Start Up: Apple hires Google’s AI chief, Trump campaign’s Facebook skillz, Grindr to stop sharing HIV data, and more

Why does Foxconn want Belkin? Because margins are good on accessories. Photo by Harsh Agrawal on Flickr.

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.

A selection of 12 links for you. Safety first. I’m @charlesarthur on Twitter. Observations and links welcome.

Woman opens fire at YouTube headquarters, wounding four before taking her own life • LA Times

James Queally, Benjamin Oreskes, Richard Winton, Tracey Lien and Angel Jennings:


A woman opened fire at the YouTube headquarters in San Bruno, Calif., wounding four people before taking her own life, authorities said.

San Bruno Police Chief Ed Barberini told reporters Tuesday afternoon that one person, believed to be the shooter, was found dead of a self-inflicted gunshot wound. Four other people were taken to area hospitals for medical treatment, according to Barberini.

Lisa Kim, a spokeswoman for Stanford Medical Center, said at least four patients from the shooting were expected to be admitted at the medical center.

“We do not know their conditions,” she said…

…At least two people were reportedly struck by gunfire, according to a law enforcement source who was not authorized to speak publicly about the incident. The shooter was described as a woman. So far authorities do not believe this was an act of terrorism and appears instead to have been a case of domestic or workplace violence — although the investigation has just begun.


So it looks like a domestic incident (as far as was known as 2200GMT). But those on all sides will use this incident to make points that aren’t justified by the facts. A take I liked: “they say the only way to stop a bad guy with a gun is a good guy with a gun but that just sounds like someone trying to sell TWO guns”.
link to this extract

Apple hires Google’s AI chief • The New York Times

Jack Nicas and Cade Metz:


Apple has hired Google’s chief of search and artificial intelligence, John Giannandrea, a major coup in its bid to catch up to the artificial intelligence technology of its rivals.

Apple said on Tuesday that Mr. Giannandrea will run Apple’s “machine learning and A.I. strategy,” and become one of 16 executives who report directly to Apple’s chief executive, Timothy D. Cook.

The hire is a victory for Apple, which many Silicon Valley executives and analysts view as lagging its peers in artificial intelligence, an increasingly crucial technology for companies that enable computers to handle more complex tasks, like understanding voice commands or identifying people in images.

“Our technology must be infused with the values we all hold dear,” Mr. Cook said in an email to staff members obtained by The New York Times. “John shares our commitment to privacy and our thoughtful approach as we make computers even smarter and more personal.”


Wow. That’s a hell of a coup. Giannandrea joined Google in 2010 from Metaweb (which Google bought). He’s got to be on a gigantic options deal with some big incentives around Siri et al.
link to this extract

Trump’s campaign said it was better at Facebook. Facebook agrees • Bloomberg

Sarah Frier:


Donald Trump’s presidential campaign has boasted often that it made better use of Facebook Inc.’s advertising tools than Hillary Clinton’s campaign did. An internal Facebook white paper, published days after the election, shows the company’s data scientists agree.

“Both campaigns spent heavily on Facebook between June and November of 2016,” the author of the internal paper writes, citing revenue of $44m for Trump and $28m for Clinton in that period. “But Trump’s FB campaigns were more complex than Clinton’s and better leveraged Facebook’s ability to optimize for outcomes.”

The paper, obtained by Bloomberg and discussed here for the first time, describes in granular detail the difference between Trump’s campaign, which was focused on finding new donors, and Clinton’s campaign, which concentrated on ensuring Clinton had broad appeal. The data scientist says 84% of Trump’s budget asked people on Facebook to take an action, like donating, compared with 56% of Clinton’s…

…Trump ran 5.9 million different versions of ads during the presidential campaign and rapidly tested them to spread those that generated the most Facebook engagement, according to the paper. Clinton ran 66,000 different kinds of ads in the same period.


And yet (I feel faintly obligated to point out) Trump lost the popular vote by millions of votes, and prevailed because 77,000 votes in three key states went his way; his winning margins in each of those states was smaller than the number of votes that went to the most popular third candidate.

Trump did Facebook better, but we need even more granularity to get clarity on what happened.
link to this extract

A Cambridge Analytica whistleblower claims that “cheating” swung the Brexit vote • The New Yorker

John Cassidy:


[Cambridge Analytica whistleblower Christopher] Wylie said that he was pretty sure A.I.Q. used C.A.’s data, adding, “You can’t have targeting software that doesn’t access the database. Cambridge Analytica would have a database and A.I.Q. would access that database, otherwise the software wouldn’t work.” He also argued that “A.I.Q. played a very significant role in Leave winning,” because the online ads that the pro-Brexit groups purchased with the help of the firm were “incredibly effective,” with very high conversion rates. On this basis, he said, “I think it is incredibly reasonable to say A.I.Q. played a very significant role in Leave winning.” Asked directly if this made the difference in the outcome of the referendum, Wylie replied, “I think it is completely reasonable to say there could have been a different outcome of the referendum had there not been, in my view, cheating.”

In response to Wylie’s statements, C.A. accused him of spreading “false information, speculation, and completely unfounded conspiracy theories.” Taking its defense to Twitter, the firm described Wylie as “a part-time contractor who left in July 2014 and has no direct knowledge of our work or practices since that date.” Contradicting its earlier claims, the firm said, “We played no role in the referendum on EU membership.” It also distanced itself from Aggregate I.Q., saying, “The suggestion that Cambridge Analytica was somehow involved in any work done by Aggregate IQ in the 2016 EU referendum is entirely false.”


link to this extract

Google shutting down URL shortener next year, existing links will keep working • 9to5Google

Abner Li:


The link is very common on the web and was first launched by Google in 2009. However, the company announced today that it’s winding down the URL Shortener beginning next month, with a complete deprecation by next year. Fortunately, existing links will continue to work.

The URL shortener service launched in 2009 for FeedBurner and the Google Toolbar. With neither of those services available, the same is now happening to for both consumers and developers. The latter group is being directed to Firebase Dynamic Links with today’s announcement meant to “refocus” Google’s efforts.

many popular URL shortening services have emerged and the ways people find content on the Internet have also changed dramatically, from primarily desktop webpages to apps, mobile devices, home assistants, and more.

However, for average users that just want to truncate a link, there is no new alternative from the company, with Google suggesting Bitly and


Link shortener advantages for the operator: you get to see where traffic is going; you get to see which users click on what and which users create what.

Link shortener disadvantages for the operator: spammers hammer your traffic; you have to watch for dead/dying links; you have to keep checking that old links aren’t taken over by spammers; it costs you a little.

I’d guess that Google saw the number of newly created links diminishing quite fast. I’d guess that Twitter and Facebook are the biggest players, via the and shorteners they use.
link to this extract

Grindr will stop sharing users’ HIV data with other companies • Buzzfeed

Azeen Ghorayshi:


The popular gay hookup app Grindr said late on Monday that it would stop sharing information about its users’ HIV status with third-party analytics companies.

The announcement came after BuzzFeed News revealed that Grindr had been securely providing two companies — Apptimize and Localytics, commonly used services to help optimize apps — with some of the information that Grindr users include in their profiles, including HIV status and “last tested date.”

The company decided to stop sharing the information with Localytics “based on the reaction — a misunderstanding of technology — to allay people’s fears,” chief security officer Bryce Case told BuzzFeed News. It will happen when the app’s next update is released, he said.

Still, Case defended Grindr’s decision to share the data, arguing that Apptimize and Localytics are simply tools to help apps like Grindr function better, and that the information was not shared to make money or for other nefarious purposes.


It was sharing this data with third-party analytics companies? What sort of world do these people live in?
link to this extract

Facebook delays home-speaker unveil amid data crisis • Bloomberg


Facebook Inc. has decided not to unveil new home products at its major developer conference in May, in part because the public is currently so outraged about the social network’s data-privacy practices, according to people familiar with the matter.

The company’s new hardware products, connected speakers with digital-assistant and video-chat capabilities, are undergoing a deeper review to ensure that they make the right trade-offs regarding user data, the people said. While the hardware wasn’t expected to be available until the fall, the company had hoped to preview the devices at the largest annual gathering of Facebook developers, said the people, who asked not to be named discussing internal plans.

The devices are part of Facebook’s plan to become more intimately involved with users’ everyday social lives, using artificial intelligence — following a path forged by Inc. and its Echo in-home smart speakers. As concerns escalate about Facebook’s collection and use of personal data, now may be the wrong time to ask consumers to trust it with even more information by placing a connected device in their homes. A Facebook spokeswoman declined to comment.


So “a deeper review to ensure that they make the right trade-offs regarding user data”? They hadn’t thought to do that before? This thing is dead in the water.
link to this extract

The curious case of the Belkin buy • Om Malik

Malik on Foxconn’s $866m of the US Belkin company:


Anker, for example, has come from nowhere and has become a dominant brand for accessories. Others such as Native Union and Mophie, too are well-known players. Amazon Basics and Bestbuy’s in-house brands, RocketFish, are other examples of companies that are aggressively trying to capture the consumer electronics business.

Against this backdrop, Belkin has done a good job of surviving in the market. They have managed not to become yet another “commodity brand.” But the question is for how long could they stay independent. In Foxconn, they have found an excellent parent to keep them growing. They can use Foxconn’s more significant infrastructure to their advantage.

As the press release notes:


Belkin International and its family of brands will continue to operate as a subsidiary of FIT under the leadership of CEO and founder Mr. (Chet) Pipkin and his executive team. Mr. Pipkin is expected to join FIT’s management team.


So what does Foxconn get? Well, if you are working as a contract manufacturing company for Apple, you aren’t making that huge a margin. Apple’s financial team makes sure that its suppliers and vendors are squeezed hard. It needs to figure out ways to boost its revenues and more importantly, margins. Belkin brings that to the stable, just like Sharp, which they bought in 2016.

Foxconn is also the financial backer for Nokia phones business, HMD Global.

It more than just money. It is also a company that is facing a lot of competition from the home-grown Chinese companies such as (former cable maker) Luxsher, which were upstarts ten years ago, but now have started to eat into Foxconn’s dominance. Belkin is yet another step towards making sure it has a future.


Not mentioned: accessories have better margins (percentage-wise) than lots of other electronics devices, so if you can sell a lot of them, you actually have a better chance of making money than if you’re just an ODM – original device manager – for someone else (as Foxconn often is).

And it has bought into a known brand. Another Chinese (Taiwanese) purchase of a western brand.
link to this extract

The death of the newsfeed • Benedict Evans

The aforesaid Evans:


If you have 1,500 or 3,000 items a day, then the chronological feed is actually just the items you can be bothered to scroll through before giving up, which can only be 10% or 20% of what’s actually there. This will be sorted by no logical order at all except whether your friends happened to post them within the last hour. It’s not so much chronological in any useful sense as a random sample, where the randomizer is simply whatever time you yourself happen to open the app. ’What did any of the 300 people that I friended in the last 5 years post between 16:32 and 17:03?’ Meanwhile, giving us detailed manual controls and filters makes little more sense – the entire history of the tech industry tells us that actual normal people would never use them, even if they worked. People don’t file. 

This is the logic that led Facebook inexorably to the ‘algorithmic feed’, which is really just tech jargon for saying that instead of this random (i.e. ‘time-based’) sample of what’s been posted, the platform tries to work out which people you would most like to see things from, and what kinds of things you would most like to see. It ought to be able to work out who your close friends are, and what kinds of things you normally click on, surely? The logic seems (or at any rate seemed) unavoidable. So, instead of a purely random sample, you get a sample based on what you might actually want to see.

Unavoidable as it seems, though, this approach has two problems. First, getting that sample ‘right’ is very hard, and beset by all sorts of conceptual challenges. But second, even if it’s a sucessful sample, it’s still a sample.

Looking at the first of these, there are a bunch of problems around getting the algorithmic newsfeed sample ‘right’, most of which have been discussed at length in the last few years. There are lots of incentives for people (Russians, game developers) to try to manipulate the feed. Using signals of what people seem to want to see risks over-fitting, circularity and filter bubbles. People’s desires change, and they get bored of things, so Facebook has to keep changing the mix to try to reflect that, and this has made it an unreliable partner for everyone from Zynga to newspapers. Facebook has to make subjective judgements about what it seems that people want, and about what metrics seem to capture that, and none of this is static or even in in principle perfectible.


This is a terrific post about the realities that social networks face. This is just a small part of a much longer post; all worth reading.
link to this extract

Instagram suddenly chokes off developers as Facebook chases privacy • TechCrunch

Josh Constine:


Without warning, Instagram has broken many of the unofficial apps built on its platform. This weekend it surprised developers with a massive reduction in how much data they can pull from the Instagram API, shrinking the API limit from 5,000 to 200 calls per user per hour. Apps that help people figure out if their followers follow them back or interact with them, analyze their audiences or find relevant hashtags are now quickly running into their API limits, leading to broken functionality and pissed off users.

Two sources confirmed the new limits to TechCrunch, and developers are complaining about the situation on StackOverflow.

In a puzzling move, Instagram is refusing to comment on what’s happening while its developer rate limits documentation site 404s. All it would confirm is that Instagram has stopped accepting submissions of new apps, just as Facebook announced it would last week following backlash over Cambridge Analytica. Developers tell me they feel left in the dark and angry that the change wasn’t scheduled or even officially announced, preventing them from rebuilding their apps to require fewer API calls.

Some developers suspect the change is part of Instagram parent company Facebook’s scramble to improve data privacy in the wake of its non-stop string of data scandals…

…Causing this kind of platform whiplash could push developers away from the Instagram ecosystem, not that the company was too keen on some of these apps. For example, Reports+ charges $3.99 per month to give people analytics about their Instagram followers. Sensor Tower tells TechCrunch that Reports+ has grossed more than $18m worldwide since October 2016 on the App Store and Google Play, and made more than $1.2m last month alone.


All the stable doors being slammed. But that’s quite the ecosystem there.
link to this extract

Amazon Music may be bigger than we thought • The Verge

Dani Deahl:


Amazon Music has tens of millions of active subscribers, the company tells The Verge. While analysts have often pointed to research that indicates it’s the third-largest company for on-demand streaming music (behind Spotify and Apple Music), Amazon has remained mum on confirming numbers.

Amazon has two tiers of music subscription: Prime Music (free for Prime subscribers) and Music Unlimited, which has monthly fees ranging from $3.99 to $14.99 depending on the number of devices, users, and if you’re already an Amazon Prime member. While Prime Music offers around 2 million songs ad-free, Music Unlimited provides more songs, greater control, and it’s cheaper than competitors’ $9.99 monthly fee for a single account.

Amazon launched Music Unlimited in April 2017 to compete with major streaming players while leveraging its Echo smart speakers, already deeply integrated with its music offerings and in millions of homes. Last year, Steve Boom, the vice president of Amazon Music, said in an interview that he “ See[s] us as one of the top global streaming services … I expect us to grow faster than everybody else.”

It appears that those predictions are being met. Amazon Music Unlimited subscriptions have grown more than 100% in the past six months…


Neatly timed to rain on Spotify’s IPO parade. What does “tens of millions” mean though? It’s between 20m and 90m. I think it’s a lot closer to the 20m.

When will Amazon get off the “no numbers number” habit? It feels like an invitation to industrial espionage, or an industry regulator that can audit these numbers. Otherwise they’re meaningless. (As a reminder, Spotify has 71m paid subscribers, Apple has 36m.)
link to this extract

Spotify’s stock falls from $165.90 opening price • The Washington Post

Hamza Shaban and Renae Merle:


Spotify made its highly anticipated Wall Street debut on Tuesday, with an opening price of $165.90, giving the music streaming company a valuation of $29.5 billion.

The price was 25% more than the reference price set by the New York Stock Exchange, based on how the stock traded on private markets before public trading began.

During the first moments of its public listing, Spotify’s stock experienced stable trading before falling more than 9% in the afternoon, to $150. Analysts had anticipated volatility during Spotify’s market debut because the company chose an unusual path to go public.

The streaming service giant, which trades under the symbol SPOT, bypassed many of the traditional steps of a Wall Street public offering. Company executives did not conduct a roadshow to convince big institutional investors, such as pension and mutual funds, to buy shares. Its chief executive even skipped the usual New York Stock Exchange ritual of ringing the opening bell. Epic Players, a theater group, preformed the honors.

What made Spotify’s public debut most notable, however, was how it offered its stock. Rather than issuing new shares, Spotify instead conducted a direct listing, in which no money was raised but existing shares were sold by employees and investors.

“Normally, companies ring bells. Normally, companies spend their day doing interviews on the trading floor touting why their stock is a good investment,” Daniel Ek, Spotify’s founder and chief executive said in a blog post Monday. “As I mentioned during our Investor Day, our focus isn’t on the initial splash. Instead, we will be working on trying to build, plan, and imagine for the long term.”


And so Spotify gets out of its tight spot with $1bn of debt raised in March 2016 that it had to pay off. Well played, Mr Ek.

Though some of the earlier investors might feel peeved. From that March 2016 WSJ story:


Fidelity Investments held its Spotify shares at $1,643 a share in January, down 27% from last August, according to regulatory filings. Another mutual fund, Vanguard International Growth, paid $2,229 a share for a stake in Spotify and still held it at that price as of December.


link to this extract

Errata, corrigenda and ai no corrida: none notified

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.