Start Up: Facebook’s worried about you, Apple’s adblocking bites, Volvo’s self-driving detour, and more

Some Americans are receiving packages from China they didn’t order. Why? An e-commerce scam. Photo by Crouching Donkey on Flickr.

Charity time: ahead of Christmas, I’m encouraging readers to make a donation to charity; a different one each day. Today’s is:
BookTrust: give £10 and a child in social care will receive books for Christmas.

A selection of 14 links for you. Got your tree yet? I’m @charlesarthur on Twitter. Observations and links welcome.

Facebook conceded it might make you feel bad. Here’s how to interpret that • The New York Times

Farhad Manjoo:


The blogpost pointed out several recent and coming changes to Facebook that the company said encouraged active interactions on the service. That’s the real message: Once you discover how much more you can get out of Facebook with this new stuff, you’ll feel super.

O.K., sure, the post can be read this way. But I’m more optimistic about it, because it’s in line with an evolving corporate posture from the company.

After initially dismissing Facebook’s role in the 2016 election, Mark Zuckerberg, Facebook’s co-founder and chief executive, has spent much of the last year publicly grappling with Facebook’s role in the world. He published a lengthy letter to Facebook’s community attempting to establish new social goals for the company. He apologized for glibly dismissing the idea that Facebook could have altered the outcome of the election. And in the company’s last earnings report to investors, he said he was willing to risk the company’s profitability to improve its community.

To be sure, Facebook is putting its own favorable spin on these studies. Yet its willingness to shine a light on critical research, and its pledge to take the findings into account when it designs its products, has to be welcomed as something new.


link to this extract

Silicon Valley techies still think they’re the good guys. They’re not • WIRED

Erin Griffith:


Talking to tech founders every day, it’s clear how little their lives have changed in the last year, even as the world around them has shifted. Even top bosses who’ve noticed the change in public opinion aren’t willing to adjust. On his blog, Y Combinator president Sam Altman argued that political correctness was damaging the tech industry. “This is uncomfortable, but it’s possible we have to allow people to say disparaging things about gay people if we want them to be able to say novel things about physics,” he wrote. On the ground, the startup kings haven’t changed their behavior. They’re still pitching me their companies with the same all-out exuberance. They’re continuing their quest to move fast and break things—regardless of what broken objects are left in their wake.

Outside the bubble, things are different. We’re not egging on startups that willingly flaunt regulations. We’re wary of artificial intelligence and its potential to eliminate jobs. We’re dubious of tech leaders’ promises to make their products safe for their kids to use. We are all sick of the jokes that no longer feel funny: lines about the lack of women in tech, about obscenely rich 20-somethings, about awkward coders with bad people skills, about “hustling” and growth at any cost. It all feels inappropriate.

But this backlash against tech is difficult to see from inside the Silicon Valley bubble. And it’s not hard to understand how we got here. In the late 2000s, just after the financial crisis, the world was eager to hear positive stories about tech. The fast rise of services like Twitter and Facebook was thrilling—a spot of optimism in the gloomy aughts—and their geek genius founders made better heroes than the greedy Wall Street jerks that had just tanked the economy.


link to this extract

Google Tango augmented reality project is shutting down • Business Insider

Matt Weiberger:


Google has announced that it will be “turning down” Tango, its ambitious augmented-reality project for redefining what a smartphone camera can do, in March 2018.

First launched in 2014, Tango (formerly Project Tango) was a design for a camera that could actually detect depth and motion, opening up all kinds of new applications. The problem, and Tango’s biggest obstacle to success, was that you needed a special, high-tech, Tango-compatible camera to take advantage — a normal camera just wouldn’t do.


A formality – Google has shifted to doing this on any (capable) smartphone with ARCore. The losers are Lenovo and Asus, which put time and (perhaps?) money into building Project Tango hardware which absolutely nobody bought. Unless Google subsidised it, which I would have thought the OEMs might have asked for.
link to this extract

March 2017: Maybe Android tablet apps will be better this year • The Verge

Dieter Bohn, back in March:


There’s a new Android tablet you can go and buy, the Samsung Galaxy Tab S3. Here’s our review of it, where Jake notes that apps freeze if they’re not in the foreground. Which is a good reminder: Android apps on tablets have never really been very good. They usually end up feeling like stretched-out phone apps.

Things have gotten better in the past couple years, but it’s still a problem. In fact, it has always been a problem. I wonder if anybody ever told Google that it was a problem and it should try to do a better job incentivizing developers to make apps that work better on tablets.

Oh, wait, somebody has.


There follows a list of times when it’s been pointed out since 2011 that Android’s tablet apps really could do better. One concludes this isn’t going to happen, if it hasn’t happened in six-plus years.
link to this extract

Criteo tumbles on impact of latest iPhone web ad limiting tech • Bloomberg

Mark Gurman:


Web advertising company Criteo SA dropped the most in more than two years after announcing that its previous forecast underestimated the effect of Apple Inc.’s latest iPhone software update on its business.

In September, Apple released new versions of its iOS and macOS operating systems, the software that runs on iPhones, iPads, and Macs, that included a feature called Intelligent Tracking Prevention in the Safari web browser. The feature limits advertisers’ abilities to collect data on users such as the websites they visit. Criteo, and other ad targeting companies, collect user data in order to show people ads more relevant to their interests.

Criteo said Nov. 1 it anticipated Apple’s changes to have a 9% to 13% net-negative impact on its 2018 revenue, but now expects the impact to be about 22% net-negative. Apple’s iOS 11.2 software update, released earlier this month, “disables the solution that some companies in the advertising ecosystem, including Criteo, currently use to reach Safari users,” Paris-based Criteo said Thursday in a statement. However, the security changes began rolling out in September with the release of iOS 11 and macOS High Sierra.


That’s really pretty big; suggests that Criteo must be very reliant on the US market.

link to this extract

Scientists link Hurricane Harvey’s record rainfall to climate change • The New York Times

Henry Fountain:


Two research groups found that the record rainfall as Harvey stalled over Texas in late August, which totaled more than 50 inches in some areas, was as much as 38% higher than would be expected in a world that was not warming.

While many scientists had said at the time that Harvey was probably affected by climate change, because warmer air holds more moisture, the size of the increase surprised some.

“The amount of precipitation increase is worse than I expected,” said Michael J. Wehner, a senior scientist at Lawrence Berkeley National Laboratory and an author of a paper on his group’s findings, which included the 38% figure. Based on how much the world has warmed, Dr. Wehner said, before the analysis he had expected an increase of only about 6 or 7%.

The other study, by an international coalition of scientists known as World Weather Attribution, found that Harvey’s rainfall was 15% higher than would be expected without climate change. Geert Jan van Oldenborgh of the Royal Netherlands Meteorological Institute and the lead author of the second study, said that climate change also made such an extreme rainstorm much more likely.


Important finding – this sort of determinism is unusual. (But did it have to be Henry Fountain?)
link to this extract

AI can be a tough sell in the enterprise, despite potential • WSJ


Artificial intelligence and machine learning tools are expected to boost productivity across all industries in the years ahead. Yet, as many early-stage applications falter, they can run into resistance in the workplace, from the shop floor to the executive suite.

Take Monsanto Co., which expects a vast majority of its early AI and deep-learning projects to fail, says Anju Gupta, the agricultural giant’s director of digital partnerships and outreach.

A 99% failure rate with a current slate of 50-plus deep-learning projects is acceptable because “that 1% is going to bring exponential gain,” Ms. Gupta told a crowd of enterprise IT managers gathered here at an AI industry conference.

The stakes are high, according to Heath Terry, a managing director at Goldman Sachs Group Inc. It estimates that AI-enabled processes will result in up to $20bn in annual savings in the agricultural sector alone, he said.

Across the board, Goldman Sachs expects AI to add between 51 to 154 basis points to U.S. productivity by 2025, the most significant boost in productivity in decades, Mr. Terry said. Already, he adds, 13% of S&P 500 firms have mentioned AI in earnings calls, as of the second quarter, while venture capital funding for AI has doubled this year to more than $10bn.

Still, failures in early tests can risk creating a backlash to AI deployments across a company, despite the potential gains, Ms. Gupta said.


link to this extract

Amazon’s fake review problem • Brian Theodore Bien

Bien was looking for a sunrise alarm clock:


Here, we see both the top and bottom review with the sentence,

The light can be pretty bright, you can adjust it where it’ll be dim and slowly brighten 30 minutes before the alarm time.

Did “Becky” and “Dione Milton” really both happen to write a review with the exact same 23-word sentence? Or, is it more likely that they are agents sourcing reviews from a script, and they sloppily pasted their reviews without rewriting them (as they were presumably instructed to do)? Note also the post dates: December 12, 2017. “Becky” and “Dione Milton” both had private profiles, where their 5-6 reviews were hidden – very similar looking.

Amazon – who has some of the world’s most advanced ML – really needs to step up its review fraud detection game. Imagine how great the Amazon shopping experience would be if we could trust its reviews.

Third party meta review sites like Fakespot will identify problems for us (in this case, the product got an “F” grade) – so why doesn’t Amazon?


As one of the commenters says, Amazon has a problem just like Google had – has? – a spam problem back in 2010 or so.
link to this extract

Americans are receiving unordered parcels from Chinese e-criminals – and can’t do anything to stop them • Forbes

Wade Shepard on an e-commerce method called “brushing”:


Chinese agents shipping ridiculous amounts of hair ties to [Pennsylvania resident Heaven] McGeehan is merely an unscrupulous way for them to fraudulently boost sales and obtain positive feedback for their clients’ products on e-commerce sites.

Basically, a “brushing” firm somehow got hold of McGeehan’s name and address – she imagines this happened from placing legitimate orders on AliExpress, the international wing of China’s Alibaba – and then created user profiles for “her” on the e-commerce sites that they wish to have higher sales ratings and favorable reviews on. They then shop for orders via the fake account, compare prices, and mimic everything an actual customer would do, before finally making a purchase from their client’s store. When delivery is confirmed, they then leave positive reviews that appear to the e-commerce platform as “verified.”

The hair ties that McGeehan receives are more than likely not the actual items the Chinese brushers are leaving reviews for. Basically, they are low cost stand-ins for the real products. It doesn’t really matter what is shipped in the packages in this case, as the person receiving it has nothing to do with the exchange. But at least McGeehan is actually receiving packages that contain something. I’ve also been receiving reports from unsuspecting and often confused people in the U.S. whose mailboxes are being filled with parcels from China which contain nothing.

Due to the unbalanced pricing policies of the United Postal Union and subsidies from the U.S. Postal Service, it costs people in China virtually nothing to ship small packages to the U.S. That, combined with the super cheap price they pay for the junk they ship, makes brushing a quick and cost effective way to move up the sales rankings – which means everything for e-commerce merchants.


link to this extract

Volvo’s Drive Me takes detour on road to full autonomy • Automotive News

Douglas Bolduc:


Volvo’s Drive Me autonomous driving project is taking some detours compared with promises the automaker made when it announced the program four years ago, but Volvo says the changes will make its first Level 4 vehicle even better when it arrives in 2021.

In early announcements about Drive Me, Volvo promised to have 100 self-driving vehicles on the road but that has been downgraded. Volvo now says it will have 100 people involved in the Drive Me program within the next four years. Initially, the people taking part in Drive Me will test the cars with the same Level 2 semiautonomous assistance systems that are commercially available to anyone who purchases the vehicle in markets such as Europe and the U.S.

Drive Me is a public autonomous driving experiment that now includes families in Sweden and will be extended to London and China later. The goal is to provide Volvo with customer feedback for its first model with Level 4 autonomy, which means the car can drive itself but still has a steering wheel and pedals so that the driver can take control when needed.


“When needed”? I don’t like that phrase. How quickly might I be needed?
link to this extract

There never was a real tulip fever • Smithsonian

Lorraine Boissoneault:


According to popular legend, the tulip craze took hold of all levels of Dutch society in the 1630s. “The rage among the Dutch to possess them was so great that the ordinary industry of the country was neglected, and the population, even to its lowest dregs, embarked in the tulip trade,” wrote Scottish journalist Charles Mackay in his popular 1841 work Extraordinary Popular Delusions and the Madness of Crowds. According to this narrative, everyone from the wealthiest merchants to the poorest chimney sweeps jumped into the tulip fray, buying bulbs at high prices and selling them for even more. Companies formed just to deal with the tulip trade, which reached a fever pitch in late 1636. But by February 1637, the bottom fell out of the market. More and more people defaulted on their agreement to buy the tulips at the prices they’d promised, and the traders who had already made their payments were left in debt or bankrupted. At least that’s what has always been claimed.

In fact, “There weren’t that many people involved and the economic repercussions were pretty minor,” [professor of early modern history at Kings College London, Anne] Goldgar says. “I couldn’t find anybody that went bankrupt. If there had been really a wholesale destruction of the economy as the myth suggests, that would’ve been a much harder thing to face.”

That’s not to say that everything about the story is wrong; merchants really did engage in a frantic tulip trade, and they paid incredibly high prices for some bulbs. And when a number of buyers announced they couldn’t pay the high price previously agreed upon, the market did fall apart and cause a small crisis—but only because it undermined social expectations.


OK, a tulip flu then? But one can see parallels with bitcoin: there’s only a small number of people who own it, and yet the coverage of it is bonkers.
link to this extract

q2vq2 • Ghostbin

“Dr Cyborkian a.k.a. janit0r, conditioner of ‘terminally ill’ devices”:


I am now here to warn you that what I’ve done was only a temporary band- aid and it’s not going to be enough to save the Internet in the future.

The bad guys are getting more sophisticated, the number of potentially vulnerable devices keep increasing, and it’s only a matter of time before a large scale Internet-disrupting event will occur. If you are willing to believe that I’ve disabled over 10 million vulnerable devices over the 13-month span of the project then it’s not far-fetched to say that such a destructive event could’ve already happened in 2017.

YOU SHOULD WAKE UP TO THE FACT THAT THE INTERNET IS ONLY ONE OR TWO SERIOUS IOT EXPLOITS AWAY FROM BEING SEVERELY DISRUPTED. The damage of such an event is immeasurable given how digitally connected our societies have become, yet CERTs, ISPs and governments are not taking the gravity of the situation seriously enough.

ISPs keep deploying devices with exposed control ports and although these are trivially found using services like Shodan the national CERTs don’t seem to care. A lot of countries don’t even have CERTs. Many of the world’s biggest ISPs do not have any actual security know-how in-house, and are instead relying on foreign vendors for help in case anything goes wrong. I’ve watched large ISPs withering for months under conditioning from my botnet without them being able to fully mitigate the vulnerabilities (good examples are BSNL, Telkom ZA, PLDT, from time to time PT Telkom, and pretty much most large ISPs south of the border).


HE seems to be the author of “Brickerbot”, an IoT-attacking malware strain which just seems to wreck them. If history is a guide, he’s releasing the code for this (linked earlier in his post) because law enforcement is close enough that he’s about to be caught, so he wants deniability – he uploads the code somewhere and then downloads it, and denies he wrote it. (Paras Jha, who recently pleaded guilty with others to writing the Mirai IoT bot, did the same.)
link to this extract

25,000 children in Britain are problem gamblers, report finds • The Guardian

Rob Davies:


Fruit machines remain the most common introduction to gambling for young people at 24%, followed by the National Lottery at 21%.

But the Gambling Commission said children were increasingly being exposed to gambling in less traditional ways, such as through eSports (computer games competitions) and via social media.

The report found that 11% of children took part in skins betting, whereby online gamers can bet using in-game items, such as weapons or outfits, which can have real monetary value if traded.

Skins betting, an industry worth up to $5.1bn (£3.8bn) last year according to one US report, is a common feature of games such as Counter Strike: Global Offensive.

And earlier this year, two men were convicted for running a website that allowed children to bet on the Fifa series of online football games.

More than one in 10 children reported having played casino-style games, which simulate roulette or fruit machines, on Facebook or smartphone apps.

The commission’s statistics indicate that children who play such games, many of which have a PEGI (Pan European Game Information) 12 age rating, are more likely to gamble in real life.


A downward spiral, started early. There’s a lot of nonsense pleading by the companies that run the gambling machines, and gaming generally, about “jobs at risk”. Gambling like this quickly goes out of control, and puts livelihoods at risk.
link to this extract

The Amazon machine • Benedict Evans

Evans points out that a key to big companies is how well they’re able to make the things that make the things they offer; in Amazon’s case, it’s teams to sell stuff:


Amazon, then, is a machine to make a machine – it is a machine to make more Amazon. The opposite extreme might be Apple, which rather than radical decentralization looks more like an ASIC, with everything carefully structured and everyone in their box, which allows Apple to create certain kinds of new product with huge efficiency but makes it pretty hard to add new product lines indefinitely. Steve Jobs was fond of talking about saying ‘no’ to new projects – that’s not a very relevant virtue to Amazon.

For both Amazon and Apple (and indeed Google or Facebook), this means that there are certain kinds of project that they can deliver very well and very repeatably and predictably, but also, crucially, that there are certain kinds of project that they are much less well suited to deliver. Google doesn’t tend to be better at cloud platforms than Apple and worse at UIs because there are better or worse people in each team, but because each company is set up to deliver certain kinds of things, and the closer a project is to that machine’s direction the more reliable the result. If the machine is designed to do X, it will struggle at Y no matter how clever the people. A lot of the story of Amazon for the last 20 years is of how many Ys turned out to by Xs – how many categories that people thought could not be sold online and could not be sold as commodities turned out to be both.


link to this extract

Errata, corrigenda and ai no corrida: none notified

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.