How Bodega typifies Silicon Valley’s cultural ignorance

On Wednesday a story about two ex-Google employees receiving an obscene amount of money for a bad idea hit social media and was met with a level of outrage you could feel through the screen. If you’re online in any way whatsoever, you likely know I’m talking about Bodega.

The excellent article, Two Ex-Googlers Want To Make Bodegas And Mom-And-Pop Corner Stores Obsolete, hit several raw nerves with a wide range of people.

This fury is so crystallized because “Bodega” — an overfunded, probably doomed, glorified vending machine startup positioned as a bodega killer — stands for everything Silicon Valley represents to us. Whereas in reality, the very concept of a bodega stands for the absolute opposite of Silicon Valley.

It’s almost like someone said “Siri, show me why everyone hates and fears the things wearing human suits known as techies.”

Piles of money for trivial garbage

Bodega isn’t just an offensive idea, it’s an idea so bad and obviously worthless it’s maddening. Part of the visceral backlash was directed at the bourgeois wastefulness of the whole startup ecosystem, of which we are all angrily exhausted.

People who shop at the same Bay Area corner stores that Bodega wants to eliminate, like me, aren’t worried about any problem the startup wants to solve. We’re fretting about paying rent, affording health insurance, and the extreme gap between Bay Area’s rich and poor created by local tech companies that’s making the homeless problem a third world nightmare in our streets. A world in which Bodega gets a truckload of cash to almost literally burn right before our eyes.

Fast Company informs us,

About a year ago, McDonald and Rajan secured funding from notable investors to launch the concept, including Josh Kopelman at First Round Capital, Kirsten Green at Forerunner Ventures, and Hunter Walk at Homebrew. They also secured angel investment from senior executives at Facebook, Twitter, Dropbox, and Google.

According to TechCrunch, the startup’s first financing round was a cool $2.5 million.

No, instead of all the money that could help the Bay Area’s crushing homeless problem (and serve as models for other cities if done well), or help other worthy tech causes like Hack The Hood, we get Silicon Valley’s “best and brightest” reinventing the vending machine in the most dystopian, community-destroying way imaginable.

You can almost see the pitch meeting. One VC remarks to another, “You think that’s crazy? Hold my Juicero.”

Replacing community with soulless automation

Deeper outrage was directed at the hubris, ignorance, and privilege it takes to want to make a business out of replacing the cornerstones of community known on the East Coast and parts of Los Angeles as bodegas.

The idea of the Bodega product is to remove human contact from the neighborhood shopping equation, to do away with the actual bodega. “The vision here is much bigger than the box itself,” McDonald told Fast Company. “Eventually, centralized shopping locations won’t be necessary, because there will be 100,000 Bodegas spread out, with one always 100 feet away from you.”

Right after the Fast Company article came out, the Association for Neighborhood and Housing Development in New York issued a statement. “The awful irony of naming the company ‘Bodega’ after the very brick and mortar institutions they aim to displace, to say nothing about the cat their logo is based on that will similarly be displaced, is offensive, utterly misguided, and frankly disrespectful to New Yorkers,” it wrote.

Until the backlash hit, Bodega seemed fine with everything written about it killing corner stores and appropriating the bodega name. The Fast Company article that got all the attention was a really well-done piece and had confronted CEO and co-founder Paul McDonald. In it, Elizabeth Segran wrote, “I asked McDonald point-blank about whether he’s worried that the name Bodega might come off as culturally insensitive. Not really.” He told her in response, “I’m not particularly concerned about it.”

When it became clear that replacing bodegas with a box wasn’t a great selling point — nor was naming the company after the thing it seemed keen to undercut — McDonald shot out a backpedaling apology.

In contrast to his brush off about the issue of co-opting “bodega” to Fast Company, McDonald wrote:

When we first came up with the idea to call the company Bodega we recognized that there was a risk of it being interpreted as misappropriation. We did some homework — speaking to New Yorkers, branding people, and even running some survey work asking about the name and any potential offense it might cause. But it’s clear that we may not have been asking the right questions of the right people.

Regarding the headlines echoing McDonald’s quote about Bodega’s aim to eliminate the necessity of stores he wrote, “Challenging the urban corner store is not and has never been our goal.”

Then the clicks and whirrs of a robot that had only learned about corner stores from a machine learning algorithm kicked in as McDonald continued:

Corner stores have been fixtures of their neighborhoods for generations. They stock thousands of items, far more than we could ever fit on a few shelves. Their owners know what products to carry and in many cases who buys what. And they’re run by people who in addition to selling everything from toilet paper to milk also offer an integral human connection to their patrons that our automated storefronts never will.

Can you imagine handing your bank details to these clowns? Or trusting them to make the right decisions about audits and security, or your home address and third parties, or … any Google, Facebook, Uber — any major company’s products — that affect millions of users every day?

Maybe there was a point in time when we could, but I don’t remember it. Every day seems to bring a new terrifying (and insulting) reason to distrust Silicon Valley’s companies and eager little startups. Which is probably because, like with Bodega, they’re all founded, operated, funded, and secured by the same kinds of people.

That wacky “integral human connection”

I think we’re morbidly fascinated with who the people behind these tech companies are, the ones making huge decisions about our lives (and our security and privacy). They create and gatekeep the technologies that arbitrate our relationships and our communities. These are the people who are shaping our future, and yet we end up with over-engineered, pricey juice presses and vending machines.

Paul McDonald served as a product manager at Google for 13 years. His partner Ashwath Rajan was an associate product manager at the company for just over a year. In case you’re not sure what that means, a Google Product Manager shepherds to completion new products and features that impact the lives of millions of users every day.

McDonald and Rajan have been testing out their Bodega Boxes at 30 locations around the Bay Area since late last year, placing the bespoke toilet paper vending machines in what me must assume are crime-free (read: upper class) apartment lobbies and offices.

Bodega is not unique in any way. Remember SceneTap, the facial recognition startup for telling its app’s users how many women were in a bar at any given time? They tried to launch in San Francisco. To their surprise, but no one else’s, the backlash was huge.

I can’t imagine a solution for the Bodega problem, mostly because of the fact that it exists at all.

Here in the Bay Area, where these ex-Googlers got their pedigrees to formulate and pitch their startup, we don’t actually call them bodegas. We call them “corner stores.” This little detail is even more conspicuous when you consider that Paul McDonald says he has not lived in New York.

Which is why even just the startup’s name “Bodega” told us locals and natives — the involuntary first-wave recipients of Silicon Valley’s fucked up experiments with our lives — all we needed to know.

Images: Alain-Christian/Flickr (Bodega cat); Spencer Platt/Getty Images (An NYC bodega)

Google has targeted ads based on hate speech, too

Yesterday, ProPublica released a report on its investigation into the sorts of ad categories Facebook makes available to advertisers. It found that the website allowed it to target ads to users based on categories like “Jew hater” and “How to burn jews” among other antisemitic options. Today, BuzzFeed reports that Google has a similar problem.

In a comparable investigation, BuzzFeed News purchased and published an ad campaign through Google that targeted users who search with keywords like “why do jews ruin everything.” Further, when BuzzFeed reporters typed in those words into the ad-buying platform, Google suggested other phrases to use like “the evil jew” and “jewish parasites.” When the team tried phrases like “white people ruin,” Google’s suggestions included “black people destroy everything” and “black people ruin neighborhoods.” After the ads were purchased, they appeared on Google’s page when BuzzFeed staff searched for the phrases they had targeted. However, the ads that made it through only accrued 17 impressions before their keywords were disabled by Google after the company received screenshots from BuzzFeed alerting them to the ads.

Google has since disabled nearly all of the terms targeted by BuzzFeed’s ad campaign, save for “black people destroy everything,” which was still an eligible target when BuzzFeed published their post. But an eligible keyword doesn’t mean it’s an approved keyword, just that the ad targeting it is under review and has the possibility of being approved. Many of BuzzFeed’s ads weren’t approved in the end and when that was the case, Google’s message in response was, “We value diversity and respect for others, so we strive to avoid offending users with ads or promoted content that’s inappropriate for our ad network. Please remove any content that promotes hatred, intolerance, harassment, intimidation, exploitation, violence, or self-harm.”

While it appears that Google caught many of the ads targeting racist language, the fact that some got through means there’s most certainly room for improvement.

In a statement to Engadget, Google’s senior vice president of ads, Sridhar Ramaswamy, said, “Our goal is to prevent our keyword suggestions tool from making offensive suggestions, and to stop any offensive ads appearing. We have language that informs advertisers when their ads are offensive and therefore rejected. In this instance, ads didn’t run against the vast majority of these keywords, but we didn’t catch all these offensive suggestions. That’s not good enough and we’re not making excuses. We’ve already turned off these suggestions, and any ads that made it through, and will work harder to stop this from happening again.”

Images: Google / BuzzFeed News

Source: BuzzFeed

Senate bill calls for free credit freezes after Equifax breach

US Senator Elizabeth Warren and a handful of her Democratic peers have introduced a bill intended to give consumers more control over the information collected by credit-reporting agencies including Equifax, TransUnion and Experian. The Freedom From Equifax Exploitation Act is in response to a massive security breach at Equifax that compromised the personal information of 143 million people. Equifax reported the hack on September 7th, though the attack itself was live from mid-May through the end of July.

Above all, the FREE Act would allow consumers to freeze their credit at any time for free. This would prevent companies like Equifax from selling that information to other entities. The bill also expands fraud alert protections and would force Equifax to refund any fees it charged for credit freezes in the wake of the breach.

On top of the bill, Warren is spearheading an investigation into the Equifax hack and the wider credit-reporting industry. This is her wheelhouse — Warren helped establish the Consumer Financial Protection Bureau in 2011, following the 2007 financial crisis. In a letter to the CFPB and Federal Trade Commission today, Warren asked whether the bureau would benefit from additional authorities to properly oversee credit-reporting agencies.

The FTC is also investigating the incident itself.

Days after the Equifax breach, Democratic Senators Brian Schatz, Claire McCaskill, Richard Blumenthal, Bernie Sanders, Jeff Merkley and Warren reintroduced the Stop Errors in Credit Use and Reporting Act, which would introduce operational standards for the credit-reporting industry and give consumers more control over their information.

Hackers were able to access Equifax’s bank of names, addresses, birth dates, social security numbers and other private information because of a flaw in the Apache Struts Web Framework that was made public in March. Equifax didn’t properly patch the loophole — the hack began two months later and was allowed to linger through July.

It isn’t just the government making moves after the Equifax incident: Credit Karma this week announced it’s added Equifax to its free credit monitoring service.

Source: Elizabeth Warren

Facebook opens a new AI research lab in Montreal

Facebook’s Artificial Intelligence Research team is expanding. The company announced a brand new AI lab that just opened in Montreal, which joins the network of existing labs based in Menlo Park, New York City and Paris. “The Montreal lab will house research scientists and engineers working on a wide range of ambitious AI research projects, but it will also have a special focus on reinforcement learning and dialog systems,” Facebook’s chief AI scientist, Yann LeCun, said in a post.

The lab will be helmed by Joelle Pineau, a computer science professor at McGill University interested in learning and the co-director of the university’s Reasoning and Learning Lab. Her research has focused on planning, learning and decision-making as well as human-robot interactions. She has published a number of studies on dialog systems and worked on a team that developed an intelligent robotic wheelchair.

LeCun says that, like other Facebook AI research labs, the Montreal site will engage with research communities via publications, conferences and collaborations. Facebook will also partner with McGill University as well as the Canadian Institute for Advanced Research, the Montreal Institute for Learning Algorithms and Université de Montréal.

“Montreal already has an existing fantastic academic AI community, an exciting ecosystem of startups, and promising government policies to encourage AI research. We are excited to become part of this larger community, and we look forward to engaging with the entire ecosystem and helping it continue to thrive,” said LeCun.

Source: Facebook

Twitter also has a problem with ads targeted towards hate speech

ProPublica found that Facebook allows ads to be targeted at users based on antisemitic keywords and BuzzFeed has reported that Google similarly allows ads to be targeted through racist and hateful phrases. So, it probably shouldn’t be a surprise at this point that it turns out Twitter is laden with similar issues.

The Daily Beast reports today that its staff was successful in launching ad campaigns on Twitter using terms like “wetback” and “Nazi” to target particular audiences. When purchasing ads, Twitter’s platform asks advertisers to select keywords that aid in helping the site place ads in front of relevant users. When Daily Beast reporters typed “wetback” into this feature, Twitter informed them that 26.3 million users might respond to that term. Similarly, it said 18.6 million accounts would probably be interested in the term “Nazi.” After confirming that their ad campaigns were successful, The Daily Beast terminated each of them.

Twitter’s ad policy notes that it prohibits the promotion of hate content and says that advertisers are responsible for their own ads. And while in its policies for keyword targeting, Twitter states that racial or ethnic origin categories can’t be included, hate speech isn’t specifically mentioned.

We’ve reached out to Twitter for a comment and we’ll update this post when we receive more information.

Source: The Daily Beast

Study says Apple data-mining safeguards don’t protect privacy enough

During last year’s WWDC in June 2016, Apple noted it would be adopting some degree of differential privacy methods to ensure privacy while the company mined user data on iOS and Mac OS. In short, the technique adds noise to data that scrambles it enough to prevent it from becoming identifiable — though the company made clear at the time that its data collection process was opt-in. Over a year later, a study claims that Apple’s methods fall short of the digital privacy community’s expectations for how much a user’s data is kept private.

As they reveal in their study (PDF), researchers from the University of Southern California, Indiana University and China’s Tsinghua University evaluated how Apple injects static into users’ identifiable info, from messages to your internet history, to baffle anyone looking at the data, from the government to Apple’s own staff. The metric for measuring a setup’s differential privacy effectiveness is called a “privacy loss parameter” or, as a variable, “epsilon.” In this case, the researchers discovered that Apple’s epsilon on MacOS allowed a lot more personal data to be identifiable than digital privacy theorists are comfortable with, and iOS 10 permits even more.

Apple has refuted the study’s findings, especially on its alleged ability to link data to particular users. But Apple still hasn’t released much information on how it specifically implements its differential privacy. As Wired points out, the most unsettling part is that Apple keeps its epsilon numbers secret, meaning it could change the amount of privacy-enabling static conceivably at any time.

We reached out to Apple for additional comment and will add when we hear back.

Via: Wired

Source: “Privacy Loss in Apple’s Implementation of Differential Privacy on MacOS 10.12” (PDF)

Apple bans misleading apps on iOS

As soon as you get your shiny new iPhone 8, you’ll probably want to download some apps. So Apple took the opportunity to update its developer app review guidelines to make sure the App Store is in tip top shape. According to 9to5 Mac, it’s now officially cracking down on misleading apps.

The App Store has already had policies against such apps already, but the recent update makes it more explicit. The ban is against apps “including content or services that it does not actually offer.” As those fake antivirus apps don’t actually work, they’re definitely on this list. Apple has already removed a number of these apps from the App Store, but this new guideline makes the ban explicit so that they can’t get on in the first place.

In addition, the new guidelines also include clauses for ARKit and Face ID. Apple wants the ARKit apps you see in the App Store to be more than just one-note apps with single objects, adding that it needs to offer “rich and integrated augmented reality experiences.” As for Face ID, Apple added that it must offer an alternate unlocking method for children under the age of 13.

It’ll also now be possible for apps to offer 100 percent of funds to be gifted from one person to another without using In-App Purchases, thus bypassing the need to give 30 percent of it to Apple. The caveat is that giving said money must be an optional choice.

Last but not least, there is also a new clause that explicitly prohibits apps that “facilitate human trafficking and/or the exploitation of children.” They’ve always been illegal anyway, but Apple is simply adding language to make it more explicit that it will take legal action if it finds such apps.

Via: The Verge

Source: 9to5Mac

Apple Watch Series 3 Limited to 3G-Like Speeds on T-Mobile

Apple’s new Apple Watch Series 3 models support LTE and UMTS connectivity, but on T-Mobile, LTE connection speeds won’t be available. The carrier is limiting the Apple Watch to a maximum speed of 512kb/s, which is more like a 3G connection than an LTE connection.

The speed limitation is noted in T-Mobile’s fine print, which says the maximum wearables speed is 512kb/s. It’s also noted on the pre-order page when adding an Apple Watch Series 3 to your cart.

At issue is T-Mobile’s ONE plan, which limits both tethering and all wearable devices to 512kb/s. While most users are unlikely to be streaming video or doing other tasks that require high connectivity speeds on the Apple Watch, the Apple Watch is also unlikely to be a data hog, so it’s unclear why T-Mobile has this limitation in place.

A T-Mobile representative told MacRumors reader Tony that its “High Speed Data with paired DIGITS” plan would provide 4G LTE data. DIGITS is priced at $25 per month without autopay, and $20 per month with Autopay.

No other major carrier in the United States, including Sprint, Verizon, or AT&T appears to be throttling Apple Watch speeds with their standard unlimited plans.

All four of the carriers are charging the same $10 per month fee to add an Apple Watch to an existing iPhone plan, and each one is offering a limited time promotion that waives monthly fees for the first three months.

The LTE Apple Watch Series 3 models became available for pre-order early this morning, with new devices slated to arrive on Friday, September 22, the official launch date for the device.

Related Roundups: Apple Watch, watchOS 3, watchOS 4
Tag: T-Mobile
Buyer’s Guide: Apple Watch (Buy Now)
Discuss this article in our forums

MacRumors Giveaway: Win a 22,000mAh Battery Pack from RAVPower

For this week’s giveaway, we’ve teamed up with RAVPower to offer MacRumors readers a chance to win a 22,000mAh Power Bank for charging iPhones, iPads, the Apple Watch, and more.

In addition to offering several Power Banks for us to give away, RAVPower has also provided MacRumors with an exclusive 20 percent off discount code on the power bank. When ordering from Amazon, use promo code MacRumo1 to get it for just over $30.

RAVPower is mainly known for a whole line of power banks that are known for being reliable, affordable, and fast.

The 22,000mAh Power Bank features three 2.4A USB-A ports that are capable of charging three devices at the same time. At 2.4A, an iPhone will charge more quickly than with a standard 1A power adapter. There’s a total current output of 5.8A, and RAVPower says its Power Bank detects the connected devices and adjusts output accordingly to provide adequate power to each one, and there’s an “iSmart 2.0” feature that adjusts voltage to minimize charging time.

RAVPower says that the Power Bank can be fully recharged in approximately 11 to 12 hours using a the 2.4A input. It comes with microUSB charging cables and a carry pouch, but to charge iOS devices, you’ll need to supply your own cables.

Power banks are generally built for functionality over looks, and the RAVPower model is a simple black (or white) rectangular power bank that is about as tall as an iPhone 7 Plus and as wide as an iPhone 7. It measures in at 6.5 inches by 2.8 inches. It’s not quite pocketable, but ideal for a purse or a backpack. Four LEDs on the front let you see power level at a glance.

22,000mAh is enough to charge an iPhone 7 to full approximately eight times and an iPhone 7 Plus approximately four and a half times.

RAVPower normally charges $40 for the Power Bank, but we have six to give away to MacRumors readers. To enter to win, use the Rafflecopter widget below and enter an email address. Email addresses will be used solely for contact purposes to reach the winner and send the prize. You can earn additional entries by subscribing to our weekly newsletter, subscribing to our YouTube channel, following us on Twitter, or visiting the MacRumors Facebook page.

Due to the complexities of international laws regarding giveaways, only U.S. residents who are 18 years of age or older are eligible to enter. To offer feedback or get more information on the giveaway restrictions, please refer to our Site Feedback section, as that is where discussion of the rules will be redirected.

a Rafflecopter giveawayThe contest will run from today (September 15) at 11:00 a.m. Pacific Time through 11:00 a.m. Pacific Time on September 22. The winners will be chosen randomly on September 22 and will be contacted by email. The winners will have 48 hours to respond and provide a shipping address before new winners are chosen.

Tag: giveaway
Discuss this article in our forums

Apple’s Chip Partner TSMC Shares Details on 7nm Node and Advanced InFO Package Process for 2018

At the Open Innovation Platform Ecosystem Forum in Santa Clara on Wednesday, chip foundry TSMC provided an update (via EE Times) on the progress of its forthcoming technology nodes, several of which would be candidates for upcoming Apple chips. Most notably, the company’s first 7-nanometer process node has already had several tape-outs (finalized designs) and expects to reach volume capacity in 2018.

TSMC’s 10 nm node, which first showed up in Apple’s A10X chip in the iPad Pro, followed by the A11, has been fraught with issues (paid link) such as low chip yield and performance short of initial expectations. TSMC looks to change its fortune with the new 7 nm node, which would be suitable for the successor to the A11 chip given current timelines.

In addition to the 7 nm node, TSMC also shared information on the follow-up revision to this node, dubbed, N7+. Featuring the long-beleaguered Extreme Ultraviolet Lithography (EUV), the revision would promise 20 percent better density, around 10 percent higher speeds, or 15 percent lower power with other factors held constant.

While EUV has faced delays for over a decade at this point, it seems to finally be coming to fruition, and a 2019 volume availability update would allow Apple to update its chip process in subsequent years yet again. Apple had previously updated process nodes with every iPhone since the transition to 3GS before being forced to use TSMC’s 16 nm node in consecutive years with the A9 and A10. Moving forward, that annual cadence is again in jeopardy as chip foundries deal with the realities of physics and minimum transistor geometry sizes.

TSMC also unveiled some low power and low leakage processes that are suitable for Apple’s other custom designs, such as its line of wireless chips like the W1 and successor W2. TSMC is targeting availability next year of a 22 nm ultra low leakage node, which is suitable for analog and RF designs such as cellular basebands or Wi-Fi chips.

This will ultimately help Apple further lower power consumption on the Apple Watch and headphones featuring the W line of wireless chips. It is also likely to be adopted by Qualcomm for its line of modem products. The W1 and W2 manufacturing processes are not currently publicly known, but it is likely that one of TSMC’s RF-focused processes powers the Apple chips.

Finally, TSMC announced a revision of its integrated fan-out packaging process (InFO) that is targeted at integrating high bandwidth memory (HBM) into the assembly, dubbed InFO-MS. HBM has generated a lot of interest from applications where very high sustained memory bandwidths are desired, such as consumer graphics cards.

HBM and similar standards such as Wide I/O promise not only to improve memory bandwidth, but also improve power consumption for a given bandwidth, making it a suitable evolution for mobile SoC designs. This type of memory interface has yet to appear in a mobile design, though it should be considered a near-term eventuality. Despite advances in mobile memory, it still lags behind desktop and laptop systems in total bandwidth, which can be important in some tasks such as graphics rendering.

Tags: TSMC, W1, A12, W2
Discuss this article in our forums