer.

“It’s my current best bet about how things fit together,” Hinton says from his home office in Toronto, where he’s been sequestered during the pandemic. If his bet pays off, it might spark the next generation of artificial neural networks—mathematical computing systems, loosely inspired by the brain’s neurons and synapses, that are at the core of today’s artificial intelligence. His “honest motivation,” as he puts it, is curiosity. But the practical motivation—and, ideally, the consequence—is more reliable and more trustworthy AI.

A Google engineering fellow and cofounder of the Vector Institute for Artificial Intelligence, Hinton wrote up his hunch in fits and starts, and at the end of February announced via Twitter that he’d posted a 44-page paper on the arXiv preprint server. He began with a disclaimer: “This paper does not describe a working system,” he wrote. Rather, it presents an “imaginary system.” He named it, “GLOM.” The term derives from “agglomerate” and the expression “glom together.”

Hinton thinks of GLOM as a way to model human perception in a machine—it offers a new way to process and represent visual information in a neural network. On a technical level, the guts of it involve a glomming together of similar vectors. Vectors are fundamental to neural networks—a vector is an array of numbers that encodes information. The simplest example is the xyz coordinates of a point—three numbers that indicate where the point is in three-dimensional space. A six-dimensional vector contains three more pieces of information—maybe the red-green-blue values for the point’s color. In a neural net, vectors in hundreds or thousands of dimensions represent entire images or words. And dealing in yet higher dimensions, Hinton believes that what goes on in our brains involves “big vectors of neural activity.”

By way of analogy, Hinton likens his glomming together of similar vectors to the dynamic of an echo chamber—the amplification of similar beliefs. “An echo chamber is a complete disaster for politics and society, but for neural nets it’s a great thing,” Hinton says. The notion of echo chambers mapped onto neural networks he calls “islands of identical vectors,” or more colloquially, “islands of agreement”—when vectors agree about the nature of their information, they point in the same direction.

“If neural nets were more like people, at least they can go wrong the same ways as people do, and so we’ll get some insight into what might confuse them.”

Geoffrey Hinton

In spirit, GLOM also gets at the elusive goal of modelling intuition—Hinton thinks of intuition as crucial to perception. He defines intuition as our ability to effortlessly make analogies. From childhood through the course of our lives, we make sense of the world by using analogical reasoning, mapping similarities from one object or idea or concept to another—or, as Hinton puts it, one big vector to another. “Similarities of big vectors explain how neural networks do intuitive analogical reasoning,” he says. More broadly, intuition captures that ineffable way a human brain generates insight. Hinton himself works very intuitively—scientifically, he is guided by intuition and the tool of analogy making. And his theory of how the brain works is all about intuition. “I’m very consistent,” he says.

Hinton hopes GLOM might be one of several breakthroughs that he reckons are needed before AI is capable of truly nimble problem solving—the kind of human-like thinking that would allow a system to make sense of things never before encountered; to draw upon similarities from past experiences, play around with ideas, generalize, extrapolate, understand. “If neural nets were more like people,” he says, “at least they can go wrong the same ways as people do, and so we’ll get some insight into what might confuse them.”

For the time being, however, GLOM itself is only an intuition—it’s “vaporware,” says Hinton. And he acknowledges that as an acronym nicely matches, “Geoff’s Last Original Model.” It is, at the very least, his latest.

Outside the box

Hinton’s devotion to artificial neural networks (a mid-2oth century invention) dates to the early 1970s. By 1986 he’d made considerable progress: whereas initially nets comprised only a couple of neuron layers, input and output, Hinton and collaborators came up with a technique for a deeper, multilayered network. But it took 26 years before computing power and data capacity caught up and capitalized on the deep architecture.

In 2012, Hinton gained fame and wealth from a deep learning breakthrough. With two students, he implemented a multilayered neural network that was trained to recognize objects in massive image data sets. The neural net learned to iteratively improve at classifying and identifying various objects—for instance, a mite, a mushroom, a motor scooter, a Madagascar cat. And it performed with unexpectedly spectacular accuracy.

Deep learning set off the latest AI revolution, transforming computer vision and the field as a whole. Hinton believes deep learning should be almost all that’s needed to fully replicate human intelligence.

But despite rapid progress, there are still major challenges. Expose a neural net to an unfamiliar data set or a foreign environment, and it reveals itself to be brittle and inflexible. Self-driving cars and essay-writing language generators impress, but things can go awry. AI visual systems can be easily confused: a coffee mug recognized from the side would be an unknown from above if the system had not been trained on that view; and with the manipulation of a few pixels, a panda can be mistaken for an ostrich, or even a school bus.

GLOM addresses two of the most difficult problems for visual perception systems: understanding a whole scene in terms of objects and their natural parts; and recognizing objects when seen from a new viewpoint.(GLOM’s focus is on vision, but Hinton expects the idea could be applied to language as well.)

An object such as Hinton’s face, for instance, is made up of his lively if dog-tired eyes (too many people asking questions; too little sleep), his mouth and ears, and a prominent nose, all topped by a not-too-untidy tousle of mostly gray. And given his nose, he is easily recognized even on first sight in profile view.

Both of these factors—the part-whole relationship and the viewpoint—are, from Hinton’s perspective, crucial to how humans do vision. “If GLOM ever works,” he says, “it’s going to do perception in a way that’s much more human-like than current neural nets.”

Grouping parts into wholes, however, can be a hard problem for computers, since parts are sometimes ambiguous. A circle could be an eye, or a doughnut, or a wheel. As Hinton explains it, the first generation of AI vision systems tried to recognize objects by relying mostly on the geometry of the part-whole-relationship—the spatial orientation among the parts and between the parts and the whole. The second generation instead relied mostly on deep learning—letting the neural net train on large amounts of data. With GLOM, Hinton combines the best aspects of both approaches.

“There’s a certain intellectual humility that I like about it,” says Gary Marcus, founder and CEO of Robust.AI and a well-known critic of the heavy reliance on deep learning. Marcus admires Hinton’s willingness to challenge something that brought him fame, to admit it’s not quite working. “It’s brave,” he says. “And it’s a great corrective to say, ‘I’m trying to think outside the box.’”

The GLOM architecture

In crafting GLOM, Hinton tried to model some of the mental shortcuts—intuitive strategies, or heuristics—that people use in making sense of the world. “GLOM, and indeed much of Geoff’s work, is about looking at heuristics that people seem to have, building neural nets that could themselves have those heuristics, and then showing that the nets do better at vision as a result,” says Nick Frosst, a computer scientist at a language startup in Toronto who worked with Hinton at Google Brain.

With visual perception, one strategy is to parse parts of an object—such as different facial features—and thereby understand the whole. If you see a certain nose, you might recognize it as part of Hinton’s face; it’s a part-whole hierarchy. To build a better vision system, Hinton says, “I have a strong intuition that we need to use part-whole hierarchies.” Human brains understand this part-whole composition by creating what’s called a “parse tree”—a branching diagram demonstrating the hierarchical relationship between the whole, its parts and subparts. The face itself is at the top of the tree, and the component eyes, nose, ears, and mouth form the branches below.

One of Hinton’s main goals with GLOM is to replicate the parse tree in a neural net—this is would distinguish it from neural nets that came before. For technical reasons, it’s hard to do. “It’s difficult because each individual image would be parsed by a person into a unique parse tree, so we would want a neural net to do the same,” says Frosst. “It’s hard to get something with a static architecture—a neural net—to take on a new structure—a parse tree—for each new image it sees.” Hinton has made various attempts. GLOM is a major revision of his previous attempt in 2017, combined with other related advances in the field.

“I’m part of a nose!”

GLOM vector

Hinton face grid


A generalized way of thinking about the GLOM architecture is as follows: The image of interest (say, a photograph of Hinton’s face) is divided into a grid. Each region of the grid is a “location” on the image—one location might contain the iris of an eye, while another might contain the tip of his nose. For each location in the net there are about five layers, or levels. And level by level, the system makes a prediction, with a vector representing the content or information. At a level near the bottom, the vector representing the tip-of-the-nose location might predict: “I’m part of a nose!” And at the next level up, in building a more coherent representation of what it’s seeing, the vector might predict: “I’m part of a face at side-angle view!”

But then the question is, do neighboring vectors at the same level agree? When in agreement, vectors point in the same direction, toward the same conclusion: “Yes, we both belong to the same nose.” Or further up the parse tree. “Yes, we both belong to the same face.”

Seeking consensus about the nature of an object—about what precisely the object is, ultimately—GLOM’s vectors iteratively, location-by-location and layer-upon-layer, average with neighbouring vectors beside, as well as predicted vectors from levels above and below.

However, the net doesn’t “willy-nilly average” with just anything nearby, says Hinton. It averages selectively, with neighboring predictions that display similarities. “This is kind of well-known in America, this is called an echo chamber,” he says. “What you do is you only accept opinions from people who already agree with you; and then what happens is that you get an echo chamber where a whole bunch of people have exactly the same opinion. GLOM actually uses that in a constructive way.” The analogous phenomenon in Hinton’s system is those “islands of agreement.”

“Geoff is a highly unusual thinker…”

Sue Becker

“Imagine a bunch of people in a room, shouting slight variations of the same idea,” says Frosst—or imagine those people as vectors pointing in slight variations of the same direction. “They would, after a while, converge on the one idea, and they would all feel it stronger, because they had it confirmed by the other people around them.” That’s how GLOM’s vectors reinforce and amplify their collective predictions about an image.

GLOM uses these islands of agreeing vectors to accomplish the trick of representing a parse tree in a neural net. Whereas some recent neural nets use agreement among vectors for activation, GLOM uses agreement for representation—building up representations of things within the net. For instance, when several vectors agree that they all represent part of the nose, their small cluster of agreement collectively represents the nose in the net’s parse tree for the face. Another smallish cluster of agreeing vectors might represent the mouth in the parse tree; and the big cluster at the top of the tree would represent the emergent conclusion that the image as a whole is Hinton’s face. “The way the parse tree is represented here,” Hinton explains, “is that at the object level you have a big island; the parts of the object are smaller islands; the subparts are even smaller islands, and so on.”

Figure 2 from Hinton’s GLOM paper. The islands of identical vectors (arrows of the same color) at the various levels represent a parse tree.

According to Hinton’s long-time friend and collaborator Yoshua Bengio, a computer scientist at the University of Montreal, if GLOM manages to solve the engineering challenge of representing a parse tree in a neural net, it would be a feat—it would be important for making neural nets work properly. “Geoff has produced amazingly powerful intuitions many times in his career, many of which have proven right,” Bengio says. “Hence, I pay attention to them, especially when he feels as strongly about them as he does about GLOM.”

The strength of Hinton’s conviction is rooted not only in the echo chamber analogy, but also in mathematical and biological analogies that inspired and justified some of the design decisions in GLOM’s novel engineering.

“Geoff is a highly unusual thinker in that he is able to draw upon complex mathematical concepts and integrate them with biological constraints to develop theories,” says Sue Becker, a former student of Hinton’s, now a computational cognitive neuroscientist at McMaster University. “Researchers who are more narrowly focused on either the mathematical theory or the neurobiology are much less likely to solve the infinitely compelling puzzle of how both machines and humans might learn and think.”

Turning philosophy into engineering

So far, Hinton’s new idea has been well received, especially in some of the world’s greatest echo chambers. “On Twitter, I got a lot of likes,” he says. And a YouTube tutorial laid claim to the term “MeGLOMania.”

Hinton is the first to admit that at present GLOM is little more than philosophical musing (he spent a year as a philosophy undergrad before switching to experimental psychology). “If an idea sounds good in philosophy, it is good,” he says. “How would you ever have a philosophical idea that just sounds like rubbish, but actually turns out to be true? That wouldn’t pass as a philosophical idea.” Science, by comparison, is “full of things that sound like complete rubbish” but turn out to work remarkably well—for example, neural nets, he says.

GLOM is designed to sound philosophically plausible. But will it work?

Chris Williams, a professor of machine learning in the School of Informatics at the University of Edinburgh, expects that GLOM might well spawn great innovations. However, he says, “the thing that distinguishes AI from philosophy is that we can use computers to test such theories.” It’s possible that a flaw in the idea might be exposed—perhaps also repaired—by such experiments, he says. “At the moment I don’t think we have enough evidence to assess the real significance of the idea, although I believe it has a lot of promise.”

The GLOM test model inputs are ten ellipses that form a sheep or a face.

Some of Hinton’s colleagues at Google Research in Toronto are in the very early stages of investigating GLOM experimentally. Laura Culp, a software engineer who implements novel neural net architectures, is using a computer simulation to test whether GLOM can produce Hinton’s islands of agreement in understanding parts and wholes of an object, even when the input parts are ambiguous. In the experiments, the parts are 10 ellipses, ovals of varying sizes, that can be arranged to form either a face or a sheep.

With random inputs of one ellipse or another, the model should be able to make predictions, Culp says, and “deal with the uncertainty of whether or not the ellipse is part of a face or a sheep, and whether it is the leg of a sheep, or the head of a sheep.” Confronted with any perturbations, the model should be able to correct itself as well. A next step is establishing a baseline, indicating whether a standard deep-learning neural net would get befuddled by such a task. As yet, GLOM is highly supervised—Culp creates and labels the data, prompting and pressuring the model to find correct predictions and succeed over time. (The unsupervised version is named GLUM—“It’s a joke,” Hinton says.)

At this preliminary state, it’s too soon to draw any big conclusions. Culp is waiting for more numbers. Hinton is already impressed nonetheless. “A simple version of GLOM can look at 10 ellipses and see a face and a sheep based on the spatial relationships between the ellipses,” he says. “This is tricky, because an individual ellipse conveys nothing about which type of object it belongs to or which part of that object it is.”

And overall, Hinton is happy with the feedback. “I just wanted to put it out there for the community, so anybody who likes can try it out,” he says. “Or try some sub-combination of these ideas. And then that will turn philosophy into science.”

Continue Reading


Pakistan temporarily blocks social media



Pakistan has temporarily blocked several social media services in the South Asian nation, according to users and a government-issued notice reviewed by TechCrunch.

In an order titled “Complete Blocking of Social Media Platforms,” the Pakistani government ordered Pakistan Telecommunication Authority to block social media platforms including Twitter, Facebook, WhatsApp, YouTube, and Telegram from 11am to 3pm local time (06.00am to 10.00am GMT) Friday.

The move comes as Pakistan looks to crackdown against a violent terrorist group and prevent troublemakers from disrupting Friday prayers congregations following days of violent protests.

Earlier this week Pakistan banned the Islamist group Tehrik-i-Labaik Pakistan after arresting its leader, which prompted protests, according to local media reports.

An entrepreneur based in Pakistan told TechCrunch that even though the order is supposed to expire at 3pm local time, similar past moves by the government suggests that the disruption will likely last for longer.

Though Pakistan, like its neighbor India, has temporarily cut phone calls access in the nation in the past, this is the first time Islamabad has issued a blanket ban on social media in the country.

Pakistan has explored ways to assume more control over content on digital services operating in the country in recent years. Some activists said the country was taking extreme measures without much explanations.

Continue Reading