Start Up No.1899: LinkedIn battles fake Apple and Amazon accounts, AI coding lawsuit, wind turbines with no blades, and more

Mosquitoes are attracted to a specific chemical that humans exude – but different people give off more, making them tasty targets for the bloodsuckers. CC-licensed photo by John Tann 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.

Holiday! I’m having one, because the world has used up its available supply of news in the past couple of weeks, and needs replenishing. So The Overspill will be on a break for three weeks: back on Monday 14 November.

There’s another post coming at the Social Warming Substack at about 0845 UK time. Free signup.

A selection of 11 links for you. Bite me. I’m @charlesarthur on Twitter. Observations and links welcome.

Battle with bots prompts mass purge of Amazon, Apple employee accounts on LinkedIn • Krebs on Security

Brian Krebs:


On October 10, 2022, there were 576,562 LinkedIn accounts that listed their current employer as Apple Inc. The next day, half of those profiles no longer existed. A similarly dramatic drop in the number of LinkedIn profiles claiming employment at Amazon comes as LinkedIn is struggling to combat a significant uptick in the creation of fake employee accounts that pair AI-generated profile photos with text lifted from legitimate users.

Jay Pinho is a developer who is working on a product that tracks company data, including hiring. Pinho has been using LinkedIn to monitor daily employee headcounts at several dozen large organizations, and last week he noticed that two of them had far fewer people claiming to work for them than they did just 24 hours previously.

Pinho’s screenshot below shows the daily count of employees as displayed on Amazon’s LinkedIn homepage. Pinho said his scraper shows that the number of LinkedIn profiles claiming current roles at Amazon fell from roughly 1.25 million to 838,601 in just one day, a 33% drop.

…In late September 2022, KrebsOnSecurity warned about the proliferation of fake LinkedIn profiles for Chief Information Security Officer (CISO) roles at some of the world’s largest corporations. A follow-up story on Oct. 5 showed how the phony profile problem has affected virtually all executive roles at corporations, and how these fake profiles are creating an identity crisis for the businesses networking site and the companies that rely on it to hire and screen prospective employees.

A day after that second story ran, KrebsOnSecurity heard from a recruiter who noticed the number of LinkedIn profiles that claimed virtually any role in network security had dropped seven% overnight. LinkedIn declined to comment about that earlier account purge, saying only that, “We’re constantly working at taking down fake accounts.”


When you think about it, of course this was going to be LinkedIn’s moderation problem. It just happens out of the light, because who realises what’s going on at LinkedIn?
unique link to this extract

Copilot: the next stage in the AI copyright wars? • TechnoLlama

Andres Guadamuz:


Copilot is an AI tool that writes code based on prompts. The program has been trained on the corpus of code submitted to the open source software repository Github, and it uses OpenAI’s Codex.

Almost from the start, Copilot has proven to be controversial, some people complained that this was a violation of open source principles (and potentially infringing copyright), yet it appears to be widely used by some developers. According to Github the tool has been used by 1.2 million users in a period of 12 months.

…While it may be difficult to find infringement in [Copilot] outputs, the question of inputs is really where things are starting to heat up. The most interesting legal debate is happening with the data used to train machine learning models. This has been a large part of the ongoing debate with art models (discussed here), but the first shot in the future of litigation may very well involve Copilot.

Programmer and lawyer Matthew Butterick has been getting a lot of attention when he announced that he was starting an investigation into Copilot with the intention of eventually starting a class-action lawsuit against Github and their parent company Microsoft.

…There is also a very strong ethical element to the complaint. Open source software communities are there to share code, but Copilot takes that code and closes it in a walled garden that contributes nothing to the community.

This is probably the biggest potential challenge to AI that we have witnessed yet, and its reach cannot be underestimated. I have been getting a few questions about this: is Butterick right?


Could be an important case, but with US copyright laws tending to be looser than British ones it feels like this is unlikely to go against Microsoft. The UK already allows data mining for training ML, and so does the EU.
unique link to this extract

Some people really are mosquito magnets, and they’re stuck that way • Scientific American

Daniel Leonard:


In a new paper published on October 18 in the journal Cell, researchers suggest that certain body odours are the deciding factor. Every person has a unique scent profile made up of different chemical compounds, and the researchers found that mosquitoes were most drawn to people whose skin produces high levels of carboxylic acids. Additionally, the researchers found that peoples’ attractiveness to mosquitoes remained steady over time, regardless of changes in diet or grooming habits.

“The question of why some people are more attractive to mosquitoes than others—that’s the question that everybody asks you,” says study co-author Leslie Vosshall, a neurobiologist and mosquito expert at the Howard Hughes Medical Institute and the Rockefeller University. “My mother, my sister, people in the street, my colleagues—everybody wants to know.” That public interest is what drove Vosshall and her colleagues to design this study, she says.

…Vosshall and her colleagues gathered 64 participants and had them wear nylon stockings on their arms. After six hours, the nylons were imbued with each person’s unique smell. “Those nylons would not have a smell to me or, I think, to anyone really,” says Maria Elena De Obaldia, a senior scientist at the biotech company Kingdom Supercultures and lead author of this new study, which she conducted while at Rockefeller. Still, the stockings were certainly odorous enough to entice mosquitoes.

The researchers cut the nylons into pieces and placed two (from different participants) into a closed container housing female Aedes aegypti mosquitoes. Did they migrate to subject number one’s sample en masse or prefer the scent of subject number two’s? Or were both equally appealing? The researchers continued these head-to-head battles over several months, Vosshall says, collecting new samples from the participants as needed. When the tournament was over, the team had clear proof that some people were more attractive than others.


I’ve lived in multiple tropical countries, and it was always our belief that mosquitoes preferred particular people’s scent. Anyhow, as I’m going to be holidaying somewhere with mosquitoes, this is only mildly encouraging news.
unique link to this extract

Adobe plays catch-up with Project Blink, an AI-powered video editor • Ars Technica

Benj Edwards:


On Wednesday, Adobe announced Project Blink, an AI-powered web app that can recognize objects, people, sounds, and words spoken in videos to help edit them faster. It uses text editing to make the process of editing a video similar to using a word processor, Adobe claims.

Adobe debuted Project Blink during a “sneaks” session at its Adobe MAX conference, according to CNET. The firm also released a demonstration video that depicts editing video by editing a text transcription of words spoken during the video. “Just upload your video, and our AI engine will figure out what happened,” it says.

Adobe claims that Project Blink can search a video for specific people, objects, or feelings—or locate sections where people are laughing or singing. Additionally, Adobe says you’ll be able to delete silent sections or remove filler words like “ums” in text, with changes automatically reflected in the video.

Some of Project Blink’s editing capabilities strongly resemble existing AI-powered video editors such as Runway, which we reported on last month, and Descript, which can edit videos based on written transcripts similar to a word processing document. It’s been a busy year for deep learning AI applications, including text-to-image and text-to-video products that have turned assumptions about creative content generation on their head. All that AI activity has prompted some to wonder how Adobe would respond, and now we’re seeing some reaction from the creative app giant.


Microsoft and Adobe. Everyone’s doing it.
unique link to this extract

These mini wind generators with no spinning blades can power homes and buildings • Singularity Hub

Vanessa Bates Ramirez:


In 2021, wind turbines generated more than 9% of US utility-scale electricity. The majority of the turbines making up that figure are the horizontal-axis variety (a rotor mounted at the top of a pole, blades twirling windmill-style). But wind energy isn’t limited to this classic design. There are also small vertical-axis turbines; large offshore vertical-axis turbines; and now, rooftop wind generators that aren’t really turbines and don’t have an axis at all, at least not in the traditional sense.

The generators are called Aeromines. Made by Aeromine Technologies, they harness wind and convert it to energy differently than conventional turbines. The latter use wind to turn blades attached to a rotor, and the spinning rotor powers a generator. Aeromines don’t have rotors or blades; instead, they have two airfoils or “wings” shaped like spoilers, angled towards each other on either side of a pole.

The generators sit on the edge of a building’s roof, taking advantage of the aerodynamic effect created by the wall below. As wind hits the airfoils, it creates a low pressure zone that sucks air through perforations and turns a propeller at the bottom of the unit. The spinning propeller is connected to a generator, which can be hooked up to a battery or connect directly to the building to provide electricity.

…The average US household uses about 9,000 kWh of electricity each year, which requires a 6.6 kW array of solar panels (about 21 standard panels). That means one Aeromine could meet the needs of an energy-conserving home, and two would be more than enough for an energy-gobbling home.


This link suggests that Aeromines generate 50% more energy per dollar spent than solar panels, which sounds promising. But a cursory glance suggests they need a flat roof, which not that many domestic houses in the UK (not many domestic houses generally?) have.
unique link to this extract

Three fundamental problems still plaguing Meta’s Enterprise XR ambitions • Creative Strategies

Oliver Blanchard:


The first point of friction in Meta’s enterprise metaverse play is the cost of entry of its hardware: Even under the best economic tailwinds, adding $1,500 per head to budget allocations that already include phones, tablets and PCs would be hard to justify, especially when the reason behind the extra spending is an XR headset whose purpose and value for the average employee remain difficult to quantify. Granted, $1,500 per unit isn’t a lot of cash for large orgs so long as the value is there, and we have seen the math work well in niche use cases. But if Meta and Microsoft are setting their sights on mass adoption and scale, either the ROI of that investment will have to be made clear and indisputable, or the pricing will have to come down. Way down.

The second point of friction is the form factor of the hardware itself. As impressive as the Quest 2 Pro’s specs and capabilities are, it still looks and feels more like a deconstructed fighter jet helmet than a pair of AR glasses: Bulbous, wide, on the wrong side of light and comfortable, and not particularly portable. The question almost asks itself: Does anyone really want to wear a cumbersome XR headset at work all day? Probably not, at least not unless you absolutely have to.

…The third point of friction is Meta’s unfortunate dehumanization of collaboration in the metaverse. As cute as it may initially seem to transform coworkers into Wii-themed digital avatars (with or without digital legs) I worry that replacing human faces with digital ones will have a negative emotional and psychological impact on remote workers over time, and we would do well not to overlook or underestimate this potential problem in the making.


Hard to disagree with, though as he says price isn’t much of a barrier. Companies were happy to lay out thousands of dollars for the first PCs because they saw they might be transformative. Though there’s a fair bit of persuasion to be done here on that.
unique link to this extract

MubertAI/Mubert-Text-to-Music: a simple notebook demonstrating prompt-based music generation via Mubert API • Github

Ilya Belikov:


We’re glad to present you our new Text-to-Music demo interface. Now as a Google Colab, and soon we’ll add this feature as a simple form on our website. This has already gone viral, so the community has questions about how everything works

People ask how generative this music really is. Each time you send a request, our API generates a unique combination of sounds. The probability of repetition is extremely small. Music is not taken from the database of finished tracks, it is created at the time of the request.

How are sounds selected for generation? The input prompt and Mubert API tags are both encoded to latent space vectors of a transformer neural network. Then the closest tags vector is selected for each prompt and corresponding tags are sent to our API for music generation.

All sounds (separate loops for bass, leads etc.) are created by musicians and sound designers, they are not synthesized by any neural network. Our paradigm is “from creators to creators”. We are musicians ourselves and it is important for us that musicians stay in the equation.


View it on YouTube. It’s quite weird (the prompt “Vladimir Lenin smoking weed with Bob Marley” produces a reggae tune with strange overtones).

Anyhow, another AI-generated content source.
unique link to this extract

Damaged cable leaves Shetland cut off from mainland • BBC News


Communications to Shetland have been severely disrupted after a subsea cable was damaged.

Police have declared a major incident after the south subsea cable between the islands and the mainland was cut. The force said some landlines and mobiles were not usable and that officers were patrolling to try to reassure residents.

Repairs to another cable connecting Shetland and Faroe are ongoing after it was damaged last week.
First Minister Nicola Sturgeon said it was an emergency situation for the island.

The Scottish government’s resilience committee had met and was working with partner agencies to ensure support was provided, she added. She said the assumption was the damage was accidental, adding: “There is nothing to suggest otherwise, but work is continuing to assess exactly what the cause of the problem has been.”

MP for Orkney and Shetlands Alastair Carmichael told the BBC he had raised the issue with the UK government, but understood it could be days before communications were restored.

…The cable that was damaged between Faroe and Shetland last week will be repaired on Saturday, according to Faroese Telecom’s head of infrastructure Páll Vesturbú. He said: “The damage is affecting most of telecom services to Shetland. There are some services still working but we will try to establish more services during the day if that’s possible.

“We expect it will be fishing vessels that damaged the cable but it is very rare that we have two problems at the same time.” MP Alistair Carmichael added that the damage had caused “catastrophic impact”.


Fishing vessels, one hopes. Or possibly.. Russian submarines?
unique link to this extract

Warner Bros.’ Lord of the Rings NFT ‘experience’ sounds like a nightmare • The Verge

Charles Pulliam-Moore:


On Thursday, Warner Bros. announced the impending arrival of “The Lord of the Rings: The Fellowship of the Ring (Extended Version) Web3 Movie Experience,” a cumbersomely named rerelease of Peter Jackson’s 2001 film that will live on Eluvio’s branded Content Blockchain. In a press release about their partnership, Eluvio CEO Michelle Munson sang Warner Bros.’ praises for its commitment to NFT projects and said that their partnership is poised to help bring films-as-NFTs to an even bigger audience of consumers.

“Fans of The Lord of the Rings can now acquire, participate, and trade in an epic living media experience that will undoubtedly surprise and delight them,” said Munson. “It’s truly designed for a mass consumer audience, not just Web3 enthusiasts, which is why it should, and does, feel so remarkable and engaging.”

You can already buy and permanently own physical copies of The Fellowship of the Ring and all of Warner Bros.’ Lord of the Rings films in 4K. But the studio’s banking on a handful of NFT-related features being enough to convince people to buy either the “Mystery” or “Epic” editions of the movie as their first step toward becoming embedded in WB’s Movieverse.

…it’s hard to imagine Warner Bros. will really be able to sell people on what sounds very much like a gussied-up, browser-based DVD selection menu masquerading as a collectible item.


Reasonable points, though I feel that it shouldn’t be impossible to find a person – analyst, cynic, whatever – who would make the point about the uselessness of the “Web3 Movie Experience”.
unique link to this extract

Year of the Four Emperors • Wikipedia


The Year of the Four Emperors, AD 69, was the first civil war of the Roman Empire, during which four emperors ruled in succession: Galba, Otho, Vitellius, and Vespasian.[1] It is considered an important interval, marking the transition from the Julio-Claudians, the first imperial dynasty, to the Flavian dynasty. The period witnessed several rebellions and claimants, with shifting allegiances and widespread turmoil in Rome and the provinces.


Hmm, can’t think what it was that prompted Overspill reader G to point us to this one. The UK is so far only on its third prime minister (and, OK, second monarch, but cool your jets), though of course the 12 months began in the summer, so plenty of time yet for prime minister No.4 to make an entrance some time in 2023.
unique link to this extract

How a secret rent algorithm pushes rents higher • ProPublica

Heather Vogel, Haru Coryne and Ryan Little:


On a summer day last year, a group of real estate tech executives gathered at a conference hall in Nashville to boast about one of their company’s signature products: software that uses a mysterious algorithm to help landlords push the highest possible rents on tenants.

“Never before have we seen these numbers,” said Jay Parsons, a vice president of RealPage, as conventiongoers wandered by. Apartment rents had recently shot up by as much as 14.5%, he said in a video touting the company’s services. Turning to his colleague, Parsons asked: What role had the software played?

“I think it’s driving it, quite honestly,” answered Andrew Bowen, another RealPage executive. “As a property manager, very few of us would be willing to actually raise rents double digits within a single month by doing it manually.”

The celebratory remarks were more than swagger. For years, RealPage has sold software that uses data analytics to suggest daily prices for open units. Property managers across the United States have gushed about how the company’s algorithm boosts profits.

“The beauty of YieldStar is that it pushes you to go places that you wouldn’t have gone if you weren’t using it,” said Kortney Balas, director of revenue management at JVM Realty, referring to RealPage’s software in a testimonial video on the company’s website.

The nation’s largest property management firm, Greystar, found that even in one downturn, its buildings using YieldStar “outperformed their markets by 4.8%,” a significant premium above competitors, RealPage said in materials on its website. Greystar uses RealPage’s software to price tens of thousands of apartments.


Depressing, honestly. Rent income is effectively unearned, yet this is about squeezing money from people doing actual work.
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

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your 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.