Looking for the Broomway? Just follow the footpath that has killed around a hundred people. CC-licensed photo by Adrian Miller 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 11 links for you. Use them wisely. I’m @charlesarthur on Twitter. Observations and links welcome.
The government’s mass testing Moonshot project looks like a 90s Silicon Valley PowerPoint nightmare • Diginomica
Derek du Preez:
The idea is that if millions and millions of tests are readily available and results can be delivered in minutes, rather than days, then people can live a fairly normal life knowing their COVID-19 status. That’s all well and good, but this major project is being pushed by the same government that has still failed to deliver a contact tracing app and also didn’t have the foresight to recognise that maybe using algorithms to predict students’ grades would be a disaster. Amongst a litany of other technology blunders over the years…
Not to mention that the testing technology in question isn’t yet available and the alleged cost of the project is almost the equivalent of the entire NHS budget.
But hey, we like ambition! And no one ever did anything worthwhile without breaking new ground and trying something new. And as everyone knows, these are ‘unprecedented times’ and we’d all like to be able to go to the pub with our mates…
Well, that might be your thinking until you glance over the leaked briefing documents for project Moonshot. A quick scan over the slides – which are full of Venn diagrams, misaligned text, unexplained colour coding, BS consultancy catchphrases – and even the most optimistic amongst us will start to lose confidence.
It honestly looks and reads like a 90s Silicon Valley PowerPoint presentation from a sales exec that knows they’re pitching a piece of technology and an idea that doesn’t yet exist and isn’t achievable.
So, let’s take a look at how taxpayer money could be used to deliver 70 million COVID-19 tests a week. Because, you’d imagine that if the government was considering forking out £100bn it would want a clear, understandable, workable plan in place to instil public confidence and stand up to peer review.
I think this idea will quietly sink under the waves. Meanwhile Dominic Cummings had told the Department of Culture, Media and For Some Reason Sport that he wants the UK to have a trillion-dollar company. Perhaps they could set up a company with a million shares and buy one of them for million pounds. Voila! (American trillion.)
unique link to this extract
Starting September 17th, the platform will label or remove tweets involving election rigging or premature election results, treating those categories as particularly likely to cause immediate harm.
“The conversation happening on Twitter is never more important than during elections,” Twitter said in a blog post rolling out the new policy. “Twitter is where people come to hear directly from elected officials and candidates for office, it’s where they come to find breaking news, and increasingly, it’s an integral source for information on when and how to vote in elections.”
Twitter’s new rules will likely bring the platform into conflict with President Trump. Over the last few months, Trump has sent out numerous tweets making false or misleading statements about the November election and voting process. Under this new policy, confusing posts regarding ballot box tampering, elections results, and election rigging will either be fact-checked or removed.
Looking forward to Twitter trying to keep up with Trump’s sure-to-be-bonkers tweets on November 3 and afterwards. Perhaps just put his account in Twitter jail for a bit.
A wider question: assuming the best outcome, should Twitter delete his account after next January?
unique link to this extract
If you consult a large-scale map of the Essex coastline between the River Crouch and the River Thames, you will see a footpath – its route marked with a stitch-line of crosses and dashes – leaving the land at a place called Wakering Stairs and then heading due east, straight out to sea. Several hundred yards offshore, it curls northeast and runs in this direction for around three miles, still offshore, before cutting back to make landfall at Fisherman’s Head, the uppermost tip of a large, low-lying and little-known marshy island called Foulness.
This is the Broomway, allegedly “the deadliest” path in Britain, and certainly the unearthliest path I have ever walked. The Broomway is thought to have killed more than 100 people over the centuries; it seems likely that there were other victims whose fates went unrecorded. Sixty-six of its dead are buried in the little Foulness churchyard; the other bodies were not recovered. Edwardian newspapers, alert to the path’s reputation, rechristened it “The Doomway”.
You start out thinking “oh, it can’t be that bad can it?” and then he heads out onto it and you realise that yes, a hundred people could have been killed on it.
unique link to this extract
According to IC Insights, a firm of analysts, HiSilicon broke into the global top-ten design companies by revenue in the first half of 2020, the first Chinese firm to do so. Since it will no longer be able to design chips for its owner [Huawei] after September 14th, HiSilicon could profitably focus on doing so for third parties in China. That would generate a new revenue stream for Huawei. If instead Huawei were forced to shut HiSilicon, its laid-off engineers would be snapped up by chip-design teams at other Chinese technology giants like Alibaba, Tencent and ByteDance. Or they could start new design firms of their own; many are said to be slipping out pre-emptively.
Each scenario worries firms like Qualcomm. The big American chip-designer lists Chinese competition as a risk in its annual filings. Last year Chinese sales made up $11.6bn out of Qualcomm’s $24.3bn in revenue. A HiSilicon liberated from Huawei would threaten those sales.
Huawei is putting on a brave face. It says it will spend over $20bn on research and development this year, $5.8bn more than in 2019 and about as much as Amazon, a firm with double its sales. It hopes to gain new revenue streams less vulnerable to American attacks. These are unlikely to let up even if Joe Biden becomes president next year. But as Uncle Sam tightens the grip, it risks squeezing Chinese technology into a form which it no longer controls. Huawei hopes to hang on until then.
the documentary carries an air of gravitas. It prosecutes its case like a trial lawyer, calling one witness after another up to the stand. They include many of the great architects of social media as we know it today—people like Tim Kendall, Facebook’s former director of monetization; Justin Rosenstein, who invented the Like button; and Guillaume Chaslot, who created the recommended-video infrastructure for YouTube—all of whom denounce their former work.
But while The Social Dilemma establishes that there is a problem, it struggles to locate the source of the stink. The film begins with an offscreen producer asking technologists what, exactly, is wrong with social media. It ends with those same technologists offering their prophecies for the future. Mostly, it shows the technologists squirming in their seats, unsure of where to begin.
Eventually, though, they start talking. According to them, the problems are thus: We spend too much time on social media. We do this because, in essence, we have no choice. The people who work at tech companies have invested infinite money, time, and engineering power to design systems that keep us hooked, and which predict our every move. It’s how they make money: We are not the user, we are the product (such clichés are repeated frequently). Mark Zuckerberg and Susan Wojcicki are billionaires; meanwhile, everyone else has given up happiness, knowledge, intimacy, spontaneity, time with our families, free will.
On Netflix now. I haven’t seen it yet, but certainly sounds intriguing.
unique link to this extract
Russian intelligence hackers are back, Microsoft warns, aiming at officials of both parties • The New York Times
David Sanger and Nicole Perlroth:
The Russian military intelligence unit that attacked the Democratic National Committee four years ago is back with a series of new, more stealthy hacks aimed at campaign staff, consultants and think tanks associated with both Democrats and Republicans.
That warning was issued on Thursday by the Microsoft Corporation, in an assessment that is far more detailed than any yet made public by American intelligence agencies.
The findings come one day after a government whistle-blower claimed that officials at the White House and the Department of Homeland Security suppressed intelligence concerning Russia’s continuing interference because it “made the president look bad,” and instructed government analysts to instead focus on interference by China and Iran.
Microsoft did find that Chinese and Iranian hackers have been active — but often not in the way that President Trump and his aides have suggested.
Contrary to an assessment by the director of national intelligence last month that said China preferred former Vice President Joseph R. Biden Jr. win the election, Microsoft found that Chinese hackers have been attacking the private email accounts of Mr. Biden’s campaign staff, along with a range of other prominent individuals in academia and the national security establishment, including groups like the Atlantic Council and the Stimson Center.
With OneNote, I’ve loved brainstorming and taking notes with the $100 Surface Pen (sold separately). I’d love it even more if the pen could keep up with my writing. Another performance issue. Unfortunately, key Microsoft apps like Excel and Skype haven’t been optimized for two screens.
Microsoft and Google are also working with third-party app developers. The Kindle app, for instance, places a page on each screen to make this one adorable little e-reader. (Or at least it should. It glitched midway through testing, but began working again later, after I complained to Microsoft.)
You can also launch one app on each screen—Edge browser on left, Word on right, for instance. One of my favorite features is App Groups, which lets you pair two apps together to simultaneously launch. I have Twitter and TikTok in one with the label, “Bad for My Brain.”
One screen is still better suited to many of our current needs, and that makes this wide device feel awkward more often than not. Talking phone-style on the folded Duo is like holding a baking pan up to your head (cue sales pitch for $200 Surface Earbuds), and the display definitely gets in the way when you’re just responding to a quick text or snapping a quick photo. (And don’t get me started on the 11-megapixel camera and its position on the top left screen.)
Mr. Barlow said he often keeps his Duo in single-screen mode so he doesn’t have to unfold it when he wants to do something quickly.
The review video is five minutes, but they zip by. There are definitely quite a few tablet-y things that it does well, and quite a few phone-y things (particularly taking photos and making calls) that it does badly. A folding tablet? That wouldn’t be so bad. Dieter Bohn’s review at The Verge is worth reading for the software section too.
unique link to this extract
David Pierce with the story of how the Surface Duo came to be; the most interesting details being how the two organisations struggled to mesh:
A few weeks after their initial meeting, [Microsoft’s chief product officer Panos] Panay brought a small team to Mountain View and redid his pitch for a team of Microsoft and Google employees, this time with [Google head of Android, Hiroshi] Lockheimer by his side. From that point on, things kicked into gear. There was no hands-in moment, no signing on the dotted line. The two teams just started building a product together. Small cultural differences — should meetings happen on Teams or Meet? Do they last 25 minutes (Google’s way) or 30 (Microsoft’s)? — mostly fell by the wayside.
The hardest part of the project, both Lockheimer and Panay said, was getting the rest of their companies on board. It was a bit easier for Google, Lockheimer said, since the Android team is used to working with partners that look suspiciously like competitors. Still, every time a new team within either company got involved, it took time and convincing to bring them up to speed. Panay described it this way: “It starts with, ‘Hey, we’re building this new Surface.’ And they’re like, ‘Whoa, that’s awesome, what are we doing?’ ‘We’re building it on Android.’ And then … there’s a pause.” People would ask Panay, you know you’re in charge of Windows, right? And he’d say, yes, but we’re doing this with Android. “That’s a hurdle” for people, Panay said, “and an emotional one.”
Current and former employees said Microsoft remains mostly a top-down culture, driven by leaders and corporate edicts, while Google allows employees huge latitude to make and ship stuff. In theory that would make it easier for Panay to get buy-in from the rest of Microsoft, who might be expected to fall in line with the big new product. And it could make it harder within Google, where an Android exec might have trouble persuading a Maps or Gmail product manager to make dual-screen devices a priority.
Apps, social media platforms and online games that are specifically targeted at children will now have to put privacy at the heart of their design.
A code of practice outlining how children’s data should be protected has come into force and firms have 12 months to comply with the new rules. If they do not, they could face huge fines imposed by the Information Commissioner’s Office.
Some questioned whether the code would bring about real change. Information commissioner Elizabeth Denham said it was an important step towards protecting children online. “A generation from now we will all be astonished that there was ever a time when there wasn’t specific regulation to protect kids online. It will be as normal as putting on a seatbelt.
“This code makes clear that kids are not like adults online, and their data needs greater protections.”
She said the Information Commissioner’s Office (ICO) recognised that it could be difficult for smaller businesses to comply with the code and would offer “help and support” over the coming year.
…The scope of protections needed for children online was huge and the ICO might not be up to the job, said one digital rights campaigner, Jen Persson.
“The code is well-intentioned, and if enforced, may bring about some more focused change in the approach of some apps and platforms to stop collecting excessive data from children for example, and start to meet the requirements of core data protection law in place for over 20 years.
“The key risks are that since the ICO has not enforced to date on behalf of children in its current remit of concrete data protection law, that it may be seen as not having the capability to enforce those new things in the code that go beyond that and are subjective, such as the best interests of the child, or that outstrip the ICO technical knowledge and capacity.”
Zuckerberg acknowledged [in the interview on Axios on HBO] that conservative voices and opinions ranked as Facebook’s most engaged content.
“It’s true that partisan content often has kind of a higher% of people … engaging with it, commenting on it, liking it,” Zuckerberg said. “But I think it’s important to differentiate that from, broadly, what people are seeing and reading and learning about on our service.”
Also in the Axios interview, the Facebook chief said he would not remove anti-vaxxer posts, even as the leading virus experts express cautious optimism that a Covid-19 vaccination may become available late this year or early next year.
“If someone is pointing out a case where a vaccine caused harm or that they’re worried about it – you know, that’s a difficult thing to say from my perspective that you shouldn’t be allowed to express at all,” Zuckerberg said.
But he denied that Facebook’s algorithms are designed to push viewpoints “that are going to kind of enrage people somehow, and that’s what we try to show people”.
“That’s not actually how our systems work,” he added.
Zuckerberg is being so absurdly disingenuous here. He knows, because the point has been made to him so often, that this isn’t about “a case” where “a vaccine caused harm”. It’s about anti-scientific propaganda allied to people selling harmful “cures”. Facebook is derelict. I haven’t used the original Axios writeup because it’s all over the place. Rather than a transcript, it’s bite-size junk.
unique link to this extract
TikTok has acknowledged that it restricts LGBT-related hashtags in some countries as part of its “localised” approach to moderation.
A report by the Australian Strategic Policy Institute (ASPI) think-tank said many LGBT hashtags were “shadow-banned” in Bosnia, Jordan and Russia.
A shadow ban limits the discovery of content without indicating that a particular hashtag is on a ban list. TikTok said that some hashtags were restricted to comply with local laws.
According to the ASPI, terms that were not linking to content included:
• “gay” in Russian and Arabic
• “I am a lesbian” and “I am gay” in Russian
• “transgender” in Arabic
TikTok said that while some terms were restricted to comply with local laws, others were limited because they were primarily used to discover pornographic content.
There’s that Chinese censorship/moderation (one of those irregular verbs: I moderate, you censor), just as you might expect. What you don’t know is what else gets banned.
unique link to this extract
Errata, corrigenda and ai no corrida: none notified