The reason why video conference calls are so exhausting comes down to the sound – and not being there. CC-licensed photo by Nick Doty 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.
There’s another post at the Social Warming Substack due at 0845 UK time. Free signup.
A selection of 9 links for you. Artisinal. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.
Not remotely cool: the science of “Zoom fatigue” • Big Think
Richard Cytowic:
When face-to-face we process a slew of signals without having to consciously think about them: facial expression, gesture, posture, vocal tone and rhythm, and the distance between speakers. We read body language and make emotional judgments about whether others are credible or not. This is easy to do in person, whereas video chats force us to work to glean the same cues. This consumes a lot of energy. Recall that compared to electronic devices, the human brain operates at ridiculously slow speeds of about 120 bits (approximately 15 bytes) per second. Listening to one person takes about 60 bits per second of brainpower, or half our available bandwidth. Trying to follow two people speaking at once is fairly impossible for the same reason multitaskers fare poorly: attempting to handle two or more simultaneous tasks quickly maxes out our fixed operating bandwidth.
As attention flags, we fatigue. Yet it is the audio gaps, not the video, that makes Zoom sessions draining. All languages have clear rules for conversation that assure no overlap but no long silences. Online meetings disrupt that convention because the separate sound and video streams are chopped into tiny digital packets and sent via different pathways to the recipient’s end where they are electronically reassembled. When some packets arrive late the software must decide whether to wait to reassemble them — causing a delay — or stitch together whatever packets are available, giving rise to stuttering audio.
Video conferencing platforms have opted to deliver audio that arrives quickly but is low in quality. Platforms aim for a lag time of less than 150 milliseconds. Yet that is long enough to violate the no-overlap/no-gap convention to which speakers are accustomed. A round-trip signal can take up to 300 milliseconds before one gets a reply, a pause that makes speakers seem less convincing and trustworthy. Repeatedly having to sort out talking over one another and who goes first is also tiresome and draining to everyone on the call.
Cytowic has a new book – Your Stone Age Brain in the Screen Age: Coping with Digital Distraction and Sensory Overload – which has just come out in the US. Now you can refer to it for why you don’t want to do a Zoom meeting.
New crypto-state emerges in the Himalayas: Bhutan has twice as many bitcoins as El Salvador • EL PAÍS English
Álvaro Sánchez:
A tiny nation squeezed between China and India deep within the Himalayas, Bhutan has become an unlikely cryptocurrency hub. The kingdom might be more accustomed to making the travel pages for its bucolic landscapes and Buddhist monasteries, but it has now leapt to the forefront of the cryptosphere after the firm Arkham Intelligence revealed that the state-owned conglomerate Druk Holdings owns 13,011 bitcoins, slightly more than double the amount declared by El Salvador’s Nayib Bukele (5,877). At current prices, this stash is valued at about $780m, which for a population of about 780,000 inhabitants represents $1,000 in bitcoins per citizen.
Arkham explains that this small fortune comes from bitcoin mining operations carried out by Bhutan’s investment arm, the aforementioned Druk Holdings, a name that means “thunder dragon.” This dragon appears on the country’s flag, holding jewels as a symbol of wealth. “We were able to corroborate the chronology of the mining activity with evolving satellite images of the facilities’ construction,” Arkham notes. The largest of these infrastructures is located on the grounds of the failed Education City, with which the authorities sought to tackle emigration and reduce unemployment, but which has ended up housing bitcoin factories that are in operation 24/7 instead of classrooms and books.
“Unlike most governments, Bhutan’s bitcoins come not from law enforcement seizure of assets, but from bitcoin mining operations, which have increased dramatically since early 2023,” states Arkham. Bhutan now ranks fourth among those countries with the most bitcoins, trailing only the US, China and the UK.
Didn’t realise the UK was such a big hub for bitcoins. Is that including all the ones in the Welsh landfill? Also, nobody seems to have written about what effect, if any, the adoption of bitcoin has had on El Salvador’s economy. Or have I just missed it?
The AI boom has an expiration date • The Atlantic
Matteo Wong:
All of this [AI] infrastructure will be extraordinarily expensive, requiring perhaps trillions of dollars of investment in the next few years. Over the summer, The Information reported that Anthropic expects to lose nearly $3bn this year.
And last month, the same outlet reported that OpenAI projects that its losses could nearly triple to $14bn in 2026 and that it will lose money until 2029, when, it claims, revenue will reach $100bn (and by which time the miraculous AGI may have arrived).
Microsoft and Google are spending more than $10bn every few months on data centers and AI infrastructure. Exactly how the technology warrants such spending—which is on the scale of, and may soon dwarf, that of the Apollo missions and the interstate-highway system—is entirely unclear, and investors are taking notice.
When Microsoft reported its most recent earnings, its cloud computing business, which includes many of its AI offerings, had grown by 29%—but the company’s stock had still tanked because it hadn’t met expectations. Google actually topped its overall ad-revenue expectations in its latest earnings, but its shares also fell afterward because the growth wasn’t enough to match the company’s absurd spending on AI.
Even Nvidia, which has used its advanced AI hardware to become the second-largest company in the world, experienced a stock dip in August despite reporting 122% revenue growth: Such eye-catching numbers may just not have been high enough for investors who have been promised nothing short of AGI [artificial general intelligence].
Absent a solid, self-sustaining business model, all that the generative-AI industry has to run on is faith.
Cleaning up “Scientific Reports”: can it be done? • Science
Derek Lowe:
I have had some problems with the journal Scientific Reports over the years, and I’m not alone. At the same time, I’ve read some interesting and useful papers published there as well. But worthless/faked manuscripts showing up in a journal tend to contaminate everything else that shows up there, which is a problem that you’d hope that scientific publishers are concerned about. To put things in the style of my late father, his one of his analogies was that if he had a gallon of urine and put a shot glass of wine into it, he still had a gallon of urine. On the other hand, if he had a gallon of wine and put a shot glass of urine into that, he now had a second gallon of urine. That’s the problem.
This open letter, signed by many well-known literature fraud experts, is (to me) more than enough evidence that Scientific Reports has some serious problems with the papers it’s letting through, and that the publishers (Springer Nature) are not doing enough to address them. It shows numerous examples of papers with odd and questionable references in them and with phrases that are redolent of (unstated) chatbot use, those apparently in attempts to bypass automated plagiarism-detection software. The authors of the letter note that even when the editors have taken action, that can be just to republish the same paper with slightly altered phrases…
…As the letter goes on to note, deploying more AI and automated systems is not going to be enough to fix this problem. Actual humans are going to have to hit some buttons here, and some of those buttons need to be labeled “delete”. The journal needs to show what editors handle each paper (which is currently invisible), because it’s likely that a small number of them are responsible for an outsize fraction of the problem.
Springer Nature doesn’t come out of this looking good. What if it likes the money it gets from subscriptions more than it worries about the reputational damage?
Meta fires staff for abusing $25 meal credits • Financial Times
Hannah Murphy and Stephen Morris:
Meta has fired about two dozen staff in Los Angeles for using their $25 meal credits to buy household items including acne pads, wine glasses and laundry detergent.
The terminations took place last week, just days before the $1.5tn social media company separately began restructuring certain teams across WhatsApp, Instagram and Reality Labs, its augmented and virtual reality arm, on Tuesday.
The revamp has included cutting some staff and relocating others, several people familiar with the decisions said, in a sign that chief executive Mark Zuckerberg’s recent efficiency drive is still under way.
Like most big tech companies, Meta offers free food to employees based out of its sprawling Silicon Valley headquarters as a perk. Staff based in smaller offices without a canteen are offered Uber Eats or Grubhub credits, for example, for food to be delivered to the office.
Staff are given daily allowances of $20 for breakfast, $25 for lunch and $25 for dinner, with meal credits issued in $25 increments.
Those who were fired were deemed to have abused the food credit system over a long period of time, said one person familiar with the matter. Some had been pooling their money together, they said, while others were getting meals sent home even though the credits were intended for the office.
Those who violated the company rules only on occasion were reprimanded but not terminated, the person added.
At least one of these people was on a $400k salary. It seems incredibly petty on both sides: you might hope people wouldn’t need such credits (then again, it’s not necessarily cheap to live in LA), but how has the company really lost out?
2004 was the first year of the future • The Verge
In early 2004, the world was shaking it like a Polaroid picture, flocking to theaters to see what was going to happen with all those hobbits, and wondering if that Tom Brady guy was something special. Meanwhile, a few folks around the world were inventing the web as we know it now: A world-shaking social network was brewing in a Harvard dorm room. A Google employee was dreaming up the future of email in their spare time. The coolest cellphone of all time was just about to drop. The internet was still a niche activity, but that was about to change — and fast.
In so many ways, the digital world in which we now all live was created 20 years ago. Google went public and began to ascend to rule the web. Facebook, Gmail, Firefox, Flickr, and Digg all launched — the year Web 2.0 became the web. “Blog” and “the long tail” were on no one’s radar before 2004, and since then, they’ve been everywhere. The United States went through a contentious election, a bunch of sequels dominated the box office, and Apple launched a new product that looked very cool but was ultimately eclipsed by a better product a year later. Okay, some things never change.
Every year is a big year in tech, of course, but 2004 was an especially big one. And The Verge didn’t exist yet! So, this week, we’ll have stories on the best and most important gadgets and platforms that launched that year and pieces about the cultural events that still affect the way we live now. Basically, we’re going to blog like it’s 2004.
Neat idea, given that it can feel difficult to find news in a tech world where the new things aren’t that thrilling: give people a bit of nostalgia by rewinding to 20 years ago.
Unfortunately, they chose the wrong year for the best iPod, which was the 2005 iPod nano. Tolerable miss, though.
Android 15’s security and privacy features are the update’s highlight • Ars Technica
Kevin Purdy:
Android 15 started rolling out to Pixel devices Tuesday and will arrive, through various third-party efforts, on other Android devices at some point. There is always a bunch of little changes to discover in an Android release, whether by reading, poking around, or letting your phone show you 25 new things after it restarts.
In Android 15, some of the most notable involve making your device less appealing to snoops and thieves and more secure against the kids to whom you hand your phone to keep them quiet at dinner. There are also smart fixes for screen sharing, OTP codes, and cellular hacking prevention, but details about them are spread across Google’s own docs and blogs and various news site’s reports.
Here’s what is notable and new in how Android 15 handles privacy and security.
Happy Android 15 for all who celebrate, but I just wanted to observe: Android has been on a long journey towards more privacy and security, while Apple’s iOS has been forced to open up more and more (mostly by the EU, but the effects go wider), with increased customisation that used to be the province of Android. This mirrors the way in which their notifications and so on have been converging for absolutely years.
Weight-loss drugs cut drug and alcohol abuse, according to new study • WSJ
Dominic Chopping:
Drugs such as Novo Nordisk’s blockbuster Ozempic can cut drug and alcohol abuse by up to 50% according to a new study, adding to mounting evidence that the drugs yield health benefits beyond diabetes and weight loss.
In a study published Thursday in scientific journal Addiction, around 500,000 people with a history of opioid use disorder were analyzed, of which just more than 8,000 were taking either GLP-1 drugs such as Ozempic or the similar GIP class of drugs that Eli Lilly’s Mounjaro belongs to.
GLP-1 drugs work by mimicking a gut hormone to control blood sugar and suppress appetite while GIP medications take a dual-target approach by mimicking both the GLP-1 hormone and a second gut hormone that is believed to enhance the drug’s effectiveness.
The study found that those taking the drugs had a 40% lower rate of opioid overdose compared with those who didn’t.
Similarly, an analysis of more than 5,600 people with a history of alcohol use disorder and who took the drugs showed they had a 50% lower rate of intoxication compared with those who didn’t take them.
“Our study… reveals the possibilities of a novel therapeutic pathway in substance use treatment,” the study’s lead researcher Fares Qeadan and co-authors of the research report Ashlie McCunn and Benjamin Tingey said.
The researchers, from Loyola University Chicago, said the study opens avenues for more comprehensive and effective treatment strategies for opioid and alcohol use disorders.
It’s an interesting finding, but it’s hard not to have a sneaking suspicion that the people who can stick to an Ozempic regime are just less likely to overdose or get intoxicated, and that the drug isn’t necessarily the primary cause. (The paper does admit, near the end, the “the data limits the ability to assume causality”.)
Spanish mother and daughter train bacteria to restore church frescoes • Reuters
Horaci Garcia and Eva Manez:
As Spanish microbiologist Pilar Bosch was casting around for a subject to investigate for her PhD in 2008, she stumbled across a paper suggesting that bacteria, her field, could be used in art restoration, her mother’s own area of expertise.
At that same moment, her mother – Pilar Roig – was struggling to restore 18th-century paintings by Antonio Palomino in one of the oldest churches in Spain’s third city, Valencia.
She was finding it particularly difficult to remove glue that had been used to pull the frescoes from the walls of Santos Juanes church during restoration work in the 1960s.
“My mother had a very difficult problem to solve and I found a paper about bacteria used to clean frescoes in Italy,” Bosch, 42, said.
She did her PhD on that project. And more than a decade later, daughter and mother have joined forces on a €4m ($4.46m) project, funded by local foundations, to use some of the techniques to restore the artworks in Valencia.
The microbiologist trains bacteria by feeding them samples of the glue which was made from animal collagen. The bacteria then naturally produce enzymes to degrade the glue.
The family team then mixes the bacteria with a natural algae-based gel and spread it on the paintings – which were taken from the walls in the 1960s, then nailed back on, still covered in glue.
After three hours, the gel is removed, revealing glue-free paintings.
“In the past, we used to work in a horrible manual way, with warm water and sponges that took hours and damaged the painting,” said Roig, now 75, whose father and grandfather along with other relatives also worked in art conservation.
Petition to get them to do this for wallpaper glue.
| • 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?Read Social Warming, my latest book, and find answers – and more. |
Errata, corrigenda and ai no corrida: No individual links for this edition because Pinboard’s API was down, so this was done by hand 😭
The Overspill: when there's more that I want to say 