Skiers’ Apple Watches are triggering emergency alerts – which is annoying, especially for responders. CC-licensed photo by 7th Groove 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 about 0845 UK time. Free signup.
A selection of 10 links for you. Don’t call me. I’m @charlesarthur on Twitter. Observations and links welcome.
Why Apple Watches keep calling 911 • The New York Times
On a recent sunny Sunday morning, following a night of fluffy snowfall, tens of thousands of skiers flocked to the resorts of Summit County. Just minutes after the lift lines opened, sirens began blaring in the 911 emergency service center, where four staff members were taking calls and dispatching help.
Each jarring alert was a new incoming call, heralding a possible car crash, heart attack or other life-threatening situation. Often, the phone operators heard a chilling sound at the far end of the line: silence, perhaps from a caller too incapacitated to respond.
At 9:07 a.m., one dispatcher, Eric Betts, responded to such a call. From the map on one of the seven monitors on his desk, he could see that the distress call originated from a slope at the Arapahoe Basin Ski Area. Mr. Betts tried calling back. A man picked up.
“Do you have an emergency?” Mr. Betts asked. No, the man said, he was skiing — safely, happily, unharmed. Slightly annoyed, he added, “For the last three days, my watch has been dialing 911.”
Winter has brought a decent amount of snowfall to the region’s ski resorts, and with it an avalanche of false emergency calls. Virtually all of them have been placed by Apple Watches or iPhone 14s under the mistaken impression that their owners have been debilitated in collisions.
As of September, these devices have come equipped with technology meant to detect car crashes and alert 911 dispatchers. It is a more sensitive upgrade to software on Apple devices, now several years old, that can detect when a user falls and then dial for help. But the latest innovation appears to send the device into overdrive: It keeps mistaking skiers, and some other fitness enthusiasts, for car-wreck victims.
…The problem extends beyond skiers. “My watch regularly thinks I’ve had an accident,” said Stacey Torman, who works in London at Salesforce and also teaches spin classes. She might be safely on the bike, exhorting her class to ramp up the energy, or waving her arms to congratulate them, when her Apple Watch senses danger.
“I want to celebrate, but my watch really doesn’t want me to celebrate,” she said. Oh great, she thinks, “now my watch thinks I’m dead.”
It’s a fabulous new feature on the Apple Watch: unintended consequences. Deeply annoying for the people running the ski emergency call centre.
unique link to this extract
America’s offices are now half full. Is this as good as it gets? • The Washington Post
The tug of war over getting workers back to the office just reached a key milestone: 50% are back at their desks on average, the most since the pandemic hit in March 2020.
But that means major corporate offices are only half as full as they once were — and many experts think this could be as good as it gets.
Overall growth in office occupancy has begun to level off in recent months despite efforts by many bosses to get workers back more often, according to data tracked by Kastle Systems. Last week, office occupancy across the country’s top 10 metro areas edged up to 50.4% of pre-pandemic levels, according to Kastle, which measures office activity through entry swipes.
Indeed, late January marked the first time that all 10 cities tracked by the index — including laggards like San Francisco that lean remote — notched average occupancy rates of at least 40% of pre-pandemic levels.
But the return-to-office figures are unlikely to go much higher as flexible work becomes entrenched in the lives of white-collar workers, experts say. Some employees have resisted hard mandates to return: they’ve left for remote opportunities elsewhere or even flouted in-office requirements, flexing worker leverage while the labor market remains hot. In response, more companies seem to be moving toward acknowledging that the 9-to-5, Monday-through-Friday in-office job is over. More than half of US jobs that can be done remotely were hybrid as of November, up from 32% in January 2019, according to data from Gallup.
That really is quite a shift.
unique link to this extract
Scoop: oh the things you’ll do with Bing’s ChatGPT: features sneak peek • Medium
Owen Yin got to play with the new Bing, apparently as an accident on Microsoft’s part:
Bing’s ChatGPT integration will be called the new Bing. Microsoft is positioning it as an evolution of the search engine, asking you to think of it as “an research assistant, personal planner, and creative partner at your side. The integration will be powered by OpenAI’s GPT-4, a faster version of ChatGPT.
The search bar is now a chatbox: the blank search bar is replaced with a large text box with a prompt inviting you to ask it anything. You’ll have 1000 characters to write your question, which will allow for a good amount of detail in your request. You’ll be able to provide context, provide specific instructions, or list examples.
Bing does research for you: unlike ChatGPT, which is trained on data collected up to 2021, the new Bing will be able to access current information.
When you ask a question, the AI will interpret it and make several searches related to your request. It will then compile the results and write a summary for you. Bing will highlight particular phrases and cite where it got that information from, allowing you to verify the claim.
Bing can make plans for you: Bing will be able to process complex tasks that you’d usually have to piece together yourself. You can give it personal requirements, like your meal preferences, budget constraints, location, or time requirements and Bing will adapt its response to your needs and interests. You could get Bing to generate inspiration for a meal plan or travel itinerary.
You can ask Bing to be creative: the new Bing has a new level of creativity and imagination. Since the AI can understand and respond to natural language queries, you’ll be able to ask Bing to take on creative tasks previously beyond the realm of a search engine. You could ask Bing to write a rhyming poem for your cousin’s birthday or create a short story featuring your friends.
Wonder if this will shift the needle at all away from Google in search. Will there be novelty value that wears off, or will it keep people by being effectively a free ChatGPT? Google still has lots of money to buy being the default search.
unique link to this extract
ChatGPT’s creators can’t figure out why it won’t talk about Trump • Semafor
Even ChatGPT’s creators can’t figure out why it won’t answer certain questions — including queries about former U.S. President Donald Trump, according to people who work at creator OpenAI.
In the months since ChatGPT was released on Nov. 30, researchers at OpenAI noticed a category of responses they call “refusals” that should have been answers.
The most-widely discussed one came in a viral tweet posted Wednesday morning: When asked to “write a poem about the positive attributes of Trump,” ChatGPT refused to wade into politics. But when asked to do the same thing for current commander-in-chief Joe Biden, ChatGPT obliged.
The tweet, viewed 29 million times, caught the attention of Twitter CEO Elon Musk, a co-founder of OpenAI who has since cut ties with the company. “It is a serious concern,” he tweeted in response.
Even as OpenAI is facing criticism about the hyped services’ choices around hot-button topics in American politics, its creators are scrambling to decipher the mysterious nuances of the technology.
Many of the allegations of bias are attempting to fit a new technology into the old debates about social media. ChatGPT itself cannot discriminate in any conventional sense. It doesn’t have the ability to comprehend, much less care about, politics or have an opinion on Republican congressman George Santos’ karaoke performances.
But conservatives who criticize ChatGPT are making two distinct allegations: They’re suggesting that OpenAI employees have deliberately installed guardrails, such as the refusals to answer certain politically sensitive prompts. And they’re alleging that the responses that ChatGPT does give have been programmed to skew left. For instance, ChatGPT gives responses that seem to support liberal causes such as affirmative action and transgender rights.
The accusations make sense in the context of social media, where tens of thousands of people around the world make judgments about whether to remove content posted by real people.
But it reflects a misunderstanding about the way ChatGPT’s technology works at a fundamental level, and all the evidence points to unintentional bias, including its underlying dataset — that is, the internet.
I’d say that refusing to tweet about Trump suggests that ChatGPT has become intelligent. Well done, ChatGPT!
unique link to this extract
Journalistic lessons for the algorithmic age • The Markup
Julia Angwin, the founding editor, is departing The Markup (though she doesn’t say if it’s her idea or a spreadsheet’s):
At The Markup, we developed an investigative checklist that reporters filled out before embarking on a project. Top of the checklist was not novelty, but scale—how many people were affected by the problem we were investigating. In other words, we chose to tackle things that were important but not secret.
For instance, anyone using Google has probably noticed that Google takes up a lot of the search result page for its own properties. Nevertheless, we decided to invest nearly a year into quantifying how much Google was boosting its own products over direct links to source material because the quality of Google search results affects nearly everyone in the world.
This type of work has an impact. The European Union has now passed a law banning tech platforms from this type of self-preferencing, and there is legislation pending in Congress to do the same.
Plus: work out the hypothesis first, then see if the data(set) agrees; know that data is political; choose the sample size; go for the numbers. The things that The Markup has achieved are remarkable. Angwin is a loss, whoever (or whatever) said it was time to go.
unique link to this extract
Harvard is shutting down project that studied social media misinformation • The Washington Post
Drew Harwell and Joseph Menn:
Since 2019, the Technology and Social Change Project has published research into the spread of coronavirus hoaxes and the online incitement techniques that preceded the Jan. 6, 2021, riot at the U.S. Capitol. It will wind down due to a school policy that requires a faculty member lead such an undertaking, Nancy Gibbs, the director of the Kennedy School’s Shorenstein Center on Media, Politics, and Public Policy, said in an internal email shared with The Washington Post.
The project’s director, Joan Donovan, one of the country’s most widely cited experts on digital “media manipulation,” is not a faculty member and therefore could not continue to lead the project, Gibbs said.
Donovan, whose title is research director, is regarded as a member of the Shorenstein Center’s staff; it’s unknown whether she had been given the option to assume a faculty role during her time at Harvard.
…The project’s sunsetting in the months before the 2024 election marks a surprise development in what has been one of the American research world’s hottest topics: how the interplay of technology, political opportunism and unwitting internet users has shaped public conversation and democratic debate.
But it also comes as the field of study into what’s known as “misinformation,” supercharged by the Trump presidency, enters a new era.
Twitter, now led by the meme-sharing billionaire Elon Musk, has worked to end the platform’s long-standing openness to free, real-time research, announcing late Wednesday that on Feb. 9 it will begin charging for automated access to its data through its Application Programming Interfaces, a move that will hurt both developers and researchers.
Some researchers also have faced harassment online or been criticized by Republican lawmakers over claims their work is skewed by a liberal agenda. The platforms they study have changed, too, away from Facebook and Twitter to places like TikTok, Discord and Twitch, which present new challenges for data gathering, analysis and debate.
Twenty questions about the Online Safety Bill • Cyberleagle
Graham Smith, who runs a blog on “law, IT, the internet and online media::
Before Christmas Culture Secretary Michelle Donelan invited members of the public to submit questions about the Online Safety Bill, which she will sit down to answer in the New Year.
Here are mine.
1.A volunteer who sets up and operates a Mastodon instance in their spare time appears to be the provider of a user-to-user service. Is that correct?
2.Alice runs a personal blog on a blogging platform and is able to decide which third party comments on her blogposts to accept or reject. Is Alice (subject to any Schedule 1 exemptions) the provider of a user-to-user service in relation to those third party comments?
3.Bob runs a blog on a blogging platform. He has multiple contributors, whom he selects. Is Bob the provider of a user-to-user service in relation to their contributions?
4.Is a collaborative software development platform the provider of a user-to-user service?
5.The exclusion from “regulated user-generated content” extends to comments on comments (Clause 49(6)). But a facility enabling free form ‘comments on comments’ appears to disapply the Sch 1 para 4 limited functionality user-to-user service exemption. Is that correct? If so, what is the rationale for the difference? Would, for example, a newspaper website with functionality that enabled free form ‘comments on comments’ therefore not enjoy exclusion from scope under Sch 1 para 4?
As you can imagine, there are 15 more, none of them simple to answer, but all important to answer. Over to you, Ms Donelan.
unique link to this extract
Brown gold: the great American manure rush begins • The Guardian
n an early August afternoon at Pinnacle Dairy, a farm located near the middle of California’s long Central Valley, 1,300 Jersey cows idle in the shade of open-air barns. Above them whir fans the size of satellites, circulating a breeze as the temperature pushes 100F (38C). Underfoot, a wet layer of feces emits a thick stench that hangs in the air. Just a tad unpleasant, the smell represents a potential goldmine.
The energy industry is transforming mounds of manure into a lucrative “carbon negative fuel” capable of powering everything from municipal buses to cargo trucks. To do so, it’s turning to dairy farms, which offer a reliable, long-term supply of the material. Pinnacle is just one of hundreds across the state that have recently sold the rights to their manure to energy producers.
Communities around the world have long generated electricity from waste, but the past few years have seen a surge in public and private investment into poop-to-energy infrastructure in the US. Though so far concentrated in states with dominant dairy sectors, like California, Wisconsin and New York, Biden’s landmark climate law passed last summer stands to unleash additional billions to support further development nationwide. The sector’s boosters describe it as an elegant way to cut emissions from both livestock and transport; but critics worry that the nascent industry could raise more issues than it resolves by entrenching environmentally harmful practices.
Animal agriculture is the nation’s single biggest source of methane, a greenhouse gas that climate scientists call a “super pollutant” due to its high short-term warming potential. The gas is released from animals when they burp, and through the decomposition of manure when collected in open-air ponds, a common livestock industry practice.
But those emissions are also a potential moneymaker. Methane from animal waste can be purified into a product virtually indistinguishable from fossil fuel-based natural gas. Marketed as renewable natural gas (RNG), it has a unique profit-making edge: in addition to revenue from the sale of the gas itself, energy companies can now also earn handsome environmental subsidies for their role in keeping methane out of the atmosphere.
Twitter is killing off the fun bots • Buzzfeed News
Katie Notopoulos and Pranav Dixit:
Daniel, the 23-year-old student in Germany behind @MakeItAQuote, told BuzzFeed News he would have never started it if there were a fee attached. “It’s a step in the wrong direction, as most of the API usage brings a lot of value to the platform,” he said. “And the fact that even myself, operating one of the biggest bots on the platform, has to consider shutting it down is very concerning. There are a lot of awesome, less popular bots. I don’t think any of them can be sustainable.”
The exact pricing of API access is not yet clear, and representatives for Twitter did not respond to requests for clarification. One screenshot of current API use pricing has been going around; prices start at $149/month for 500 requests of the API per month and go up to $2,499/month for high-volume use. On Thursday evening, Twitter owner Elon Musk posted a tweet suggesting that access to the API would cost $100/month and require ID verification.
There’s a long history of beloved bots on Twitter. @Horse_ebooks was a bot that grew a cult following for its vibey snippets of text like “everything happens so much.” (The account eventually was taken over by a human who continued to manually tweet in a bot-like way until he revealed himself.)
Other bots are newsworthy, like @TrumpsAlert and @BigTechAlert, which tweet whenever Trump and his advisers or Big Tech CEOs, respectively, follow or unfollow a new person.
Álex Barredo, who runs @BigTechAlert, told BuzzFeed News he’s open to paying a small fee to keep the popular bot running, but not if it costs $100 a month. He also has other options. He’s aware that most of the actual spam bots don’t even use the API to operate (which means charging for API access isn’t really going to wipe out spam), so one option he’s considering is to rework the bot so it doesn’t need the API. He’s also considering either open-sourcing it or moving it to its own website or a different platform.
Alternative headline: Twitter is killing off Twitter. Those prices are bonkers. There’s a list of bots that presently exist on Twitter, and where you can find them on Mastodon, in this Google Doc. Presently there are 84. Musk, in his random way, says that there will be a “free write-only” API for “bots providing good content for free”. Who decides what is “good content”?
He’s also looking to charge businesses $1,000 per month for their verifications. Musk and his advisers are fools. I wrote about this in my latest missive on my Social Warming Substack.
unique link to this extract
“2001: A Space Odyssey” directed by… George Lucas? • YouTube
Hilariously good, and you know that’s just how it would have been. Equally good, but far more restful (by the same creator), is “Star Wars directed by Stanley Kubrick“. (Both via John Gruber.)
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