The ACLU is worried about Amazon’s doorbell plan

The ACLU is already working overtime on facial recognition issues, the group has already exposed flaws with Amazon’s Rekognition system.

Here’s their latest concern – plans from Amazon to add the tech to doorbells with the intent of spotting “suspicious visitors”:

While the details are sketchy, the application describes a system that the police can use to match the faces of people walking by a doorbell camera with a photo database of persons they deem “suspicious.” Likewise, homeowners can also add photos of “suspicious” people into the system and then the doorbell’s facial recognition program will scan anyone passing their home. In either case, if a match occurs, the person’s face can be automatically sent to law enforcement, and the police could arrive in minutes.

The ACLU goes on to explain how, based on what little research we have on these systems, current facial recognition technologies are less accurate when looking at people with darker skin.

Dimwitted bomb threat scam fails to pay off

A flurry of bomb “threats” across the US on Thursday. NBC News:

A wave of bomb threats were reported Thursday against businesses, schools, hospitals and other places across the country — causing panic and evacuations, although all appeared to be hoaxes.

Police in cities nationwide reported threats, some emailed, some phoned in.

The FBI in a statement said they were “aware of the recent bomb threats made in cities around the country, and we remain in touch with our law enforcement partners to provide assistance. As always, we encourage the public to remain vigilant and to promptly report suspicious activities which could represent a threat to public safety.

Thankfully, none of the threats have been credible, with no packages found. Almost immediately police forces were treating this as an obvious scam, based in part on the laughably-constructed email issuing the threat.

The note demanded $20,000-worth of Bitcoin by the end of the day. At 2pm PT, looking at the Bitcoin wallet details (which I won’t repeat here), not a single ransom has been paid.

Google puts Brazil’s destroyed Museu Nacional online

In September, Brazil’s national museum was heavily damaged by fire – a blaze that took with it some of the world’s most treasured artifacts. With an estimated 92% of its archive lost forever, the 200-year-old building is now permanently closed.

Today, Google announced it had digitally captured many of the exhibits at the Museu Nacional:

The section is a mixture of still images and virtual tours that use a Street View-like interface to look around.

It is of course little consolation for losing the physical objects, and the important history they held, but it at least acts as something of a record of what was once on show. 

The making of a computer-generated influencer

File this one under “depressing trends that will grow in 2019”, I think. From today’s Wall Street Journal:

Miquela Sousa, who recently graced billboards from London to Japan as part of an Ugg advertising campaign, ticks off all the boxes for a model of the moment: She is exotic, attractive and huge on Instagram.

She is also entirely fake, a computer-generated character who—despite what she says on Instagram—can’t feel the pain of a hangover or appreciate how hard it is to walk in stilettos.

Facial recognition ‘used to spot known stalkers at Taylor Swift gig’

Remarkable nib in Rolling Stone today:

Taylor Swift fans mesmerized by rehearsal clips on a kiosk at her May 18th Rose Bowl show were unaware of one crucial detail: A facial-recognition camera inside the display was taking their photos. The images were being transferred to a Nashville “command post,” where they were cross-referenced with a database of hundreds of the pop star’s known stalkers.

Woah boy. Lots to unpack there: if, as the story suggests, fans were unaware they were being scanned… then what? Worth burying into the T&Cs attached to concert tickets in future to see what’s lurking. This use case of facial recognition – pioneered in China – will only get more prevalent.

I’ve been in touch with the Oak View Group – whose CTO is quoted in the Rolling Stone piece as having seen the tech – to learn more. My questions:

  1. What kind of consent is given by fans who have their faces scanned by the technology?
  2. How long is the data retained?
  3. Can a concert goer opt out and still be admitted to the event?

Update:

I’ve been met with silence by Swift’s label, the Oak View Group and the Rose Bowl. Rolling Stone does not name the company involved in the tech.

But, I did have an interesting chat with Blink Identity, a company in the same space. It counts Ticketmaster among its investors (but, to be clear, is not the firm behind the tech apparently used at the Swift gig).

In 2019 Blink Identity – which claims to be able to verify a person’s face at walking speed – is going to start testing out its “your face is your ticket” system. The company likened it to the TSA pre-check line at the airport.

Rather than scanning everyone – as the Rolling Stones story suggests happened at the Rose Bowl – Blink Identity’s system provides those who opt-in (by sending a selfie) a chance to have a breezier entry into major events.

Another update:

Gizmodo is as confused as I am about this one:

The Rolling Stone report has taken off in the past day, with Quartz, Vanity Fair, the Hill, the Verge, Business Insider, and others picking up the story. But the only real information we have is from Downing. And so far no one has answered some key questions—including the Oak View Group and Prevent Advisors, which have not responded to multiple requests for comment.

For starters, who is running this face recognition system? Was Taylor Swift or her people informed this reported measure would be in place? Were concertgoers informed that their photos were being taken and sent to a facial recognition database in another state? Were the photos stored, and if so, where and for how long? There were reportedly more than 60,000 people at the Rose Bowl concert—how many of those people had their mug snapped by the alleged spybooth? Did the system identify any Swift stalkers—and, if they did, what happened to those people?

At the end of all this I’m wondering: why did Rolling Stone not name the company? 

On trend, Washington Post posts buoyant subscriber numbers

Some new numbers on the Washington Post’s subscription base. Quality newspapers, in the US at least, have found their sweet spot in attracting paying customers. Giving readers a chance to read five (or so) pieces for free means material is not locked away from the internet – like the Times Of London’s dismal paywall.

That, combined with the Trump presidency, is good news for the US capitol’s paper of record:

For context, the New York Times is doing even better. From last month:

More than three million paid digital-only subscribers. More than four million total.

The New York Times Company announced on Thursday that it surpassed those milestones during the third quarter of 2018, when the number of its digital subscribers showed a net increase of roughly 203,000.

What can’t be understated in all of this is the power of bundling things to these deals that aren’t news, as the NYT goes on to note:

Not all of the new subscribers signed on for news. The Times reported that, of the 203,000 net increase in digital subscribers, 143,000 signed on for digital news products, with the remainder paying for the company’s cooking and crossword features.

This might seem like something of a hack to boost numbers, but it’s just a modern incarnation of what has always been the case. Newspapers had TV guides and cinema listings – crossword and recipe apps fill that space, and fill it well.

Aussie teens with no tech knowledge produce hit app

If I was launching a social app, I’d put 100% of my publicity effort into finding the right celebrity ambassadors – and let the press coverage happen organically, which it certainly will.

Case in point: Two totally-green app developers, Business Insider writes, used a “break” from studying law in Sydney to drum up the kind of buzz Silicon Valley veterans dream of:

Australian twin sisters Colina and Hripsime Demirdjian decided to take a break from studying law at Sydney’s Macquarie University to create their first app, Moji Edit.

Moji Edit allows users to create and personalise an emoji of themselves.

The 24-year-olds weren’t prepared for the response. At one stage the app was ahead of Pokemon Go and celebrities started picking it up and tweeting about it.

The app got more than 500,000 users in four weeks. It featured in the UK and Ireland as Hot App of the Week, ranked #3 in Free Utilities in the US and was Canada’s #2 in free Utilities.

The key point there is the celebrity involvement. Celebrity use – not just endorsement – is the real fast track to making things happen.

Cutting-edge innovation (Not really. Stop it.)

Tis the season for pointless gadgets and Gillette is getting its announcement in early, before hordes more come our way at CES in January.

So it begins:

“Gillette today pilots a new concept designed to personalize the shaving experience in entirely new ways.”

And so it continues:

“This new platform leverages technology from 3D-printing powerhouse Formlabs to offer a series of customizable 3D-printed handles.”

Ok, so at this point you’re thinking – “Oh, its not a customised blade, just a handle, but that’s kinda cool”. But, reader, it’s not even that. Read on:

Visitors choose from a range of 48 designs so intricate they can only be produced using 3D-printing technology. After selecting the handle, consumers personalize it further by choosing from seven color options (black, white, red, blue, green, grey and chrome) and by adding text to the design. Handles can accommodate either Gillette’s MACH3 or Fusion5 ProGlide razor cartridges.

I guess “Gillette announces new handles” doesn’t have quite the same ring to it. Please, technology press, focus your attentions on meaningful innovation in 2019, not dross like this. 

(H/T @CMG_BBC)

The verdict on Sundar Pichai’s visit to Congress

As we head into 2019, its indisputable that our political systems need to assess how to better hold technology companies to account. After watching, for more than three hours, Sundar Pichai testify before Congress, my overriding thought is simply: there must be a better way.

Embed from Getty Images

The US needs to devise new formats for hearings like this. A format that  that caters to progress, rather than the egotistical whims of ill-informed members of Congress.

To compensate for gaps in lawmakers’ knowledge, these hearings need trusted experts, people who can stop the politicians from waddling down lines of questioning that give bosses the chance to trot out a prepared statement.

Or, in Pichai’s case today, a chance to point out the blatently obvious. In all of the tech hearings so far, this has to have been the lowest point:

Still, for now, it is what it is.

On Tuesday, Pichai only looked uncomfortable when being harangued with unanswerable gotchas, such as Rep Poe of Texas demanding to know if Google would track his location if he moves from one side of the room to the other. The answer is, of course, it depends what’s on your phone – not that Pichai was given the chance to respond.

Besides those brief moments, it was a typically calm display from a man apparently mostly on top of one of the most complexes businesses in the world.

Here’s my take for BBC News:

The Verge’s Casey Newton wrote that Pichai skated through the hearing:

After months of polite deferrals, Sundar Pichai finally went before Congress on Tuesday, and over the course of three and a half hours, said as little as possible. The hearing before the House Judiciary Committee was defined, as had been the Facebook hearings before it, by the widespread befuddlement of our lawmakers.

While Wired said Congress “blew” its opportunity:

It was a foreboding reminder of Congress’s continued technological ignorance, and a sign that while lawmakers almost unilaterally agree that something must be done about tech giants’ tremendous power, they remain unwilling to set aside partisan squabbles to actually do anything about it.

Slate Magazine’s Aaron Mak said Republicans “embarrassed themselves” during the session. Though, for what it’s worth, I don’t think many Democrats did much better, opting to attack Republican colleagues rather that using their privileged position to ask a proper question to an extremely powerful man.

But Mak did rightly give praise to the representatives that pressed Pichai on Project Dragonfly, the secretive China plan we know about thanks to various leaks and some great reporting by The Intercept.

From that same Slate piece, here’s a summary of what we learned (or at least, had confirmed):

Pichai acknowledged that there was such an effort but said that there were no current plans to follow through with the project. He also revealed, during questioning from Pennsylvania Rep. Keith Rothfus, that Google at one point had more than 100 people working on the effort. Rhode Island Rep. David Cicilline also asked, “Will you, Mr. Pichai, rule out launching a tool for surveillance and censorship in China while you’re CEO of Google?” The CEO talked about the importance of providing information but would not make any commitments on this issue either.

The Washington Post had Pichai’s exit interview, as it were. No great revelations in there, but they’ve promised a follow-up piece that will run on Wednesday. 

Jack Dorsey responds to criticism on Myanmar trip

Jack Dorsey’s recent tweets about a meditation retreat in Myanmar went down poorly in several quarters. New York Times

Mr. Dorsey now faces a backlash from critics who described his travel missives — which included a description of how he used his Apple Watch in airplane mode to track his heart rate during meditation and rest — as politically tone deaf.

On Tuesday, he tweeted about the upset:

For what it’s worth, I tweeted some admiration about Jack’s trip – I thought it was impressive to see a chief executive taking such an isolated break from work (and he really was – I had a couple of meetings at Twitter while he was away and staff, with some element of fluster, said they didn’t have a way to reach him).

A colleague (and others) criticised me for that view: 

Quite right too. On reflection – for both Jack and me – perhaps the intention was overshadowed by ignorance about the statement made by visiting the country at all.