Start Up No.2247: free our data Keir Starmer!, Google News likes AI spam, making batteries with salt, AirPod cameras?, and more


Does the relative scarcity of lefthanded people stem from a sort of social pressure? New research offers clues. CC-licensed photo by Alex Lewis 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 coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.


A selection of 10 links for you. Dexter and Sinister. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.


TechScape: here’s four ways a new Labour government could use tech to boost Britain • The Guardian

Alex Hern:

»

In March 2006, back when the Guardian technology section was a physical supplement in Thursday’s newspaper, we ran a campaign to “free our data”. We wrote about government-owned and approved agencies such as the Ordnance Survey, the UK Hydrographic Office and the Highways Agency collecting data on our behalf. We asked: “Why can’t we get at that data as easily as we can Google Maps?”

The campaign was, over the years that followed, a mixed success. Across the public sector, a new norm was created that government data should generally be made available to the public when possible. It almost certainly influenced the direction of the gov.uk project, putting open data at the heart of the state’s digital footprint, and a glance at the top-level data.gov.uk website shows how much work has been done to that end.

Someone born on the day of that campaign’s launch will be voting for the first time on Thursday. Yet some of the most valuable pieces of our digital infrastructure are still locked up, behind restrictive terms or expensive paywalls.

The Postcode Address File (PAF) is one example. It holds 1.8m postcodes and almost 30m postal addresses, and is the ground truth for how we navigate the country. It was privatised along with Royal Mail, but remains tightly controlled by the state, with access charges regulated by Ofcom and a unique license for the public sector to use it at a flat cost.

Freeing our data is the right thing to do, but successive governments have viewed it as expensive: giving up a valuable revenue stream, in the name of abstract concepts. But a Labour party looking for growth and state renewal over the next few years should recognise that if a government dataset is valuable enough to be worth charging for, it’s even more valuable if it can be built on, improved and reused.

«

The sale of the PAF was a disgrace – but as the article points out, there are lots of big potential wins which can spark growth through making more data available on non-restrictive licences. Free Our Data lives! And it’s old enough to buy drinks!
unique link to this extract


Google Search ranks AI spam above original reporting in news results • WIRED

Reece Rogers:

»

As reported by 404 Media in January, AI-powered articles appeared multiple times for basic queries at the beginning of the year in Google News results. Two months later, Google announced significant changes to its algorithm and new spam policies, as an attempt to improve the search results. And by the end of April, Google shared that the major adjustments to remove unhelpful results from its search engine ranking system were finished. “As of April 19, we’ve completed the rollout of these changes. You’ll now see 45% less low-quality, unoriginal content in search results versus the 40% improvement we expected across this work,” wrote Elizabeth Tucker, a director of product management at Google, in a blog post.

Despite the changes, spammy content created with the help of AI remains an ongoing, prevalent issue for Google News.

“This is a really rampant problem on Google right now, and it’s hard to answer specifically why it’s happening,” says Lily Ray, senior director of search engine optimization at the marketing agency Amsive. “We’ve had some clients say, ‘Hey, they took our article and rehashed it with AI. It looks exactly like what we wrote in our original content but just kind of like a mumbo-jumbo, AI-rewritten version of it.’”

At first glance, it was clear to me that some of the images for Syrus’ blogs were AI generated based on the illustrations’ droopy eyes and other deformed physical features—telltale signs of AI trying to represent the human body.

Now, was the text of our article rewritten using AI? I reached out to the person behind the blog to learn more about how they made it and received confirmation via email that an Italian marketing agency created the blog. They claim to have used an AI tool as part of the writing process. “Regarding your concerns about plagiarism, we can assure you that our content creation process involves AI tools that analyze and synthesize information from various sources while always respecting intellectual property,” writes someone using the name Daniele Syrus over email.

«

Google News has had a longstanding problem with recency bias: a more recent but purely rewritten article will rank higher than the original, despite adding nothing to it. Now AI is making it worse.
unique link to this extract


Cleanview: see the energy transition in real time

Michael Thomas built this site, which tracks energy installs across the US and provides lots of interesting data:

»

Cleanview crunches millions of data points and tracks thousands of projects each month so you don’t have to.

«

Wind, solar, battery, plus all the other energy installations. It’s a fascinating resource for tracking how things are changing – because they are changing: just look, for example, at battery installation in California.
unique link to this extract


AI Overview study for 8,000 keywords in Google Search • Advanced Web Ranking

Philip Petrescu with the TL;DR on the new AI stuff when you do a Google query:

»

• The average AI Overview is 169 words and 912 pixels long
• Only 12.4% of the analyzed keywords display an AI Overview
• A Featured Snippet is showing on 17.6% of the analyzed keywords
• On average, AI Overviews appear alongside Featured Snippets in 7.4% of cases. For the Health niche, they show up together the most often (34.9% of queries)
• AI Overviews contain 7.2 links on average when expanded
• 33.4% of AI Overview links rank in that query’s top 10 organic results
• 46.5% of the URLs included in AI Overviews rank outside the top 50 organic results
• Five-word queries trigger an AI Overview most frequently
• Keywords from the Health and Safety niches are more likely to trigger AI Overviews
• No AI Overviews show up for brand related queries
• Navigational intent keywords are less likely to display AI Overviews
• Google Ads are displayed in 28.3% keywords that trigger AI Overviews
• From all the keywords that trigger AI Overviews, Ads at the top of the SERP appear for 8.7% of keywords. Ads at the bottom are displayed for 19.5% of these keywords
• Shopping Ads are almost never seen together with AI Overviews and when they are, they always appear below the AI Overview.

«

There’s much more if you want the in-depth version. One other point from lower down: “Do AI Overviews impact the visibility of the first organic result?”

And the answer is: yes, they do. Significantly.
unique link to this extract


Google emissions jump nearly 50% over five years as AI use surges • FT

Camilla Hodgson and Stephen Morris:

»

Google’s greenhouse gas emissions have surged 48% in the past five years due to the expansion of its data centres that underpin artificial intelligence systems, leaving its commitment to get to “net zero” by 2030 in doubt.

The Silicon Valley company’s pollution amounted to 14.3mn tonnes of carbon equivalent in 2023, a 48% increase from its 2019 baseline and a 13% rise since last year, Google said in its annual environmental report on Tuesday.

Google said the jump highlighted “the challenge of reducing emissions” at the same time as it invests in the build-out of large language models and their associated applications and infrastructure, admitting that “the future environmental impact of AI” was “complex and difficult to predict”.

Chief sustainability officer Kate Brandt said the company remained committed to the 2030 target but stressed the “extremely ambitious” nature of the goal.

“We do still expect our emissions to continue to rise before dropping towards our goal,” said Brandt.

She added that Google was “working very hard” on reducing its emissions, including by signing deals for clean energy. There was also a “tremendous opportunity for climate solutions that are enabled by AI”, said Brandt.

«

Wait, though – Google didn’t start introducing its AI search until earlier this year, surely? If this is what happens when it’s just gearing up for wide use of AI, which we know consumes far more energy than normal search, it’s a really troubling sign.
unique link to this extract


Northvolt: the battery company seeking to make batteries with salt • Climate Home News

Martin Gelin and Meli Petersson Ellafi:

»

In recent years, concerns have grown over the social and environmental harms of extracting and refining battery minerals. Reserves of lithium, nickel and cobalt are concentrated in a handful of countries, making them prone to geopolitical and trade disruptions, and subject to intense competition. Meanwhile, China dominates the lithium-ion battery supply chain, controlling 70% of global lithium refining capacity.

Sodium-based batteries offer a solution to the electric battery supply chain challenges, particularly for Western countries seeking to reduce their dependence on China for cleantech.

According to 2023 analysis by BloombergNEF, sodium batteries could displace 272,000 tonnes of lithium demand by 2035, equivalent to about 7% of the overall market projected for that year.

“The real value of sodium-ion batteries is the potential to build a European supply chain,” said Iola Hughes, research manager at London-based battery consultancy Rho Motion. “In the US and the EU, there is growing pressure to decouple from China and to build domestic supply, and sodium-ion batteries could have strong potential in that transition.”

In January, the European Investment Bank backed the company with over $1bn in financing, citing Northvolt’s ability to create the first fully integrated circular lithium-ion battery production facility outside Asia.
At the time, EU Commission Vice President Maroš Šefčovič, in charge of the European Green Deal, said the battery industry was of “strategic importance and a key battleground for global competitiveness”.

“Northvolt, our battery pioneer, showcases that the EU has what it takes to build an innovative, sustainable and globally competitive battery ecosystem,” he said.

«

Sodium is plentiful (available at a sea near you), which makes it very attractive for batteries where size isn’t an issue.
unique link to this extract


Figma disables AI app design tool after it copied Apple’s Weather app • 404 Media

Emanuel Maiberg:

»

The design tool Figma has disabled a newly launched AI-powered app design tool after a user showed that it was clearly copying Apple’s weather app. 

Figma disabled the feature, named Make Design, after CEO and cofounder of Not Boring Software Andy Allen tweeted images showing that asking it to make a “weather app” produced several variations of apps that looked almost identical to Apple’s default weather app. 

“Within hours of seeing this tweet, we identified the issue, which was related to the underlying design systems that were created. Ultimately it is my fault for not insisting on a better QA process for this work and pushing our team hard to hit a deadline for Config,” Figma CEO Dylan Field said on Twitter. Config is Figma’s annual conference where it showcased Make Design. “I have asked our team to temporarily disable the Make Design feature until we are confident we can stand behind its output. The feature will be disabled when our US based team wakes up in a few hours, and we will re-enable it when we have completed a full QA pass on the underlying design system.”

«

Wonder how that “QA pass” is going to work, exactly. The AI is, one presumes, a generative AI system, which means you have to tear up its entire learning matrix in order to take even one app out. And you’d want to take a fair number of apps out. Which would probably only leave the bad ones. Which you don’t want to have in.

One hopes this is a free upgrade.
unique link to this extract


Why most of us are right-handed, and prefer to view faces on the left • Science Museum Blog

Roger Highfield:

»

Once upon a time, scientists thought biases in behaviour were unique to humans and reflected differences in the dominance of the left and right brain hemispheres, each of which controls the opposite side of the body.   

But, as has so often been the case, research in recent decades has shown plenty of handedness in other creatures and the way that they behave. 

Studies conducted on animals suggests that brain hemisphere biases evolved because they allow the two sides of the brain to simultaneously do different things without the animals becoming muddled, or the two hemispheres conflicting.   

So being right- or left-handed free up some brain power to make animals more efficient at finding food and, in general, surviving, for instance turning the right way to stay in a shoal or a flock to cut the risk of being picked off by predators, so they can pass their genes on. 

The museum’s Live Science residency allowed the team to use museum visitors to study associations between the degree of hand bias and performance as well as direction of biases and social ability that had already been studied in animal research. 

The team measured handedness using a timed colour-matching pegboard task, along with images of faces expressing different emotion (such as surprise or anger) presented on the right- or left-hand side of a screen. 

People with mild to moderate strength hand bias (left or right) placed more colour-matched pegs correctly than those with a strong or weak bias, showing that, in humans, extremes of handedness may limit our flexibility, unlike wild animals. 

«

The full study, based on 1,600 experiments on live humans (visitors to the Science Museum) is published in Nature.
unique link to this extract


Apple rumoured to be working on camera-powered AirPods • Gizmodo

Harri Weber:

»

Apple is reportedly bringing cameras to AirPods, ear hairs be damned.

The $3 trillion company intends to mass-produce revamped earbuds with built-in infrared cameras by 2026, according to a new report from analyst and longtime Apple insider Ming-Chi Kuo. The cameras could help Apple shore up its current and future augmented-reality headsets with enhanced spatial audio features, the analyst wrote.

Citing a supply-chain survey, Kuo indicated that pairing these enhanced buds with Vision Pro goggles could make Apple’s spatial-computing experience more lifelike. For example, “if users turn their heads to look in a specific direction, the sound source in that direction can be emphasized,” said the analyst.

For folks not interested in dropping thousands on an Apple headset, the IR cameras could offer other perks, including bringing “in-air” gestures to AirPods, per Kuo. The IR module could be similar to the receiver that powers Apple’s facial recognition feature, FaceID.

The analyst’s report follows an earlier story from Bloomberg’s Mark Gurman, which noted that Apple was looking into the idea of camera-powered AirPods.

«

Seems a pretty weird idea. The Ming-Chi Kuo piece contains a paragraph saying “The IR camera can detect environmental image changes, potentially enabling in-air gesture control to enhance human-device interaction. It is worth noting that Apple has filed related patents in this area”. Still seems weird. But also a long way off.
unique link to this extract


Three million iOS and macOS apps were exposed to potent supply-chain attacks • Ars Technica

Dan Goodin:

»

Vulnerabilities that went undetected for a decade left thousands of macOS and iOS apps susceptible to supply-chain attacks. Hackers could have added malicious code compromising the security of millions or billions of people who installed them, researchers said Monday.

The vulnerabilities, which were fixed last October, resided in a “trunk” server used to manage CocoaPods, a repository for open source Swift and Objective-C projects that roughly 3 million macOS and iOS apps depend on. When developers make changes to one of their “pods”—CocoaPods lingo for individual code packages—dependent apps typically incorporate them automatically through app updates, typically with no interaction required by end users.

“Many applications can access a user’s most sensitive information: credit card details, medical records, private materials, and more,” wrote researchers from EVA Information Security, the firm that discovered the vulnerability. “Injecting code into these applications could enable attackers to access this information for almost any malicious purpose imaginable—ransomware, fraud, blackmail, corporate espionage… In the process, it could expose companies to major legal liabilities and reputational risk.”

The three vulnerabilities EVA discovered stem from an insecure verification email mechanism used to authenticate developers of individual pods. The developer entered the email address associated with their pod. The trunk server responded by sending a link to the address. When a person clicked on the link, they gained access to the account.

In one case, an attacker could manipulate the URL in the link to make it point to a server under the attacker’s control. The server accepted a spoofed XFH, an HTTP header for identifying the target host specified in an HTTP request. The EVA researchers found that they could use a forged XFH to construct URLs of their choice.

…CocoaPods maintainers disclosed and patched the vulnerabilities last October. At the time, they said they weren’t aware of any active attempts to exploit the vulnerabilities. They did, however, confirm that the scenarios described by the researchers were plausible.

«

The food chain of code packages really does mirror the complexity of life. But with code, of course.
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?

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


Errata, corrigenda and ai no corrida: none notified

Start Up No.2246: EU threatens Meta with DMA fine, the tech unemployment line, Florida man shoo(t)s drone, and more


A new scheme is injecting radioactive substances into rhinos’ horns so poaching can be detected at Customs. CC-licensed photo by .waldec 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 coming this week at the Social Warming Substack on Friday at 0845 UK time. Free signup.


A selection of 9 links for you. Not horny. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.


Brussels follows up Apple charges with Meta ‘pay or consent’ case • FT

Javier Espinoza:

»

The EU has charged Facebook’s parent Meta with breaking the bloc’s landmark digital rules, only a week after it pressed a similar case against Apple.

The European Commission, the EU’s executive body, is exercising new powers granted by the Digital Markets Act — legislation aimed at improving consumer choice and opening up markets for European start-ups to flourish. The tech giants had to comply from March this year. 

In preliminary findings issued on Monday, Brussels regulators said they were worried about Meta’s “pay or consent” model. Facebook and Instagram users can currently opt to use the social networks for free while consenting to data collection, or pay not to have their data shared.

The regulators said that the choice presented by Meta’s model risks giving consumers a false alternative, with the financial barrier potentially forcing them to consent to their personal data being tracked for advertising purposes.

The Financial Times was first to reveal the commission’s move earlier on Monday.

Under the bloc’s new digital rules, tech giants must gain consent from users “when they intend to combine or cross-use their personal data across different core platform services”, the EU said in March, when it opened compliance investigations against Meta and other Big Tech groups.

The EU executive said that Meta “users who do not consent should still get access to an equivalent service which uses less of their personal data, in this case for the personalisation of advertising”.

«

In October last year I wrote about how Meta’s proposed charges were about four times higher than was justified by their advertising revenue in the EU. However, that doesn’t seem to be the basis of the EU’s complaint; it simply wants a free service but with less data collection. Basically, challenging Meta to accept lower profits (if we presume that profits are directly dependent on amount of personal data collected).
unique link to this extract


Laid-off tech workers advised to sell plasma and personal belongings • SF Gate

Ariana Bindman:

»

Nina McCollum has been laid off so many times that the 55-year-old is basically an unofficial expert. That’s how she describes herself, at least. 

The marketing writer, who went viral in 2019 for documenting how she submitted over 200 applications during her two-year unemployment period, eventually landed her dream job at a major human resources tech company in the Bay Area. But then, in March 2023, she was let go — and suddenly back at square one. 

McCollum is not alone. Over the past two years, major tech companies in the Bay Area have haemorrhaged high-salaried workers, sending a chill throughout an industry that once seemed untouchable. Meta has let go of at least 21,000 workers, while Google has handed pink slips to hundreds of employees across San Francisco, Sunnyvale and Mountain View. Though the state government boasts about California’s growing economy and low unemployment rate, multiple people who spoke with SFGATE painted a bleak picture.

…When tech recruiter Irene Nexica was laid off in March 2023, she was “immediately transported to the world of being poor,” she told SFGATE. Since then, she’s worked a variety of “survival jobs,” ranging from catering to online retail. She also works as a career coach at a federally funded program that helps people train for employment. Anyone who makes under $35 an hour qualifies, she said. “We’re seeing a flood of tech people come in because they are super desperate now,” she continued.

Companies, sensing their desperation, are now taking advantage of it. When Nexica applied for a nonprofit recruiting job, she found that 1,400 people had applied already — and for significantly less money than the role usually pays. Instead of a salary of around $125,000, this one was offering between $80,000 and $90,000. She also found that the people in charge of hiring for roles would message her about so-called opportunities, only to ghost her without explanation.

«

unique link to this extract


Alzheimer’s scientist indicted for allegedly falsifying data in $16m scheme • Ars Technica

Beth Mole:

»

A federal grand jury has indicted an embattled Alzheimer’s researcher for allegedly falsifying data to fraudulently obtain $16m in federal research funding from the National Institutes of Health for the development of a controversial Alzheimer’s drug and diagnostic test.

Hoau-Yan Wang, 67, a medical professor at the City University of New York, was a paid collaborator with the Austin, Texas-based pharmaceutical company Cassava Sciences. Wang’s research and publications provided scientific underpinnings for Cassava’s Alzheimer’s treatment, Simufilam, which is now in Phase III trials.

Simufilam is a small-molecule drug that Cassava claims can restore the structure and function of a scaffolding protein in the brain of people with Alzheimer’s, leading to slowed cognitive decline. But outside researchers have long expressed doubts and concerns about the research.

In 2023, Science magazine obtained a 50-page report from an internal investigation at CUNY that looked into 31 misconduct allegations made against Wang in 2021. According to the report, the investigating committee “found evidence highly suggestive of deliberate scientific misconduct by Wang for 14 of the 31 allegations,” the report states.

The allegations largely centered around doctored and fabricated images from Western blotting, an analytical technique used to separate and detect proteins. However, the committee couldn’t conclusively prove the images were falsified “due to the failure of Dr. Wang to provide underlying, original data or research records and the low quality of the published images that had to be examined in their place.”

«

At least with this sort of (alleged) fraud you can see the motivation: all that funding. Though if it’s faked research, won’t the failure of the therapy become clear in the trials? Perhaps in that inevitability one just says “oh well, didn’t work” and drives back to the laboratory in the Ferrari.
unique link to this extract


Supreme Court protects the future of content moderation • The Verge

Lauren Feiner:

»

On Monday, the Supreme Court issued its decision in Moody v. NetChoice and NetChoice v. Paxton, two consequential cases about the future of speech on the internet. The court explicitly extended First Amendment protections to how social media platforms organize, curate, and moderate their feeds, drawing a comparison between internet content moderation and “traditional publishers and editors.”

The decision elaborates that the compilation and curation of “others’ speech into an expressive product of its own” is entitled to First Amendment protection and that “the government cannot get its way just by asserting an interest in better balancing the marketplace of ideas.” 

The NetChoice cases concern a pair of similar laws in Florida and Texas that aimed to limit how large social media companies could moderate content on their sites. The legislation took shape after conservative politicians in both states criticized major tech companies for allegedly exerting bias against conservative viewpoints. Tech industry groups NetChoice and the Computer & Communications Industry Association sued to block both laws. Appeals courts in each state came to different conclusions about whether the statutes could be upheld, setting up the Supreme Court to make the final call.

The Supreme Court vacated both of the appeals court decisions, ruling that neither court adequately analyzed “the facial First Amendment challenges” to the laws — that is, whether the social media content moderation laws in Florida and Texas would always be unconstitutional in all applications. The court sent the cases back down to the lower courts to reconsider.

Under the new Supreme Court decision, content moderation is generally protected by the First Amendment.

«

Hurrah, democracy is save–oh, really?
unique link to this extract


Florida man charged after shooting at Walmart delivery drone • USA Today

Saleen Martin:

»

A Florida man is facing multiple charges after authorities say he shot at a Walmart delivery drone.

The shooting happened in Clermont, about 26 miles west of Orlando. The Lake County Sheriff’s Office received a complaint about the ordeal Wednesday, the office said in a Facebook post. According to the caller, a bullet hole was found in the payload the drone was carrying. 

Witnesses pointed authorities to 72-year-old Dennis Winn, who interviewed with officials and later admitted to shooting at the drone one time with a 9mm pistol, the sheriff’s office said.

In police bodycam footage, Winn told police he tried to shoo the drone off and it didn’t go away, so he shot at it. “I fired one round at it,” he said in the footage. “They say I hit it so I must be a good shot, or else it’s not that far away … I’m going to wind up having to find a real good defense lawyer.”

He was taken into custody and charged with shooting at an aircraft, criminal mischief damage over $1,000 and discharging a firearm in public or residential property, according to the sheriff’s office.

«

He tried to “shoo off the drone”? And he’s younger than both the contenders for the US presidency. What a country.
unique link to this extract


Do be a quitter! How I broke my exercise streak – and smashed my fitness goals • The Guardian

Phil Daoust:

»

that is nothing compared with my boneheadedness 13 or 14 years ago, when I lived in the mountains and used to swim in a long, chilly lake. It was usually just two lengths, and I took them slowly, but it added up to about 3-4km – say 150 lengths of a typical leisure-centre pool. The season was short, and I usually stopped once the temperature began to drop. But this time I got in the water well into autumn. I didn’t own a wetsuit, and this was not the kind of place that had lifeguards. Even in summer, I was sometimes the only person in the water.

After 2-3km, breaststroking my way back to where I had left the car, I began to shiver. I know now that this is one of the first signs of hypothermia. The sensible thing would have been to get out of the lake and walk, but that didn’t even occur to me. It wasn’t in the plan, and perhaps the cold was getting to my brain too. So I swam on for another 40 or 50 minutes, trembling all the way. I survived (obviously), but the more I think about it, the more I see how lucky I was. If I had got into trouble, no one would have seen me, let alone saved me.

Now that I have realised that sometimes the best thing to do with a plan is to ignore it, I have been delighted to discover a lot of experts feel the same. They just don’t make a fuss about it. Take Michael Ulloa, an Edinburgh-based performance nutritionist and personal trainer. “We’re constantly told that if we can’t stick to a plan 100%, then we have somehow failed,” he says. “This couldn’t be further from the truth. It is messing people up. When we deviate from a plan, we shouldn’t overthink it. We should ask why this deviation happened and what we can do to limit the chances of it happening again. Did we try to take on more than we can chew? Are we not enjoying our current training programme? Or maybe we were just tired and we needed to give ourselves a day or two.”

“It’s easy to overanalyse and be over self-critical, but there really is no need,” he says. “Most of us are not professional athletes, we are simply everyday people doing our best – and sometimes we need to take a day when it feels too much.”

«

Giving yourself “permission to quit” turns out to be very wise. Quite possibly there’s survivor bias in it. Which means it’s a good thing, not a bad one.
unique link to this extract


Generative AI is not going to build your [software] engineering team for you • Stack Overflow

Charity Majors:

»

It is really, really tough to get your first role as an engineer. I didn’t realize how hard it was until I watched my little sister (new grad, terrific grades, some hands on experience, fiendishly hard worker) struggle for nearly two years to land a real job in her field. That was a few years ago; anecdotally, it seems to have gotten even harder since then.

This past year, I have read a steady drip of articles about entry-level jobs in various industries being replaced by AI. Some of which absolutely have merit. Any job that consists of drudgery such as converting a document from one format to another, reading and summarizing a bunch of text, or replacing one set of icons with another, seems pretty obviously vulnerable. This doesn’t feel all that revolutionary to me, it’s just extending the existing boom in automation to cover textual material as well as mathy stuff.

…People act like writing code is the hard part of software. It is not. It never has been, it never will be. Writing code is the easiest part of software engineering, and it’s getting easier by the day. The hard parts are what you do with that code—operating it, understanding it, extending it, and governing it over its entire lifecycle.

A junior engineer begins by learning how to write and debug lines, functions, and snippets of code. As you practice and progress towards being a senior engineer, you learn to compose systems out of software, and guide systems through waves of change and transformation.

Sociotechnical systems consist of software, tools, and people; understanding them requires familiarity with the interplay between software, users, production, infrastructure, and continuous changes over time. These systems are fantastically complex and subject to chaos, nondeterminism and emergent behaviors. If anyone claims to understand the system they are developing and operating, the system is either exceptionally small or (more likely) they don’t know enough to know what they don’t know. Code is easy, in other words, but systems are hard.

«

unique link to this extract


First radioactive rhino horns to curb poaching in South Africa • Phys.org

Zama Luthuli:

»

South African scientists on Tuesday injected radioactive material into live rhino horns to make them easier to detect at border posts in a pioneering project aimed at curbing poaching.

The country is home to a large majority of the world’s rhinos and as such is a hotspot for poaching driven by demand from Asia, where horns are used in traditional medicine for their supposed therapeutic effect.

At the Limpopo rhino orphanage in the Waterberg area, in the country’s northeast, a few of the thick-skinned herbivores grazed in the low savannah.

James Larkin, director of the University of the Witwatersrand’s radiation and health physics unit who spearheaded the initiative, told AFP he had put “two tiny little radioactive chips in the horn” as he administered the radioisotopes on one of the large animals’ horns.

The radioactive material would “render the horn useless… essentially poisonous for human consumption” added Nithaya Chetty, professor and dean of science at the same university.

The dusty rhino, put to sleep and crouched on the ground, did not feel any pain, Larkin said.

The radioactive material’s dose was so low it would not impact the animal’s health or the environment in any way, he said.

In February the environment ministry said that, despite government efforts to tackle the illicit trade, 499 of the giant mammals were killed in 2023, mostly in state-run parks. This represents an 11% increase over the 2022 figures.

«

What a brilliant idea. Especially if you could persuade the poachers that the radioactivity is dangerous.
unique link to this extract


Influencers aren’t getting famous like they used to • Glamour

Stephanie McNeal:

»

It used to go like this: An influencer or content creator would spend anywhere from a very short to a respectably decent amount of time building up their content and fan base. Maybe they were family vloggers, or prank YouTubers like the Paul brothers, or makeup artists or fashionistas. Then, there’d be a tipping point. Suddenly, these people were being written about in mainstream media like People and Us Weekly, landing roles in movies and on television shows, and attending high-class events.

Remember in 2021, when influencers like Chamberlain and Rae attended the Met Gala and everyone freaked out? This year, Chamberlain attended the Met again (she’s actually become somewhat of a fixture and works the red carpet for Vogue), but the rest of them have dried up. And even if Anna Wintour wanted to invite a fresh crop of internet talent, who would she choose? Can you name anyone in the past year who has ascended in a major way?

The last true influencer to truly “make it” in this way was Alix Earle, who became a household name seemingly overnight in early 2023 and has done quite well for herself. That’s not just my opinion. Sophie, who runs a business consulting on social media called Pretty Little Marketer, identified Earle as the most recent to “blow up” as well.

“I’d say she’s exceeded well past the title of influencer to global celebrity status,” she says.

This doesn’t mean that people aren’t doing well on social media. The problem is that too many people are doing well on social media. It’s not unique these days to have hundreds of thousands—or even millions—of followers, to have an agent, or to get lucrative brand deals. But with a huge pool of creators who are hard to differentiate from each other, the hard part seems to be standing out at all.

“There are many influencers who still make it ‘big’ in terms of opportunity and following, but we’ve seen fewer boom into mainstream media as we have in the past,” Sophie says.

«

Influencing’s too popular? It’s a weird world.
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?

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


Errata, corrigenda and ai no corrida: none notified

Start Up No.2245: user-replaceable smartphone batteries by 2027?, Facebook’s pro-Reform Russians, election times!, and more


Move over, human beauty pageant contestants: your AI replacements are arriving. CC-licensed photo by Paul Chin 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 9 links for you. Please don’t cry. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.


‘Miss AI’ beauty pageant and the complicated quest for the ‘perfect’ woman • CNN

Issy Ronald:

»

Ten women participating in a beauty pageant is nothing new. Some pose candidly, some play to the camera, their beauty forever frozen in this moment in time. Like many other pageants held in countries around the world, the contestants are young, thin and embody many of the standards defining traditional “beauty.”

But that is where the similarities to a traditional beauty pageant end. None of these women are real — everything about them, even the emotion that flickers across their faces, is generated by artificial intelligence (AI), for the world’s first ever AI beauty pageant. Each has a creator or team of creators, who use programmes like Open AI’s DALL·E 3, Midjourney or Stable Diffusion to generate images of the women from text prompts.

These ten contestants have been selected from a pool of more than 1,500 entrants to make the final of “Miss AI,” scheduled to be held at the end of June and broadcast online by its organizers “The World AI Creator Awards.”

For those involved, the event is an opportunity to showcase and demystify the technology’s extraordinary abilities. But for others, it represents a further proliferation of unrealistic beauty standards often linked to racial and gender stereotypes and fueled by the ever-increasing number of digitally enhanced images online.

“I think we’re starting to increasingly lose touch with what an unedited face looks like,” Dr Kerry McInerney, a research associate at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge, told CNN in a video interview.

Each of the contestants has a unique and distinctive personality, as well as face. One red-haired, green-eyed avatar named Seren Ay poses for Instagram photos as she travels around the world and through time, appearing next to Turkey’s first president Kemal Ataturk, on the Oscars red carpet or wandering through the neon-lit streets of Kyoto, Japan at night.

«

The “contestants” all have the correct number of fingers, at least in the still pictures. I guess we’d need to wait for the video round.
unique link to this extract


Making sense of the EU’s fight for user-replaceable smartphone batteries • The Verge

Jon Porter:

»

If you’ve been online in the past week, you’ve probably seen one or two headlines about the European Union voting in favor of easy-to-replace batteries in smartphones by around 2027. That’s based on a June 14th vote in which the European Parliament voted overwhelmingly in favor of an agreement that would overhaul the rules around batteries in the bloc. 

The good news is that those headlines are fundamentally accurate; the EU is moving forward with regulation designed to require smartphones to have batteries that are easier to replace, to the benefit of the environment and end users. But this being the European Union, there’s a lot more going on behind the scenes. And it’s these details that could have a significant impact on how and when manufacturers will actually have to comply.

For starters, the widely cited 2027 deadline for offering smartphones with more easily replaceable batteries isn’t quite the whole story, according to Cristina Ganapini, coordinator of Right to Repair Europe. That’s because there’s another piece of legislation currently working its way through the EU’s lawmaking process called the Ecodesign for Smartphones and Tablets. It contains similar rules about making smartphone batteries easier to replace and is expected to come into effect earlier in June or July 2025. So by the time 2027 rolls around, some smartphone manufacturers may have already been selling devices with user-replaceable batteries in the EU for over a year.

«

Will Apple claim that its repair kits, which people can hire to do various kinds of work on their own phones, fulfil this “user-replaceable” description, I wonder?
unique link to this extract


Revealed: the tech entrepreneur behind a pro-Israel hate network • The Guardian

Jason Wilson:

»

A prime mover behind the Shirion Collective, a conspiracy-minded, pro-Israel disinformation network seeking to shape public opinion about the Gaza conflict in the US, Australia and the UK, is a tech entrepreneur named Daniel Linden living in Florida who co-wrote a guidebook for OnlyFans users, the Guardian can reveal.

Shirion has harassed pro-Palestinian activists, including many Jews, offered bounties for the identity of pro-Palestinian protesters, spread conspiracy narratives centered on figures like George Soros, and boasted of an AI-surveillance platform but offered few concrete details of how the technology functions.

The Guardian investigation used public records and open source materials to corroborate information originally provided by the White Rose Society, an Australian anti-fascist research collective.

Linden set up Shirion’s crowdfunding efforts, appears to play a central role in operating the network’s social media accounts, and coordinates the group’s efforts on a Telegram channel. Public records and online materials indicate he lives in Gainesville, Florida, but he has also had recent stints in Durango, Colorado, and Medellin, Colombia.

The Guardian emailed Linden at several addresses associated with him and his business ventures, and attempted to contact him via phone, text, a direct message on Reddit and a post tagging an X account associated with one of his ventures seeking comment on this reporting, but received no response.

Heidi Beirich, co-founder and chief strategy officer at the Global Project Against Hate and Extremism, said of Linden’s Shirion campaigning that his apparent “grifting” is common among extremists, “but his ideology seems very confused”.

“Regardless,” she added, “he is spreading hateful messages.”

The revelations shed light on the nature of Shirion, which has been criticized in the US congress and attracted media attention around the world, and its role in pushing back against criticisms of Israel’s conduct in its invasion of Gaza.

«

Notable that this is categorised under news article about the “far right”.
unique link to this extract


Biden aides plotted debate strategy for months. Then it all collapsed • The Washington Post

Tyler Pager:

»

President Biden’s debate prep went fine.

In the sessions, the president still spoke haltingly. He sometimes confused facts and figures. He tripped over words and meandered. Debate prep would not fix his stutter or make him appear any younger, aides knew.

But as Biden boarded Marine One to leave the rustic Camp David presidential retreat for Atlanta, they sought to reassure anxious allies. The president, they said, was prepared and would perform well. Some said the debate might even be boring.

This story is based on conversations with eight individuals involved in or briefed on the president’s debate preparation, all of whom spoke on the condition of anonymity to describe private meetings. The Biden campaign declined to comment.

For a full week, the president sequestered himself at Camp David with more than a dozen aides to prepare for Thursday’s presidential debate with former president Donald Trump. He rehearsed answers, met with policy aides and participated in mock debates, with his personal lawyer, Bob Bauer, playing the part of Trump.

Every topic he was asked about Thursday, he had practiced answers for — including the final one about his age.

So aides were bewildered by his performance. Many felt they had never seen him collapse so dramatically. After all, Biden was a veteran of numerous debates — as a senator, vice-presidential nominee and presidential candidate. And they did not understand why he gave an entirely different answer on the age question than the one they spent more than a week perfecting.

The president did not just stumble over words. He appeared to lose his focus and often was unable to finish sentences. His voice was raspy and thin, and when the debate concluded, first lady Jill Biden appeared to help her husband down the stairs.

«

The debate was at 9pm. Up past his bedtime, basically. CNN had an article ahead of the debate about how each contender was preparing. Trump’s went better on the night, you could say.
unique link to this extract


Tories ‘highly alarmed’ by network of pro-Russian Facebook pages interfering in UK election • ABC News

Michael Workman and Kevin Nguyen:

»

Revelations of foreign interference in the UK election, uncovered by the Australian Broadcasting Corporation (ABC(, have been described as “highly alarming” by the Conservative Party, which will be writing to the Cabinet Office seeking urgent advice about how to combat it.

Ahead of the UK elections, the ABC has been monitoring five coordinated Facebook pages which have been spreading Kremlin talking points, with some posting in support of Nigel Farage’s populist Reform UK party — a key challenger to the Conservatives in the July 4 poll.

The five pages identified by ABC Investigations as being part of a coordinated network appear to have little in common. One page presents itself as a pro-refugee left-wing group, while others reference white supremacist conspiracy theories and use AI-generated images of asylum seekers to stoke anti-immigration fears.

The ABC has been able to link these seemingly disparate pages by examining the location data attached to the pages’ administrators, tracking paid ads, and by analysing the pages’ similar or shared content.

The ABC shared its findings with disinformation experts, who said the network’s activity had the hallmarks of a Russian influence operation.

“For me, it’s Russian,” said AI Forensics head of research Salvatore Romano.

«

You may be surprised to hear that Nigel Farage has rejected this as “cobblers”. Makes a change from actors, though still a form of worker.
unique link to this extract


What time will we know who won? Hour-by-hour guide to election night • The Guardian

Jamie Grierson, Jim Waterson and Ashley Kirk:

»

After months of speculation on when the election might be held, six weeks of actual campaigning, D-day blunders, gambling scandals, smashing the gangs, stopping the boats, surrendering finances, triple-lock-pluses, national service, VAT on private schools, taxes up and taxes down, the election night will soon be upon us. Here’s how it may unfold.

«

Basically: exit polls 10pm, nothing much until 3am, and then mayhem; all over bar the shouting (and weeping) by 7am.
unique link to this extract


What happened when British GQ stopped trying to ‘feed the algorithm’ • Press Gazette

Charlotte Tobitt:

»

GQ‘s European director of audience development, analytics and social, Neha-Tamara Patel, joined in July 2022 and told Press Gazette about a strategy shift that has seen the legacy men’s lifestyle brand move away from quick wins towards more considered content with the aim of a more engaged core audience.

She said: “It was apparent to me that we were doing a lot of what I call ‘feeding the algorithm content’: lots of short-form news, a lot of quick fashion news, all of which was still within GQ’s world but from an audience perspective wasn’t really serving us long term.

“It meant that we had a lot of churn, a lot of people coming in for that quickfire content and then leaving again without really accessing the broader spectrum of what we do as a brand.

“So we really made a conscious decision to slow things down, not necessarily feed the news cycle. We are a lifestyle magazine brand at the end of the day, not business of fashion or anything like that. So we obviously do touch on fashion news, but we try and think about where we can add to the conversation rather than it being like pure reportage. We’re not just about headlines. It’s like, how are we moving the story on and what else are we bringing to the table?”

…Ipsos iris data shared with Press Gazette shows gq-magazine.co.uk had a UK audience reach of 888,117 in May, down 25% in two years. But its total minutes were down by a lesser 12% to 2.2 million, with recovery of 19% in the past year after a slump in May 2023.

Across last year British GQ saw a 47% year-on-year increase in engagement among British users with 71 million engaged minutes in total, according to figures shared by the publisher.

«

What are “engaged minutes”? Though it’s evident that for a monthly magazine to not attempt to keep up with the relentless hour-by-hour news cycle is a far more sensible approach.
unique link to this extract


Einstein and his peers were ‘irrationally resistant’ to black holes. This illustrated story explores why • BBC Future

Ben Platts-Mills wrote the piece, and illustrated it too:

»

Einstein showed that space and time were continually stretched and distorted by masses such as stars and planets, and that this accounted for gravity. The way bodies are drawn together is not due to a “force” attracting them, he argued, but to the “curvature” of the Universe caused by mass. The greater the mass, the more curvature it causes and so the greater the gravitational effect.

When Einstein first published his theory, he hadn’t pinned down the solutions to his own equations, which would have revealed to him the full implications of his discovery. It was another scientist who made this step.

In November 1915, Karl Schwarzschild was an artillery lieutenant in the German army, on the Eastern Front. He read Einstein’s new theory while working at a weather station close to the front line and wrote a letter in response.

His letter supplied the missing solutions and showed how they could be used to model a star’s gravity. One feature of the model, Schwarzschild noted later, was a radius of compression below which a star – or any other spherical mass – would begin to implode indefinitely under its own gravity. If applied to the physics of the real world, this had horrifying implications. It meant that a star would continue collapsing forever, its mass being crushed ever smaller. Its gravitation would become ever more powerful as it insatiably devoured surrounding masses until, finally, it reached the point of “singularity”, a moment where the laws of physics break down, and time and space cease to exist.

Decades later, the Schwarzschild singularity would be recognised as a turning point in theoretical physics – the first time black holes had been hinted at. Schwarzschild himself, however, dismissed the idea as a mathematical artefact.

The critical radius was, Schwarzschild concluded, simply the limit for a star’s compression – the point it would stop collapsing. Instead of discovering black holes, then, he became the first person to reject the evidence for them on principle. We will never know if he might have revised his ideas because he died of an autoimmune disease in 1916. 

Einstein’s reaction to Schwarzschild’s solutions was mixed.

«

Which goes to show: even the greatest minds will dismiss real possibilities as quirks and artefacts because they’re too difficult to contemplate. And even in the midst of a war, people find ideas too tantalising to ignore.
unique link to this extract


The solar industrial revolution is the biggest investment opportunity in history • Casey Handmer’s blog

Handmer is a scientist and entrepreneur who founded Terraform Industries (which aims to generate fuel from air, using solar energy:

»

To a good approximation, even as global population climbs towards 10 billion, the number of humans enduring extreme poverty has fallen in both absolute and relative terms, from nearly universal just 200 years ago to rapidly vanishing today. We are living in the first time in the history of any life form where we can put a finite upper bound on the number of poverty-stricken human years left to be endured by the human race: Almost certainly fewer than 10 billion. This is a lot but it’s a lot fewer than infinity.

This unprecedented improvement in the human condition has been unlocked by social and political innovation, and underwritten by the consumption of copious quantities of cheap energy, almost all of it from fossil coal, oil, and gas.

To a good approximation, oil is the antidote to poverty.

But oil is finite. The good stuff is gone. Fracking is expensive. Most places don’t have oil. Climate and scarcity will force us to use other forms of energy, most of them not as useful as oil. Are we headed for economic difficulties as a result of this? A handful of nations have endured severe energy shortages due to political instability, and it has never gone well for them. North Korea. Cuba. Venezuela. Are we doomed?

No.

Solar photovoltaic (PV) power got cheap, then big, then cheaper, then bigger. Last year, we installed about 460 GW globally. Check out the knee in the curve in 2009! A learning rate of 44% means that the cost falls by 44% for every doubling of production, and production is currently doubling roughly every 18 months.

Here’s a free heretical viewpoint, or at least an early prediction: solar PV is not just a partial substitute for oil, it’s a cheaper and better energy source in every way that matters.

Corollary: Our techno-capital machine is a thermodynamic mechanism that systematically hunts for and then maximally exploits the cheapest energy it can find. When it unlocks cheaper energy, first coal, then oil, then gas, and now solar, it drives up the rate of economic growth, due to an expanded spread between energy cost and application value.

«

I think he’s right about the benefits of solar PV, at least.
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?

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


Errata, corrigenda and ai no corrida: none notified