The smog in Los Angeles has been a persistent problem for 70 years. Has it improved, or worsened? CC-licensed photo by radcliffe dacanay 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.
It’s Friday, so there’s another post due at the Social Warming Substack at about 0845 UK time.
A selection of 9 links for you. How many fingers? I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.
Elon Musk’s X fact-checking system delayed Israel corrections for days • NBC News
Ben Goggin:
»
An approved Community Notes member gave NBC News access to the feature’s volunteer interface, which showed that many false posts with hundreds of thousands of views had no notes, while other notes sat unapproved for hours and sometimes days on posts that accrued tens of thousands of views.
NBC News focused on two prominent pieces of Israel-Hamas misinformation that have already been debunked: a fake White House news release that was posted to X claiming that the Biden administration had granted Israel $8bn in emergency aid, and false reports that St. Porphyrius Orthodox Church in Gaza was destroyed.
Of 120 posts related to those stories, only 8% had a published community note, while 26% of those posts had unpublished notes from volunteers that had yet to be approved. About two-thirds of the top posts reviewed by NBC News had no proposed or published Community Notes on them.
The findings echo what one Community Notes volunteer said was a lack of response by X to efforts to debunk misleading posts.
“All weekend we were furiously vetting, writing, and approving Community Notes on hundreds of posts which were demonstrably fake news,” Kim Picazio, a Community Notes volunteer, wrote on Instagram’s Threads. “It took 2+ days for the backroom to press whatever button to finally make all our warnings publicly viewable. By that time… You know the rest of that sentence.”
Picazio told NBC News that she ended up just tweeting out proposed Community Notes herself in response to misinformation “out of total frustration.”
…According to a post from @CommunityNotes on Oct. 3, “Starting today, notes will appear an average of 1.5 hours faster, and as much as 3.5 hours faster in some cases.”
But even after these changes, users with access to the Community Notes program, and a review of content within the system by NBC News, found that the system was failing to catch misinformation posted more than 24 hours ago.
«
Saying you’ll post the 90 or 210 minutes Notes faster implies it takes you at least that long to get them posted. During which time the original can go viral. Always a weakness in these systems: virality is highest in the first 24 hours. Notes or other moves after that are effectively wasted.
unique link to this extract
Europe gives Musk 24 hours to respond about Israel-Hamas war misinformation on X • CNBC
Jonathan Vanian:
»
A European regulator has issued Elon Musk a stern warning about the spread of illegal content and disinformation on X, formerly known as Twitter, amid the Israel-Hamas conflict. Failure to comply with the European regulations around illegal content could result in fines worth 6% of a company’s annual revenue.
Thierry Breton, the European commissioner for the internal market, said in a letter addressed to Musk on Tuesday that his office has “indications” that groups are spreading misinformation and “violent and terrorist” content on X, and urged the billionaire to respond within a 24-hour period.
The letter comes after numerous researchers, news organizations and other groups have documented a rise of misleading, false and questionable content on X, creating confusion about the current conflict.
Breton shared his letter via an X post, tagging Musk’s handle and including a hashtag that refers to the Digital Services Act, the newly enacted legislation by the European Commission — the executive arm of the European Union — that requires platforms with more than 45 million monthly active users in the EU to monitor for and take down illegal content as well as detail their protocols for doing so.
«
I guess you’d have to wonder about motive, wouldn’t you, given the previous link.
unique link to this extract
L.A.’s forever war on smog • Los Angeles Times
Patt Morrison:
»
A handsome man with the physique of the college football player he had once been walks into an airtight, 8-foot-square Plexiglas booth heated to 90 degrees.
The door closes behind him.
When he staggers out, two hours later, he has won no fabulous prizes. Instead, he is a little headachey. He has trouble concentrating, his eyes are weepy, and, as a doctor will soon tell him, he has lost — temporarily at least — 22% of his lung capacity.
By now, you know this man is not a game-show contestant. And this is most definitely not a game.
But it is a stunt.
The newspaper reporters and photographers were there in 1956 to watch the man in the box, S. Smith Griswold, head of Los Angeles County’s nine-year-old Air Pollution Control District. And he had a point to prove.
He spent those two hours breathing in a potent summertime version of what millions of his fellow Angelenos were breathing every day — smog — to show them what it did to the human body, and to Los Angeles.
L.A. was like the royal baby in a fairy tale, endowed with abundant gifts — glittering sunlight, dramatic landscapes and lovely beaches that remain untarnished — until the bad wizard shows up with the asterisk:
“Yes, your landscape is gorgeous, but that magnificent rim of mountains will trap the stew of air inside your basin as surely as a lid on a cookpot. And those glittering sunbeams will curse you by creating a doomsday photochemical reaction with whatever you put into that photogenic bowl — smoke, car exhaust, wafting industrial toxins — so in time you won’t even be able to see those glorious mountains.”
«
Fascinating long read – with fascinating photos – about a problem that LA has been suffering for 80 years. I stayed in LA a few times, and went running in the hills: it was like breathing glass.
unique link to this extract
Hamas militants behind Israel attack raised millions in crypto • WSJ
Angus Berwick and Ian Talley:
»
Hamas’s lightning strike on Israel last weekend has left observers questioning how the group financed the surprise operation. One possible answer: cryptocurrency.
During the year leading up to the attacks, three militant groups—Hamas, Palestinian Islamic Jihad and their Lebanese ally Hezbollah—received large amounts of funds through crypto, according to a review of Israeli government seizure orders and blockchain analytics reports.
Digital-currency wallets that Israeli authorities linked to the PIJ received as much as $93m in crypto between August 2021 and June this year, analysis by leading crypto researcher Elliptic showed.
Wallets connected to Hamas received about $41m more over a similar time period, according to research by another crypto analytics and software firm, Tel Aviv-based BitOK.
Militants from the PIJ joined Hamas on Saturday in storming into Israel from the Gaza Strip, killing some 900 civilians and abducting at least a hundred more. At least 700 Palestinians have died since Israel retaliated with a wave of attacks on Gaza.
Hamas’s armed wing, the Izz ad-Din al-Qassam Brigades, didn’t respond to a request for comment on the groups’ use of crypto. The PIJ and Hezbollah couldn’t be reached for comment.
«
I like the effort of asking Hamas’s armed wings whether they got their money via crypto. Once you have the crypto, you can shift it to someone friendly – an arms dealer, say – who will accept it and turn it into real money at some point.
unique link to this extract
Global PC shipment decline narrows to just 7% in Q3 2023 • Canalys
»
According to Canalys’ latest data, the worldwide PC market posted another sequential improvement in Q3 2023. While total shipments of PCs amounting to 65.6 million units were down 7% year-on-year, they rose 8% compared to Q2 2023. This represents the smallest annual shipment decline for the industry in over a year and is a further sign of recovery in both inventory levels and underlying demand. Shipments of notebooks dropped 6% annually to 52.1 million units, while desktop shipments were down 8% to 13.5 million units.
“After a tough start to the year, the third quarter of 2023 brought about more positive signs for the global PC market,” said Ishan Dutt, principal analyst at Canalys. “Amid some improvements in the macroeconomic environment, key players across the industry are now expressing cautious optimism as their inventory correction efforts have been largely successful. As a result, pockets of underlying demand strength across all end-user segments are now better reflected in vendors’ sell-in shipment performance. Most major OEMs posted sequential growth in shipments, even after accounting for the demand boost from the education sector that largely materialized last quarter. Looking ahead, Canalys expects this positive trend to continue, with the market set for a return to growth during the highly anticipated holiday season.”
«
Ah, the PC business, quietly ticking over into a sort of oblivion. Lenovo and HP did better than the rest of the market; Apple did a lot worse, but still had about 10% of all sales according to both Canalys and IDC.
Canalys reckons there will be “an additional demand boost from AI” with “AI-capable PCs” becoming mainstream by 2025. Apparently Apple kicked this off with the M1 chip, with its Neural Engine? News to me.
unique link to this extract
Big tech struggles to turn AI hype into profits • WSJ
Tom Dotan and Deepa Seetharaman:
»
AI often doesn’t have the economies of scale of standard software because it can require intense new calculations for each query. The more customers use the products, the more expensive it is to cover the infrastructure bills. These running costs expose companies charging flat fees for AI to potential losses.
Microsoft used AI from its partner OpenAI to launch GitHub Copilot, a service that helps programmers create, fix and translate code. It has been popular with coders—more than 1.5 million people have used it and it is helping build nearly half of Copilot users’ code—because it slashes the time and effort needed to program.
It has also been a money loser because it is so expensive to run.
Individuals pay $10 a month for the AI assistant. In the first few months of this year, the company was losing on average more than $20 a month per user, according to a person familiar with the figures, who said some users were costing the company as much as $80 a month.
Microsoft and GitHub didn’t respond to requests for comment on whether the service is earning money. The profitability picture for GitHub Copilot and other AI-powered assistants will change if computing costs come down.
Microsoft is going with a higher price for its next AI software upgrade. On top of regular monthly charges—starting around $13 for the basic Microsoft 365 office-software suite for business customers—the company will charge an additional $30 a month for the AI-infused version. The AI-powered feature can be instructed to compose emails, create PowerPoint presentations and build Excel spreadsheets independently.
Google, which is releasing a similar AI assistant feature for its workplace software, will also be charging $30 a month on top of the regular subscription fee, which starts at $6 a month.
Microsoft, Google and others have gone with a flat monthly rate, betting that the higher additional charges will more than cover the average expenses of running the technology.
«
Two potential views: 1) this is just like search back in the 1990s, which took some time to find a working business model 2) this can never scale because every query is different. Too early, but don’t bet against the drive for profit.
unique link to this extract
How Ozempic and other weight loss drugs could reshape the food business • Axios
Emily Peck:
»
Earlier this week, Walmart’s U.S. CEO told Bloomberg that customers taking Ozempic buy less food. (Walmart mined its own pharmacy and grocery data to pinpoint customer buying patterns, per Bloomberg.)
“We definitely do see a slight change compared to the total population, we do see a slight pullback,” John Furner said. “Just less units, slightly less calories.” But he added that it’s still early days for Ozempic.
The drug itself has boosted sales of other items at Walmart — folks on the drug “tend to spend more with us overall,” another company exec said this summer.
Meanwhile, Steve Cahillane, CEO of snack maker Kellanova, told CNBC this week that his company — a Kellogg spinoff that makes Pringles and Cheez-It — is watching the Ozempic trends. “But it’s just far too early to forecast this is a headwind.”
It’s early, yes, but investors are paying close attention. In an 82-page note this summer, a team of 17 Morgan Stanley analysts, strategists and associates laid out how obesity medicine could dampen demand for food and reshape the “food ecosystem.” The firm projects that over the next 10 years, 7% of the U.S. population — 24 million people — could be taking these drugs.
Folks on the drug will likely consume 20% fewer calories, they say. In 2035 that would represent 1.3% of overall calories consumed. Analysts also modeled out a bullish scenario where calorie consumption falls by 1.7% and a bearish one at 0.9%.
Increased use of these weight-loss drugs could hurt demand for high-calorie, high-fat and sugary foods — at home or at fast-food outlets.
«
Could “hurt” demand? There’s a particular media mindset – usually reflected in business stories – that any diversion from what currently happens must be bad. Even if what currently happens is that people eat far too many empty calories in high fructose corn syrup, produced in surplus because of subsidies paid to farmers in rural states which have disproportionate representation in the US political system.
unique link to this extract
Unbundling AI • Benedict Evans
A part of a longer, typically thoughtful piece:
»
I think that natural language, voice or text are not necessarily the right interface even if you were talking to an AGI, even if it was ‘right’, and, more fundamentally, that asking a question and getting an answer might be a narrow interface, not a general-purpose one.
The simplest challenge is what to ask. You have a text box and a prompt. What do you type in? What can you ask for? All conversations about AI these days seem to be hunts for metaphors, so, as an analyst, I think it’s interesting to think about Excel. You’re given an infinite grid that could do ‘anything’, so what do you do with it? What would you make? That might be a hard question. An LLM text prompt has a lot of this ‘blank canvas’ challenge, but with even less constraint.
Some of this is familiarity, or exploration, or desire paths, and some of my objection is ‘legacy thinking’. Whenever we get a new tool, we start by forcing it to fit our existing ways of working, and then over time we change the work to fit the new tool. We try to treat ChatGPT as though it was Google or a database instead of asking what it is useful for. How can we change the work to take advance of this?
Excel, like a lot of modern software, tries to help. When you open it today, you don’t get a blank spreadsheet. You get ideas and suggestions. ChatGPT is now trying to do the same thing – ‘what should I do with this?’
This takes me to a second problem, though. Excel isn’t just giving suggestions – those tiles are documents, and documents are the start of a process, not an answer. You can see what you’ve built and what it’s doing, and how far you’ve got. The same sense of creation as process applies to Photoshop, Ableton or Powerpoint, or even a simple text editor. The operative word is editor – you can edit!
Conversely, using an LLM to do anything specific is a series of questions and answers, and trial and error, not a process. You don’t work on something that evolves under your hands. You create an input, which might be five words or 50, or you might attach a CSV or an image, and you press GO, and your prompt goes into a black box and something comes back. If that wasn’t what you wanted, you go back to the prompt and try again, or tell the black box to do something to the third paragraph or change the tree in the image, press GO, and see what happens now.
This can feel like Battleship as a user interface – you plug stuff into the prompt and wait to find out what you hit.
«
I’ve still not found a use for ChatGPT in my work.
unique link to this extract
Dogger Bank: World’s largest offshore wind farm starts exporting power • BusinessGreen News
James Murray:
»
The world’s largest offshore wind farm started exporting power to the UK grid this weekend, marking the latest major milestone for the UK’s offshore wind industry.
Developers SSE, Equinor, and Vårgrønn announced today that the first turbine from the first phase of the Dogger Bank Wind Farm began exporting power at 8:37pm on Saturday evening.
The first power was generated by one of the project’s giant GE Vernova Haliade-X 13MW turbines. It was then transmitted 130 kilometres to the shore along a high-voltage direct current (HVDC) transmission system, marking the first-time HVDC technology has been used on a UK wind farm.
The developers said every rotation of the Haliade-X turbine’s 107 metre blades generates enough power to run an average home for two days.
The first power from the 1.2GW Dogger Bank A wind farm is set to be followed in the coming years by the development of two further wind farms in the Dogger Bank zone, which should ultimately deliver 3.6GW of clean power capacity – enough to power six million homes and deliver yearly CO2 savings equivalent to removing 1.5m cars from the road.
«
Licensed in 2008, onstream 15 years later. For comparison: Hinkley C nuclear power station, licensed in 2012, might come onstream in 2028. Similar production generation. Broadly similar timescales. Very different prices: £26bn for the nuclear power station, £5.5bn for Dogger A (and B, yet to come onstream).
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 