Start Up No.1693: China’s ‘data trap’ in Africa, the metaverse’s latency problem, Qualcomm’s creepy always-on camera, and more


Is the Million Dollar Homepage the real precursor of NFTs? It’s certainly redolent of one new scheme. CC-licensed photo by charlene mcbride 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 10 links for you. Found in many countries. I’m @charlesarthur on Twitter. Observations and links welcome.


The real reason China is pushing “digital sovereignty” in Africa • Rest Of World

Yinka Adegoke:

»

it is the evolution of the “Digital Silk Road,” a term coined by Xi in a 2015 state white paper, that has quietly become a contentious topic for China-Africa watchers. The Digital Silk Road (DSR) includes everything from cross-border e-commerce, smart cities, and fintech apps through to big data, internet of things, smartphones, and undersea cables. These projects don’t grab headlines like shiny new Chinese-built airports and railways or spark panicked fears of China’s “debt trap diplomacy.” But the unfettered influence of Chinese firms developing every step of the digital ecosystem in nearly all African countries has become a growing point of concern, particularly for China’s rivals in the United States.

As Motolani Agbebi, a researcher at Tampere University in Finland, told Rest of World, significant Chinese involvement in Africa’s telecoms sector actually predates DSR. Between 1999 and 2001, Huawei and ZTE first started working consistently on the continent, supported by China’s “go out policy,” which promoted the internationalization of Chinese companies.

But the ongoing pandemic, which has forced African governments and their citizens online to participate in remote work, schooling, and delivery of government services, has underscored Africa’s dependence on Chinese tech. “There’s no doubt that technology infrastructure and the broader Digital Silk Road initiative are far more important to China today in regions like Africa than they were even just a few years ago,” said Eric Olander, managing editor of The China Africa Project.

«

Earlier this week the head of MI6 – that’s the chief of the UK’s counterintelligence service – spoke in public for the first time, and on China he warned about “data traps” and “debt traps”. This seems a bit like both together.
unique link to this extract


WTA suspends tournaments in China amid concern for Peng Shuai • The Guardian

Vincent Ni:

»

The Women’s Tennis Association (WTA) has announced the suspension of all tournaments in China amid concerns about the safety of the Chinese tennis star Peng Shuai, following weeks of a high-profile row with Beijing over the player’s wellbeing.

“With the full support of the WTA Board of Directors, I am announcing the immediate suspension of all WTA tournaments in China, including Hong Kong,” said the WTA chairman, Steve Simon, announcing the decision in a statement on Wednesday.

“In good conscience, I don’t see how I can ask our athletes to compete there when Peng Shuai is not allowed to communicate freely and has seemingly been pressured to contradict her allegation of sexual assault. Given the current state of affairs, I am also greatly concerned about the risks that all of our players and staff could face if we were to hold events in China in 2022.”

Early last month , 35-year-old Peng, a former doubles world No 1, used a post on Chinese social media website, Weibo to accuse 75-year-old Zhang Gaoli, a former Chinese vice-premier, of having coerced her into sex . Her lengthy post was quickly deleted by the censors, and Peng disappeared from public for more than two weeks as the WTA and colleagues said they were unable to reach her.

«

What’s been notable throughout this saga has been how the WTA hasn’t had to be pushed into taking this position. Even though there’s a lot of money at stake, which usually leads to sharp intakes of breath from sponsors, it has been forthright about this in a way that shames US basketball’s pusillanimous stance over a tweet referencing Taiwan by one of its bosses.
unique link to this extract


The metaverse, latency, and metalatency • Domos

Magnus Olden:

»

Are you annoyed by timing issues, like two people starting to talk simultaneously on video conference calls? Well, you are in for a new magnitude of frustration. Or maybe you are thinking that Video Conferences work surprisingly well and that the step from Video Conference to Interactive VR is easy?

This blog points out some facts around why delivering any Metaverse will be near impossible over the internet. Some of the technological building blocks and incentive structures of the internet are poorly suited for a Metaverse (or anything massively interactive). I will describe why:

• The Metaverse is different: Massive interactive VR will have more demanding network requirements than current applications, especially latency-wise. And there will be latency! 

• We create latency for each other. End-to-end Congestion Controls, the core mechanism for self-regulation of usage over the internet, relies on inducing latency. Additionally, the internet runs on shared infrastructure or frequencies, and Internet Services Providers have no incentives or simple means to change that.

• We create latency for ourselves. This is amplified in shared networks. Guaranteed latency propositions ignore these limitations and a couple of other elephants in the room.

• Removing that last 10% of variable latency is orders of magnitude more complex than removing the first 90%. It’s like asking Usain Bolt to shave another second off his 100 meter time.

«

Although, although, we used to think that the internet was just Full Up and we couldn’t get anything down the pipes and voice calls were impossible and video calls moreso, yet things improve. It might be that the real improvement needs to happen at the centres – the servers – rather than the edge, once fibre is ubiquitous.
unique link to this extract


Qualcomm’s new always-on smartphone camera is a privacy nightmare • The Verge

Dan Seifert:

»

“Your phone’s front camera is always securely looking for your face, even if you don’t touch it or raise to wake it.” That’s how Qualcomm Technologies vice president of product management Judd Heape introduced the company’s new always-on camera capabilities in the Snapdragon 8 Gen 1 processor set to arrive in top-shelf Android phones early next year.

Depending on who you are, that statement can either be exciting or terrifying. For Qualcomm, it thinks this new feature will enable new use cases, like being able to wake and unlock your phone without having to pick it up or have it instantly lock when it no longer sees your face.

But for those of us with any sense of how modern technology is used to violate our privacy, a camera on our phone that’s always recording images even when we’re not using it sounds like the stuff of nightmares and has a cost to our privacy that far outweighs any potential convenience benefits.

Qualcomm’s main pitch for this feature is for unlocking your phone any time you glance at it, even if it’s just sitting on a table or propped up on a stand. You don’t need to pick it up or tap the screen or say a voice command — it just unlocks when it sees your face. I can see this being useful if your hands are messy or otherwise occupied (in its presentation, Qualcomm used the example of using it while cooking a recipe to check the next steps). Maybe you’ve got your phone mounted in your car, and you can just glance over at it to see driving directions without having to take your hands off the steering wheel or leave the screen on the entire time.

The company is also spinning it as making your phone more secure by automatically locking the phone when it no longer sees your face or detects someone looking over your shoulder and snooping on your group chat. It can also suppress private information or notifications from popping up if you’re looking at the phone with someone else.

«

A follow-on from yesterday’s piece about concerns over surveillance. Here it comes down the road, sooner than you expected.
unique link to this extract


Cutting a Banksy into 10,000 (digital) pieces • The New York Times

Robin Pogrebin:

»

In the latest example of art market disruption, a prominent former auction executive teamed up with cryptocurrency experts in May to purchase the 2005 Banksy painting “Love Is in the Air” for $12.9m and now plans to sell off 10,000 pieces of it as NFTs, or nonfungible tokens.

The executive, Loic Gouzer, who upended the traditional auction format while he was at Christie’s — most notably orchestrating the sale of a $450.3 million Leonardo da Vinci painting in a contemporary art auction in 2017 — has helped found the company Particle, a platform that merges art and technology with a goal of reaching a broader pool of potential buyers.

…If successful, the venture could help fuel a burgeoning category of competition in the art market, with consortiums of multiple buyers challenging the pre-eminence of billionaire collectors at a time when the pandemic has accelerated online commerce. NFTs have become increasingly popular, accounting for one-third of online sales, or two% of the overall art market, according to the database Artprice.

«

Aiming for a $1,500 price tag on each (not very big) square. You know, they could do much the same with a blank web page where each pixel (or collection of them) was sold off to bidders – you could probably get a million dollars or so, easy.
unique link to this extract


Omicron found in US — plus 23 other countries in 5 of 6 global regions • Ars Technica

Beth Mole:

»

The omicron coronavirus variant has now been detected in at least 24 countries in five of six global regions—and as of this afternoon, that includes the United States.

The US Centers for Disease Control and Prevention confirmed this afternoon that the first US case was detected in a person in California who had returned to the US from South Africa on November 22 and tested positive on November 29. Researchers at the University of California, San Francisco carried out genomic sequencing identifying the omicron variant in the person, and the CDC confirmed that sequencing.

The CDC reported that the person was fully vaccinated and had only mild symptoms that are improving. In a press briefing Wednesday afternoon, top infectious disease expert Anthony Fauci said it appeared the person had not yet received a booster shot. Public health experts suggest that booster shots will significantly improve protection against the new, still poorly understood variant.

All of the person’s close contacts are being followed, Fauci added, and all have tested negative so far.

The detection of omicron in the US is unsurprising to health officials. “It was just a matter of time before the case of omicron would be detected in the United States,” Fauci said. But, “we know what we need to do to protect people,” he added, listing vaccination, boosting, and masking.

«

Turned out it wasn’t a vaccination drive in rich countries that defeated delta. It was a new variant. Nature defeating us yet again. (Side note: apparently the initial and second “O” in omicron should be short – like in “hot”, not “lone”.)
unique link to this extract


A different kind of Covid vaccine is about ready to roll • NPR

Joe Palca:

»

When Operation Warp Speed began spending billions of dollars to facilitate the development of a COVID-19 vaccine, it chose three vaccine technologies to back: mRNA vaccines being developed Pfizer and Moderna, a viral vector vaccine proposed by Johnson & Johnson, and protein subunit vaccines to be made by Sanofi and Novavax.

The first two technologies were successful, and there are now billions of vaccine doses in this country and around the world.

A year ago, Novavax was confident its vaccine would also be out there.

“We have a massive number of people working on scaling up our vaccine,” Gregg Glenn, Novavax’s president of research and development said in an interview in September 2020. “I am very optimistic by the year-end we’ll have a lot of product and we’re talking about more than 2 billion doses in 2021.”

But Glenn’s optimism was misguided. A large study of the vaccine took longer to complete than was hoped, and the company ran into manufacturing problems.

«

You’re wondering what a subunit vaccine is. From a link in the story:

»

Rather than injecting a whole pathogen to trigger an immune response, subunit vaccines (sometimes called acellular vaccines) contain purified pieces of it, which have been specially selected for their ability to stimulate immune cells. Because these fragments are incapable of causing disease, subunit vaccines are considered very safe. There are several types: protein subunit vaccines contain specific isolated proteins from viral or bacterial pathogens; polysaccharide vaccines contain chains of sugar molecules (polysaccharides) found in the cell walls of some bacteria; conjugate subunit vaccines bind a polysaccharide chain to a carrier protein to try and boost the immune response. Only protein subunit vaccines are being developed against the virus that causes COVID-19.

Other subunit vaccines are already in widespread use. Examples include the hepatitis B and acellular pertussis vaccines (protein subunit), the pneumococcal polysaccharide vaccine (polysaccharide), and the MenACWY vaccine, which contains polysaccharides from the surface of four types of the bacteria which causes meningococcal disease joined to diphtheria or tetanus toxoid (conjugate subunit).

«

Different from mRNA, which has code (the RNA) telling your cells how to make bits of the virus.
unique link to this extract


Here’s why movie dialogue has gotten more difficult to understand (and three ways to fix it) • Slashfilm

Ben Pearson:

»

I reached out to several professional sound editors, designers, and mixers, many of whom have won Oscars for their work on some of Hollywood’s biggest films, to get to the bottom of what’s going on. One person refused to talk to me, saying it would be “professional suicide” to address this topic on the record. Another agreed to talk, but only under the condition that they remain anonymous. But several others spoke openly about the topic, and it quickly became apparent that this is a familiar subject among the folks in the sound community, since they’re the ones who often bear the brunt of complaints about dialogue intelligibility. 

“It’s not easy to mix a movie,” says Jaime Baksht, who took home an Oscar for his work on last year’s excellent “Sound of Metal” and previously worked on Alfonso Cuarón’s “Roma.” “Everybody thinks you’re just moving levers, but it’s not like that.”

This problem indeed goes far beyond simply flipping a switch or two on a mixing board. It’s much more complex than I anticipated, and it turns out there isn’t one simple element that can be singled out and blamed as the primary culprit.

“There are a number of root causes,” says Mark Mangini, the Academy Award-winning sound designer behind films like “Mad Max: Fury Road” and “Blade Runner 2049.” “It’s really a gumbo, an accumulation of problems that have been exacerbated over the last 10 years … that’s kind of this time span where all of us in the filmmaking community are noticing that dialogue is harder and harder to understand.”

Join me and these industry experts as we sort through that “gumbo” and identify some of the most prominent reasons it has become more difficult to, in the paraphrased words of Chris Tucker’s Detective Carter in “Rush Hour,” understand the words that are coming out of characters’ mouths.

«

Great news for subtitle writers though.
unique link to this extract


DeepMind AI software collaborates with humans on mathematical breakthroughs • New Scientist

Matthew Sparkes:

»

AI software has collaborated with mathematicians to successfully develop a theorem about the structure of knots, but the suggestions given by the code were so unintuitive that they were initially dismissed. Only later were they discovered to offer invaluable insight. The work suggests AI may reveal new areas of mathematics where large data sets make problems too complex to be comprehended by humans.

Mathematicians have long used computers to carry out the brute force work of large calculations, and AI has even been used to disprove mathematical conjectures. But creating a conjecture from scratch is a far more complex and nuanced problem.

To disprove a conjecture an AI simply needs to churn through vast numbers of inputs to find a single example that contradicts the idea. In contrast, developing a conjecture or proving a theorem requires intuition, skill and the stringing together of lots of logical steps.

UK-based AI company DeepMind, owned by Google’s parent company Alphabet, has previously had success in using AI to beat humans at games of chess and Go, as well as solving the structures of human proteins. Now the firm’s scientists have shown that AI can provide human mathematicians with promising leads to develop theorems. That work has led to a conjecture in the field of topology and representation theory, and a proven theorem about the structure of knots.

Unlike most neural network research, in which an AI is fed large amounts of examples and learns to spot or create similar inputs, the AI here examined existing mathematical constructs for patterns. DeepMind says that its AI found both previously known and novel patterns and guided human mathematicians toward new discoveries.

«

Centaurs: humans and machines working together. (There’s also a series of Reith Lectures on BBC Radio 4 about AI. Definitely worth your time, as are the programmes by [Adam] Rutherford and [Hannah] Fry examining the ideas from the lectures further.)
unique link to this extract


Why subscription Twitter is a terrible idea, Twitter bans sharing private photos and videos, Twitter and free speech • Stratechery

Ben Thompson:

»

I promised that I was going to be ok with being wrong more often, and this week very well may be an excellent example! What are all of the reasons why a new business model for Twitter might be a bad idea?

Retention: This is the most obvious issue. Would that many people who use Twitter actually pay? I threw out the possibility that Twitter would lose a third of its users, but lots of folks think that Twitter would lose far more, even up to 90%. This would obviously be fatal, not simply from a revenue perspective, but from a content perspective.

The question is where all of those people would go; there obviously would be significant market demand for a free alternative. Another startup is probably the best guess, but I do wonder the extent to which the relatively idealistic world into which Twitter launched in 2006 was a critical factor in building a platform that is ground zero for information for nearly everything, from every point of view. That is what will be hard to replicate (and, as someone who was there in the early days noted, Twitter did evolve very rapidly in its early days, so saying it was “perfect” right off the bat isn’t quite right).

«

He also says that Growth (how do you add users?) and Risk (adding the Retention and Growth downsides together means a potentially big risk) are the other two problems. (This is from the subscription version of the newsletter, so not open to everyone.)
unique link to this extract


• Why do social networks drive us a little mad?
• Why does angry content seem to dominate what we see?
• How much of a role do algorithms play in affecting what we see and do online?
• What can we do about it?
• Did Facebook have any inkling of what was coming in Myanmar in 2016?

Social Warming, my latest book, and find answers – and more.


Errata, corrigenda and ai no corrida: none notified

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.