Start Up No.940: Pixel 3 v iPhone XS camera, how to make solar win, is China hijacking net traffic?, Gab faces mute, and more


Hard to think VR headsets haven’t caught on in a big way, isn’t it? Photo by Nan Palmero 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. Use them wisely. I’m @charlesarthur on Twitter. Observations and links welcome.

The dream of virtual reality is dying • The Outline

Joshua Topolsky:

»

Several prominent studios have shut down or ceased VR efforts, including Viacom and AltspaceVR, and Microsoft is a steadfast “no” when it comes to dipping its toes in the water via the Xbox. Sony has boasted about sales of the PSVR hitting 3 million in two years, but there are 82 million PS4 units in the hands of consumers (and keep in mind that Microsoft sold 35 million Kinects but still discontinued the product). With cumbersome hardware (which, let’s be honest, looks really stupid to most people), absurd PC requirements, and nearly no AAA titles to lure the curious into the world of VR, it’s becoming increasingly unlikely that we’ll see a major shift to virtual reality any time soon.

Also worth noting: if you’re looking to Magic Leap for a kind of bridge to the future with its AR efforts, don’t get too wound up. Brian Merchant’s excellent and detailed feature story for Gizmodo on the company’s struggles to get around the same hardware, software, and consumer adoption issues that plague VR make it clear there is no easy answer in this space.

«

The top quote is from the CEO of CCP Games, responsible for Eve:Online, who says “We expected VR to be two to three times as big as it was, period… A lot of people bought headsets just to try it out. How many of those people are active? We found that in terms of our data, a lot of users weren’t”.

But would two or three times larger have really given it enough momentum? Anyway, off it goes through the Trapdoor of Doom (a little-known opening in the Hype Cycle).
link to this extract


Stripe steps away from Gab network after synagogue shooting • Irish Times

Tim Bradshaw:

»

Stripe and PayPal, as well as hosting provider Joyent, all said they would stop Gab from using their services, citing violations of their terms of services, which do not allow hate speech.

Gab slammed the moves as “direct collusion between big tech giants” against it.

Stripe, the Silicon Valley-based online payments company established by Limerick brothers Patrick and John Collison, said over the weekend that it was suspending transfers “effective immediately”.

The company said Gab founder Andrew Torba had not “provided us sufficient evidence that Gab actually prevents violations of our policies”.

“If there’s more information you can provide on how exactly Gab will moderate its platform for adult content and other violations of our ToS [terms of service], we’re open to having a phone call this week to discuss,” the payments company said.

The moves are likely to reopen the debate about the limits of free speech online and the potential for social networks to radicalise users.

Gab was launched two years ago by tech entrepreneur Andrew Torba, who became frustrated with what he perceived as a bias against conservative views on California-based social networks such as Facebook and Twitter.

«

(Bradshaw being syndicated from the FT. This seems to be the edition where we find people being syndicated, as you’ll see.)
link to this extract


Attacks on Jewish people rising on Instagram and Twitter, researchers say • NBC News

David Ingram:

»

Researchers who study social media say that they are seeing an increase in anti-Semitic posts from far-right users of Instagram and Twitter and that the services aren’t doing enough about it.

Separate researchers who were independently looking at the two social networks said attacks on Jewish people had spiked on both services ahead of the midterm elections on Nov. 6, similar to a rise in harassment before the 2016 presidential election.

Many but not all of the posts mention billionaire investor and philanthropist George Soros, the researchers said. Soros is frequently the subject of unfounded conspiracy theories, and his home was among the targets in a series of attempted bombings this month.

Jonathan Albright, a researcher at Columbia University in New York who directs a center on digital forensics, told NBC News that the amount of anti-Semitic material posted to Instagram and tied to Soros was possibly the worst sample of hate speech he had seen on the site.

The recommended top posts for the hashtag “#soros” on Instagram on Thursday included a photo of Soros with the caption “I am the devil” and a cartoon suggesting that Soros and other targets of the explosive devices were themselves behind the bombs, a “false flag” conspiracy theory that gained traction online before the arrest of a Florida man on Friday.

«

link to this extract


China systematically hijacks internet traffic: researchers • iTnews

Juha Saarinen:

»

Researchers have mapped out a series of internet traffic hijacks and redirections that they say are part of large espionage and intellectual property theft effort by China.

The researchers, Chris Demchak of the United States Naval War College and Yuval Shavitt of the Tel Aviv University in Israel, say in their paper that state-owned China Telecom hijacked and diverted internet traffic going to or passing through the US and Canada to China on a regular basis.

Tel Aviv University researchers built a route tracing system that monitors BGP announcements  and which picks up on patterns suggesting accidental or deliberate hijacks and discovered multiple attacks by China Telecom over the past few years.

In 2016, China Telecom diverted traffic between Canada and Korean government networks to its PoP in Toronto. From there, traffic was forwarded to the China Telecom PoP on the US West Coast and sent to China, and finally delivered to Korea.

Normally, the traffic would take a shorter route, going between Canada, the US and directly to Korea. The traffic hijack lasted for six months, suggesting it was a deliberate attack, Demchak and Shavitt said.

Demchak and Shavitt detailed other traffic hijacks, including one that saw traffic from US locations to a large Anglo-American bank’s Milan headquarters being terminated in China, and never delivered to Italy, in 2016.

«

link to this extract


The global tech backlash is just beginning • The Toronto Star

Christopher Mims:

»

The largest tech businesses reach more people than any other companies have in history, and by many metrics they have also grown at unprecedented speeds. The companies themselves argue tech is bringing great benefits to people and improving their lives, yet when they enter industries, they consolidate power and make competitors miserable in ways not seen since the Gilded Age.

As people around the world become more familiar with the internet, their views tend to change from enthusiasm to caution. A survey by the Centre for International Governance Innovation reveal that in Kenya, for example, people are singularly positive about the impact of tech, whereas in North America and Europe, people are more concerned about Big Tech’s overreach.

“Familiarity breeds contempt,” says Fen Hampson, director of global security and politics at CIGI, who conducted the survey.

As the backlash plays out, it has the potential to subdivide the internet, forcing the biggest players to create separate products and procedures for different regions. The results—following a costly, complicated and protracted transition—will be better for consumers in some cases, and significantly worse in others. Europe and the U.S. The global tech backlash starts in the West, where countries have been feeling the results of Big Tech’s growing power the longest.

«

(Syndicated from his home at the WSJ. Free to read!)
link to this extract


Corporate tax and tech companies in the UK • Tax Watch UK

George Turner:

»

In this paper we seek to estimate the revenues made by five of the largest technology companies in the world – the Tech 5 – from their UK customers. The companies included in the study are: Apple, Google, Facebook, Cisco Systems and Microsoft. We then estimate the profits these companies are making from their UK sales based on the published profit margins of those companies. From there we can make an estimate of how much tax these companies would generate in the absence of profit shifting.

In total we estimate that in 2017 these five companies earned revenues of £23.4 bn from UK customers. We further estimate that profit attributable to these sales was £6.6bn, which at the prevailing rates would have given a tax liability of £1.26bn.

The profits declared in the accounts of the UK subsidiaries of these companies, and their tax liabilities, were far less. In total, the accounts of the main UK subsidiaries of the companies we looked at suggested a combined tax liability of £191m. This is more than £1bn less than we calculate would have been due if the accounts of the UK subsidiaries of the Tech 5 more accurately reflected the revenues and profits made from UK customers.

These findings bring into focus just how much money the UK government is losing to profit shifting by large multinationals every year, and how efforts to combat this practice have largely failed. To put this into context, HMRC estimates that corporation tax avoidance by all large companies costs the Treasury just £700m a year.

«

It’s budget day in the UK. But according to Turner, “We find that years of naming and shaming, tax investigations and efforts to change the tax system have largely failed.”
link to this extract


The death of FilmStruck is a dark day in the history of movies • Slate

Joanna Scutts:

»

The strangled corporate newspeak of the memo announcing the closure, with its reference to the “learnings” to be gleaned from the FilmStruck experiment, engenders the same kind of helpless rage as the tortured syntax of Donald Trump’s tweets—it’s so painfully revealing of the kind of grandiose carelessness that is the hallmark of power right now.

As Warner gears up to face down Disney with its direct-to-consumer streaming service, launching next year, it’s clear that the company has no interest in catering to passionate fans of its back catalog, only in chasing the largest possible audience for its new releases. What’s not clear is why it has to be a zero-sum game, and why efforts at preservation and education have to be eliminated in order to chase the biggest possible audience and present them with a library far broader than it is deep. As the screenwriter John August recently pointed out, there are still hundreds of movies from the home-video era that are not available to stream, and the availability of older titles is even more of a patchwork. This is a slow erosion of cultural heritage under the guise of infinite availability. Titles that are not available to stream will be harder to assign in classes, will cease to bubble up into the cultural awareness, and will eventually cease to matter.

«

It seems possible – likely, even? – that Warner could just add the Filmstruck catalogue to whatever it launches next year? Yet this seems to be completely discounted as an idea. What would Warner lose by doing that?
link to this extract


Let’s store solar and wind energy – by using compressed air • The Conversation

Seamus Garvey is professor of dynamics at the University of Nottingham:

»

The concept seems simple: you just suck in some air from the atmosphere, compress it using electrically-driven compressors and store the energy in the form of pressurised air. When you need that energy you just let the air out and pass it through a machine that takes the energy from the air and turns an electrical generator.

Compressed air energy storage (or CAES), to give it its full name, can involve storing air in steel tanks or in much less expensive containments deep underwater. In some cases, high pressure air can be stored in caverns deep underground, either excavated directly out of hard rock or formed in large salt deposits by so-called “solution mining”, where water is pumped in and salty water comes out. Such salt caverns are often used to store natural gas.

Compressed air could easily deliver the required scale of storage, but it remains grossly undervalued by policymakers, funding bodies and the energy industry itself. This has stunted the development of the technology and means it is likely that much more expensive and less effective solutions will instead be adopted. At present, three key problems stand in the way of compressed air…

«

One of them is that it’s too long-lived. Weirdly. Because money likes things that work quickly, not over the course of 50 or 100 years.
link to this extract


The secret to making green tech like solar panels go mainstream • Daily Beast

Tarpley Hitt:

»

[MIT professor David] Rand and [Yale psychology grad student Gordon] Kraft-Todd used the data—conveniently split into two distinct groups of towns, towns where representatives installed solar panels and towns where representatives did not—to figure out what made customers adopt solar energy. They found that when people saw an ambassador who used solar panels, they were inclined to buy one for themselves. As part of each campaign, Solarize tapped several locals to serve as “solar ambassadors,” or people who would act as the primary representatives for the campaign in their town.

Rand and Kraft-Todd focused on these figures, wondering what made some ambassadors better than others.

They found that figures who were central to city operations–an alderperson, a well-known community volunteer, or other public servants–were often tapped for their name recognition, even if they didn’t have solar energy in their own homes.

As a result, only a small minority of the “solar ambassadors” actually used the energy they were talking so much about. In Rand and Kraft-Todd’s survey of 58 towns, only 32.7% of the ambassadors had actually installed solar panels through the Solarize program. They didn’t walk the talk, and solar panel adoption was weak in these towns.

But in towns where the ambassador had panels themselves, 62.8% more people adopted solar panels.

«

It’s very much about “I’ll do what you do, not what you say.”
link to this extract


AIs trained to help with sepsis treatment, fracture diagnosis • Ars Technica

John Timmer:

»

The new research isn’t intended to create an AI that replaces these doctors; rather, it’s intended to help them out.

The team recruited 18 orthopedic surgeons to diagnose over 135,0000 images of potential wrist fractures, and then it used that data to train their algorithm, a deep-learning convolutional neural network. The algorithm was used to highlight areas of interest to doctors who don’t specialize in orthopedics. In essence, it was helping them focus on areas that are mostly likely to contain a break.

In the past, trials like this have resulted in over-diagnosis, where doctors would recommend further tests for something that’s harmless. But in this case, the accuracy went up as false positives went down. The sensitivity (or ability) to identify fractures went from 81% up to 92%, while the specificity (or ability to make the right diagnosis) rose from 88% to 94%. Combined, these results mean that ER docs would have seen their misdiagnosis rate drop by nearly half.

Neither of these involved using the software in a context that fully reflects medically relevant circumstances. Both ER doctors and those treating sepsis (who may be one and the same) will normally have a lot of additional concerns and distractions, so it may be a challenge to integrate AI use into their process.

«

That is the point, isn’t it: it’s great when you’re not trying to figure out which of 15 different possible wrong things is wrong with the patient.
link to this extract


Pixel 3 vs. iPhone XS camera face-off: why Google wins • Tom’s Guide

Caitlin McGarry:

»

The iPhone XS takes more natural shots with colors that are more true to life. Its dual-lens shooter takes portraits that also are more DSLR-like than the Pixel’s. But the Pixel 3 edges out the iPhone XS thanks to the help of software that turns out bright, crisp and colorful photos, even in at night. We’re betting the Pixel 3’s low-light images will look even better when the promised Night Sight features debuts in a software update. With Night Sight, the Pixel will then combine several low-light frames to fill in details and make the final image look brighter.

Some photographers don’t want software doing all the work. In that case, the iPhone XS provides a more natural-looking shot you can take to the next level with your own editing (or an artfully applied Instagram filter). But the Pixel 3’s camera will only get smarter, and we’re looking forward to seeing what other features are in store.

«

The sort of comparison we’ve been looking for. The Pixel 3’s smart stacking of exposures to do Night Sight (but it’s more than that) and for street scenes is quite something.
link to this extract


Apple: the second-best tech company in the world • The Outline

Joshua Topolsky:

»

Apple’s lack of data (and its inability or unwillingness to blend large swaths of data) actually seems to be one of the issues driving its slippage in software innovation. While Google is using its deep pool of user data to do astounding things like screen calls or make reservations for users with AI, map the world in more detail, identify objects and describe them in real-time, and yes — make its cameras smarter, faster, and better looking — Apple devices seem increasingly disconnected from the world they exist in (and sometimes even their own platforms).

As both Amazon and Google have proven in the digital assistant and voice computing space, the more things you know about your users, the better you can actually serve them. Apple, on the other hand, wants to keep you inside its tools, safe from the potential dangers of data misuse or abuse certainly, but also marooned on a narrow island, sanitized and distanced from the riches that data can provide when used appropriately.

«

I’m willing to be corrected, but I don’t think it’s deep pools of user data that Google’s using for Call Screening or Duplex. It’s AI systems which have been taught on quite different sets of data from email. (I don’t know what they have been taught on.) Certainly, user data makes maps better, and the data from Google Photos does – that’s probably a key input to the photo system on the Pixel 3.

But that data does exist, and whether Apple starts to use it more broadly is a key question for the future. It’s the collision of questions: can you improve the camera (and other systems) without embedded AI? At present the answer seems to be no. (Though might that be just because when everything’s getting AI, getting AI seems like the only answer.)
link to this extract


Errata, corrigenda and ai no corrida: none notified

3 thoughts on “Start Up No.940: Pixel 3 v iPhone XS camera, how to make solar win, is China hijacking net traffic?, Gab faces mute, and more

  1. Re Apple and AI. I think there’s a bit of overselling going on:
    1- everything with an algorithm is being rebranded “AI”. I’m sure a lot of wat goes on with cameras is edge detection etc, at this rate solving a+b=c is AI too.
    2- data and silicon are free, optics are expensive. I don’t think “AI” is the best way to make better pics, but it certainly is the cheapest and less space-hungry. It’s a compromise, not a panacea.

    I’m fairly sure the same is true of a lot of things that are being AI’d. To a hammer… and GAFA have lots of data. So we get AI’d Feeds instead of a re-thinking of what a Feed should be, etc.

  2. Re. Ambassadors. It’s not really ambassadors, it’s anyone really. I’m experiencing that with my proselytizing for Xiaomi. I got one, that made convincing people close to me much easier, and now I’m just starting to see 2nd-degree convertees I haven’t interacted with but are getting one after seeing the ones I pushed for.

    It’s very literal though, maybe because people don’t know/trust the brand yet. You’ve got to show them the Redmi Note 5 to be convincing, confronted with my huge Mi Max they couldn’t get past the size, and diving into alternatives (Pocophone F1 for games or higher perfs, Redmi not-Note for simpler needs) meets a lot of resistance. Past early adopters, people just can’t be bothered with product analysis. That makes most of Apple’s success I guess.

  3. Seems Xiaomi is even getting into software now, with a Fortnite clone: https://www.androidheadlines.com/2018/10/xiaomi-battle-royale-game-beta.html
    Why not, they also make underwear and lamps and everything in-between, and have some experience with software – MIUI started as a Launcher for any Android phone.
    I’m guessing the hardware reqs. will be much lower than Fortnite’s to better match Xiaomi’s installed base. Will it run on Mi TV boxes too ?
    It’ll be interesting to see if they can make a good product, then make it successful, then achieve some synergy off it. Same problematic as Apple and TV.

Leave a Reply

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

WordPress.com Logo

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

Google+ photo

You are commenting using your Google+ 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.