People are getting injured or even killed taking selfies, a new study shows – and it’s time to warn them properly. CC-licensed photo by Mike Goad 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. Hello, duckface. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.
Google’s 36% search revenue share with Apple is 3x what Android OEMs get • Ars Technica
Ron Amadeo:
»
How much more does Google pay for an Apple user than an Android one? A lot. It was recently revealed in the Epic v. Google trial (Google has a few monopoly lawsuits going on) that the highest tier of search revenue share for cooperative Android OEMs is only 12%, a third of what Google pays Apple. In terms of total cash amount, it’s reasonable to assume Apple gets more total money than many smaller companies but to see the direct breakdown that each Apple user is worth three times more than an Android user is a new insight.
A big part of the differing payment rates probably has to do with how threatened Google feels by each company. Apple has already proven that it has the power to dump an established Google service and go off on its own. A prime example is Apple Maps, which replaced Google Maps as a default iOS app and, according to testimony from Google VP of Finance, Michael Roszak, tanked Google Maps mobile traffic by 60% when it launched. Roszak said that Google uses the Apple Maps launch as “a datapoint” when estimating how an Apple search switch would go. No one on the Android side has this kind of power. There’s also the consideration that Apple users are generally more affluent than Android users, making them more desirable ad clickers.
On Android, Google has differing tiers of payments depending on how Google-y your phone is. As revealed in documents from Epic v. Google, Android’s “Premier Device Program” offers 12% search revenue to devices with “Google exclusivity and defaults for all key functions” and no rival app stores. The big participants in this program are/were Motorola, LG, and HMD, which had at least 98% of their devices qualify. Other brands like Xiaomi, Sony, Sharp, and BBK (that’s OnePlus, Oppo, and Vivo) were at 70%.
Android partners don’t just get search revenue; they also get a cut of Google Play app sales and ads run on their devices. In the case of Motorola and LG, they were getting another 3–6% of Play Store spending.
«
It is fascinating seeing Google’s business model, particularly around TAC (traffic acquisition costs), being picked apart in public like this.
unique link to this extract
Meta bars political advertisers from using generative AI ads tools • Reuters
Katie Paul:
»
Facebook owner Meta is barring political campaigns and advertisers in other regulated industries from using its new generative AI advertising products, a company spokesperson said on Monday, denying access to tools that lawmakers have warned could turbo-charge the spread of election misinformation.
Meta publicly disclosed the decision in updates posted to its help center on Monday night, following publication of this story. Its advertising standards prohibit ads with content that have been debunked by the company’s fact-checking partners but do not have any rules specifically on AI.
“As we continue to test new Generative AI ads creation tools in Ads Manager, advertisers running campaigns that qualify as ads for Housing, Employment or Credit or Social Issues, Elections, or Politics, or related to Health, Pharmaceuticals or Financial Services aren’t currently permitted to use these Generative AI features,” the company said in a note appended to several pages explaining how the tools work.
«
So political campaigns won’t be able to use Meta’s generative AI tools to make their ads, but they could certainly do so outside Meta and then just upload them. Hard to see what difference this makes to anything, apart from giving Meta something to say it “doesn’t allow” in US congressional hearings.
unique link to this extract
Why I just resigned from my job in generative AI • Music Business Worldwide
Ed Newton-Rex worked at Stability AI on its Stable Audio generative AI music-making platform; he’s a published classical composer in his own right:
»
I’ve resigned from my role leading the Audio team at Stability AI, because I don’t agree with the company’s opinion that training generative AI models on copyrighted works is ‘fair use’.
…I wasn’t able to change the prevailing opinion on fair use at the company.
This was made clear when the US Copyright Office recently invited public comments on generative AI and copyright, and Stability was one of many AI companies to respond. Stability’s 23-page submission included this on its opening page: “We believe that Al development is an acceptable, transformative, and socially-beneficial use of existing content that is protected by fair use”.
For those unfamiliar with ‘fair use’, this claims that training an AI model on copyrighted works doesn’t infringe the copyright in those works, so it can be done without permission, and without payment. This is a position that is fairly standard across many of the large generative AI companies, and other big tech companies building these models — it’s far from a view that is unique to Stability. But it’s a position I disagree with.
I disagree because one of the factors affecting whether the act of copying is fair use, according to Congress, is “the effect of the use upon the potential market for or value of the copyrighted work”. Today’s generative AI models can clearly be used to create works that compete with the copyrighted works they are trained on. So I don’t see how using copyrighted works to train generative AI models of this nature can be considered fair use.
«
Well, I guess we’ll see how the courts view this. But it seems to me that this is a “transformational” use and will be allowed. Will Newton-Rex ask for his job back if the courts rule that way?
unique link to this extract
The incredible shrinking heat pump • The Verge
Justine Calma:
»
Just 10% of households worldwide have heat pumps today. Those are typically bigger, more complex, and expensive systems that need to be professionally installed. For those reasons, they’re usually out of reach for renters. New York City Housing Authority (NYCHA) actually did a test run with one of those existing options, called a split system unit, which involved mounting equipment on the roof and on the wall in a tenant’s home. It ended up being too unwieldy, and the project stopped there.
Unfortunately, when it comes to new, more efficient appliances and clean energy technologies, it’s typically more affluent households that can afford to bring these new things into their homes first. The benefits don’t usually trickle down to lower-income households until later, if at all.
New York is attempting to flip that scenario now by purchasing new window heat pumps for public housing residents. “The beauty of this project is that some of the lowest-income residents in the city are experiencing the newest technology for the first time so they’re leading in this area, which is really nice and something that we’re very proud of,” says Justin Driscoll, president and CEO of New York Power Authority, the public power organization that procures electricity for NYCHA.
The big motivation to switch to heat pumps now, though, is a deadline. Back in 2019, New York state passed a law to reduce the greenhouse gas emissions causing climate change by 85% by 2050.
…Heat and hot water in buildings create about 40% of New York City’s planet-heating greenhouse gas emissions. So, in 2021, NYCHA and its partners announced a $263m investment in electric heat pumps.
«
They have huge air conditioning units festooned over houses, so why not heat pumps?
unique link to this extract
Here’s how bad climate change will get in the US—and why there’s still hope • WIRED
Matt Simon:
»
The [US Fifth National Climate] assessment notes the already staggering cost of climate change in the US, beyond wildfires. In the 1980s, on average, the US experienced one billion-dollar disaster every four months. That’s now one every three weeks. Between 2018 and 2022, the country suffered 89 billion-dollar events. Extreme weather now costs the country nearly $150bn annually. But, the report emphasizes, that’s a conservative estimate, because it doesn’t consider the costs of the aftermath, like loss of life, health care for survivors, or the damage done to ecosystems.
“I think this report really highlights how the changes we’re experiencing now are unprecedented in our nation’s history,” says Kristina Dahl, a technical contributor to the assessment and principal climate scientist for the Climate and Energy program at the Union of Concerned Scientists. “The US has warmed more quickly than the planet as a whole. So the US is really feeling this.”
The assessment also points out that in the next three decades, scientists expect sea levels along the contiguous US to rise nearly a foot. By 2050, coastal flooding will happen five to 10 times more often than today, and by the end of the century, millions of seaside residents could be displaced. But we’re dealing with a lot of uncertainty. Sea level rise could accelerate if the ice sheets atop Greenland and Antarctica start declining faster. Just last week, a study found that northern Greenland’s ice is in much worse shape than previously understood. “Uncertainty in the stability of ice sheets at high warming levels means that increases in sea level along the continental US of 3-7 feet by 2100 and 5-12 feet by 2150 are distinct possibilities that cannot be ruled out,” the assessment warns.
And keep in mind that sea level rise will not unfold uniformly across US coastlines, due to quirks in the physics involved. Some places, like the Gulf Coast, are also rapidly sinking, a phenomenon known as subsidence, which exacerbates the problem.
«
2050 actually doesn’t feel that far away.
unique link to this extract
YouTubers asked to disclose AI-generated content – or else • The Register
Katyanna Quach:
»
YouTube is slapping a bunch of rules on AI-generated videos in the hope of curbing the spread of faked footage masqueraded as legit; deepfakes that make people appear to say or do things they never did; and tracks that rip off artists’ copyrighted work.
This red tape will be rolled out over the coming months and apply to material uploaded by users, we’re told
Specifically, the Google-owned vid-sharing giant will require content creators to disclose if their videos contain believable synthetic footage of made-up events, including AI-made depictions, or deepfakes that put words in people’s mouths. In those cases, a label will be added to a video’s description declaring the content was altered or digitally generated, and a more prominent note will be added to the video player itself if the content is particularly sensitive. Breaking the rules will lead to content being torn down and accounts punished.
…Faked footage that could mislead viewers about important topics such as elections, conflicts and violence, public health issues, or popular figures must also be flagged in particular. “Creators who consistently choose not to disclose this information may be subject to content removal, suspension from the YouTube Partner Program, or other penalties,” YouTube product veeps Jennifer Flannery O’Connor and Emily Moxley warned today.
«
Well, let’s see how that works out for them. How will they know, apart from anything?
unique link to this extract
Everyone stop being ridiculous for like five minutes • Garbage Day
Ryan Broderick:
»
There’s an old journalism joke that reporters cover every new election according to the rules of the previous one. But I think the tech press does the same thing. Which explains why most of the stories you read about AI right now use the same whack-a-mole content cop strategy most news outlets and research groups spent the 2010s using to cover platforms like Facebook or Twitter. Now they’re breathlessly writing up every instance of an AI producing A Forbidden Image. And what’s worse is this attitude helps tech companies continue to undermine labor and consolidate lobbying power, allows politicians to keep dragging their feet on writing real legislation for the internet, and provides fantastic cover for online platforms that still don’t know how to moderate themselves. I have yet to see anything produced by generative AI you couldn’t do with Photoshop or After Effects or, like, Wikipedia. And if everyone stopped being ridiculous for five minutes, we’d all realize that this tech hasn’t introduced a single new problem. We still just have same old ones we refuse to deal with!
And so, my big hot AI take here is that there’s actually nothing new to moderate. I mean, my god, OpenAI is literally using the same Africa-based third-party moderation contractors that Meta and Google use. It’s all just the same stuff with a new Sci-Fi coat of paint.
«
Broderick essentially arguing the opposite position from the Tech Against Terrorism research linked here yesterday from Wired.
unique link to this extract
Sam Bankman-Fried exposed the fragility of crypto • The New York Times
Molly White:
»
Although some might try to dismiss the FTX collapse as a unique case, it is far from it: the funding model [of issuing a “coin” in an “initial coin offering”, or ICO, and putting a ludicrous value on it, and then using that as collateral for loans of real money] has become all too normal in the cryptocurrency world. Entrepreneurs thought they had found a free money machine in 2017 as initial coin offerings became popular, enabling cryptocurrency companies to bootstrap without needing to find venture investors — investors who might insist on a seat on the board or a view into the company’s operations. A crackdown on I.C.O.s in the United States shortly after failed to stop the practice, with companies either presenting the offerings in disguises designed to stymie if not entirely evade the Securities and Exchange Commission, or moving offshore in hopes of being beyond the reach of the long arms of the Securities and Exchange Commission (S.E.C.)’s enforcers.
FTX is perhaps the best known catastrophe, but the same pattern has played out for customers of the Celsius cryptocurrency lender’s CEL token, the Voyager Digital broker’s VGX and the Terra/Luna ecosystem’s LUNA. Civil and criminal cases have revealed internal conversations among Celsius executives desperately trying to support the CEL token price to keep the floundering company afloat, to no avail, knowing what the token’s collapse would mean for the company. Binance, a still-operational exchange whose balance sheets are as opaque as those of the Bankman-Fried companies before their collapse, heavily promotes its BNB token. The extent to which the company relies on BNB to finance its operations is unclear, but history provides ominous warnings.
The collapse of the FTX exchange revealed the massive duplicity underlying many crypto exchanges, but its implosion should not be attributed to that alone. It, like so many companies in the cryptocurrency industry, had propped itself up on an imaginary foundation of tokens it had invented, and that foundation was bound to fail eventually. When the next company in its position falls, the only surprise should be that people expected any other outcome.
«
I’ve long since ceased to be surprised by the things that people will believe will happen differently this time.
unique link to this extract
Selfie-related deaths at tourist sites are ‘public health problem’: researchers • NY Post
Angie Raphael:
»
Selfie-related injuries and deaths at tourist hotspots have become such a massive risk that they should be viewed as a “public health problem”, researchers suggest.
Of particular concern are selfie-related deaths at picturesque aquatic locations, such as waterfalls, according to the University of New South Wales, Australia study.
Part of the study examined how selfie-related injuries and deaths were reported in the media.
Four peer-reviewed studies identified falls from a height, such as a cliff or waterfall, as the most common incident. Drowning was the second most common cause of death. People often climbed over barriers and fenced-off areas to get to the perfect selfie spot, the report noted.
The mean age of victims was about 22, most of whom were female tourists. “The selfie-related incident phenomenon should be viewed as a public health problem that requires a public health risk communication response,” the report concluded. “To date, little attention has been paid to averting selfie-related incidents through behaviour change methodologies or direct messaging to users, including through social media apps.”
Previous research recommended “no selfie zones”, barriers and signage as ways to prevent selfie-related injuries and deaths.
«
Perhaps something like “5 people have died here so far taking selfies”, but write the number with 1, 2, 3, and 4 crossed out in front of it. The paper is in the “Journal of Medical Internet Research”, which is an intriguing title in its own right.
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
The Overspill: when there's more that I want to say 