Shared posts

26 Apr 15:46

Google will finally put Keep reminders in Tasks, where they always belonged

by Emma Roth
Google logo with colorful shapes
Illustration: The Verge

The reminders you’ve jotted down in Google Keep will soon show up in Tasks. Google says the long-requested integration is coming over the next year, allowing you to see, edit, and complete your Keep reminders across Tasks, Calendar, and Assistant.

The change will make Tasks the central hub for all to-dos and reminders, hopefully making them far easier to manage. It’s something my colleague David Pierce suggested when talking about Google’s upgraded Keep app last year, saying, “Now, Google, make Keep reminders show up in Tasks. It’s right there!”

Image: Google

Even though Tasks doesn’t support location-based reminders like Keep does, Google states (referring to Keep) that you can still “add time or location-based...

Continue reading…

26 Apr 15:44

20 Million Teams Phone users and 1 Million Teams Rooms – Microsoft Earnings FY24 Q3

by Tom Arbuthnot

Microsoft just released its latest earnings information for FY24 Q3. I always like to examine what information we can glean from a Microsoft 365 perspective, and this time, we got a lot. Here are the highlights.

Microsoft’s total revenue increased 17% to $61.86bn during the first three months of 2024, the third quarter of its financial year, surpassing analyst expectations of $60.88bn.

Earnings per share increased 20% to $2.94, ahead of the expected $2.83.

Microsoft Teams

20 Million+ Teams Phone users – note these are active users, not licenses sold (which is still a much bigger number). Up nearly 30% year-over-year. It will be interesting to see if the recently released Shared Calling will cause this activated seat number to continue to grow.

1 Million+ Teams Rooms — Microsoft has now surpassed a million Microsoft Teams Rooms. The last public number we had was 500,000 in the FY23 Q2 earnings call (January 24, 2023), which was 70% up year over year. So, we continued massive year-over-year growth.

We did not get a new Monthly Active Users number. “When it comes to Teams, we once again saw year-over-year usage growth.”. The last updated monthly active users number was in the FY24 Q1 results, 320 million monthly active users. 80% of the overall number of Office 365 monthly active users

Microsoft Copilot

Microsoft has added 150+ Copilot capabilities since the start of the year.

Nearly 60% of the Fortune 500 now use Copilot with companies like Amgen, bp, Cognizant, Koch Industries, Moody’s, Novo Nordisk, NVIDIA, and Tech Mahindra purchasing over 10,000 seats.

More than 65% of the Fortune 500 now use Azure OpenAI Service

Early adopters of Copilot show increased usage intensity, including a nearly 50% increase in Copilot-assisted interactions per user in Microsoft Teams.

Copilot in Windows is now available on nearly 225 million Windows 10 and Windows 11 PCs, up 2X quarter-over-quarter.

DAX Copilot (Nuance) is being used by more than 200 healthcare organizations, including Providence, Stanford Health Care, and WellSpan Health.

30,000 customers have used Copilot Studio to customize Copilot for Microsoft 365 or build their own, up 175% quarter-over-quarter.

Over 330,000 organizations—including over half of the Fortune 100—have used AI-powered capabilities in Power Platform.

Power Apps now has over 25 million monthly active users, up over 40% year-over-year.

GitHub Copilot is bending the productivity curve for developers.

GitHub Copilot now has 1.8 million paid subscribers, with growth accelerating to over 35% quarter-over-quarter, and continues to see increased adoption from businesses in every industry, including Itaú, Lufthansa Systems, Nokia, Pinterest, and Volvo Cars.

Microsoft 365

Office 365 commercial seats grew 8% year-over-year. Seat growth was again driven by our small and medium business and frontline worker offerings, although growth continued to moderate in SMB.

Office 365 commercial revenue increased 15% “in line with expectations, driven by healthy renewal execution, ARPU growth from continued E5 momentum, and early Copilot for Microsoft 365 progress” (Office commercial revenue grew 13% and 12% in constant currency.)

Office consumer revenue increased 4%, slightly below expectations. Microsoft 365 subscriptions grew 14% to 80.8 million.

LinkedIn revenue increased 10% and 9% in constant currency.

Check out the full earnings call/details here.

The post 20 Million Teams Phone users and 1 Million Teams Rooms – Microsoft Earnings FY24 Q3 first appeared on Tom Talks.

26 Apr 15:37

Teams Overshadowed by Copilot Again Despite Two Huge Milestones

by Tom Wright

Microsoft has revealed two huge milestones for Teams despite the collaboration platform taking a backseat during the vendor’s earnings call as Copilot dominated discussions again.

Microsoft CEO Satya Nadella said that Microsoft crossed the one million Teams Rooms threshold for the first time in the vendor’s fiscal Q3 but again stopped short of revealing the trusty monthly active user figure that had been a staple of Microsoft investor calls since the start of the decade.

Nadella reserved a short section of his prepared remarks for Teams, but the focus was mainly on artificial intelligence.

“When it comes to Teams, we once again saw year-over-year usage growth,” the chief executive said.

“We’re rolling out a new version, which is up to two times faster while using 50 percent less memory for all customers.

“We surpassed one million Teams Rooms for the first time as we continue to make hybrid meetings better with new AI-powered features like automatic camera switching and speaker recognition.”

Alongside Rooms, Nadella also revealed a milestone for Teams Phone.

The CEO said Teams Phone now has 20 million PSTN users, up nearly 30 percent on the same quarter last year.

“All of this innovation is driving growth across Microsoft 365 companies across the private and public sector, including Amadeus, BlackRock, Chevron, Ecolab, Kimberly Clark, [which] all chose our premium E5 offerings this quarter for advanced security, compliance, voice, and analytics,” he added.

Microsoft has not revealed overall Teams stats since October last year, when it pegged monthly active users at 320 million.

The Bigger Picture

Nadella was speaking as Microsoft revealed total sales of $61.9bn, up 17 percent year on year. Operating income jumped 23 percent to $27.6bn.

As is usually the case, the vendor reported sales growth across almost all of its categories and segments, with devices the sole unit to see a decline. Microsoft said this resulted from its focus on “higher-margin, premium products”.

It was, however, Microsoft’s cloud credentials that stole the show.

The vendor’s Intelligent Cloud unit, which houses its Azure public cloud infrastructure technology, brought in $26.7bn in revenue, beating analyst expectations. The sales reveal came shortly after Microsoft unveiled a huge cloud deal with Coca-Cola.

Copilot dominated discussions on the earnings call with Nadella and CFO Amy Hood, albeit without any revenue figures attached to the hype.

Nadella rattled through various Copilot use cases across Azure, Microsoft 365, app development, customer service, and more.

He revealed that nearly 60 percent of Fortune 500 companies are using the generative AI tool, with huge companies, including Cognizant, Koch Industries, Moody’s, Novo Nordisk, NVIDIA, and Tech Mahindra, purchasing over 10,000 seats.

In Other News

Microsoft Teams Classic app will no longer be supported after July 1st, 2024, as Microsoft rolls out a new Teams client.

Microsoft urged users to switch to the new version to avoid service disruptions. Various reminders will be issued leading up to the end of support date, and users can continue using the classic version until the ‘end of availability’, which varies based on operating systems. Admins have options for managing the transition, with Microsoft-managed rollouts and toggle options available.

Earlier this week, Microsoft brought former Meta executive Jason Taylor on board to join its AI supercomputing team. Taylor assumed the role of Corporate Vice President and Deputy CEO at Microsoft, reporting to Kevin Scott, Microsoft CTO.

During his tenure at Meta from 2009 to 2022, Taylor served as Vice President of Infrastructure, overseeing AI, data, and privacy infrastructure and managing the company’s server budgets. He also chaired the Open Compute Project Foundation from 2015 to 2017.

 

 

26 Apr 15:34

The end of coral reefs as we know them

by Benji Jones
An illustration of scuba divers wearing wetsuits and yellow fins swimming over the sea floor, which is strewn with white coral and gravestones.
Paige Vickers/Vox

Years ago, scientists made a devastating prediction about the ocean. Now it’s unfolding.

Several years ago, the world’s top climate scientists made a frightening prediction: If the planet warms by 1.5 degrees Celsius, relative to preindustrial times, 70 to 90 percent of coral reefs globally would die off. At an increase of 2°C, that number jumps to more than 99 percent.

These researchers were essentially describing the global collapse of an entire ecosystem driven by climate change. Warm ocean water causes corals — large colonies of tiny animals called polyps — to “bleach,” meaning they lose a kind of beneficial algae that lives within their bodies. That algae gives coral its color and much of its food, so bleached corals are not only white but also starving. And starved coral is more likely to die.

In not so great news, the planet is now approaching that 1.5°C mark. In 2023, the hottest year ever measured, the average global temperature was 1.52°C above the preindustrial average, as my colleague Umair Irfan reported. That doesn’t mean Earth has officially blown past this important threshold — typically, scientists measure these sorts of averages over decades, not years — but it’s a sign that we’re getting close.

A person wearing a snorkel, black wet suit, and flippers, swims above a coral reef while filming with an underwater camera. David Gray/AFP via Getty Images
Marine biologist Anne Hoggett swims above bleached and dead coral on the Great Barrier Reef in April 2024.
A photo taken from above shows several figures in wet suits and fins swimming in clear blue water above a multi-colored ref with many white spots. David Gray/AFP via Getty Images
Tourists snorkel above a section of the Great Barrier Reef full of bleached and dead coral on April 5, 2024.

So, it’s no surprise, then, that coral reefs are, indeed, collapsing.

In April, the National Oceanic and Atmospheric Administration (NOAA) announced that the planet is experiencing its fourth global “bleaching” event on record. Since early 2023, an enormous amount of coral in the Atlantic, Pacific, and Indian oceans has turned white, including in places like the Great Barrier Reef and the Florida Keys. A record 60 percent of the planet’s coral reef habitat has experienced enough heat to cause bleaching in the last year, NOAA marine scientist Derek Manzello said in a press conference Thursday. In some regions, nearly all of the coral has died.

“What we are seeing now is essentially what scientists have been predicting was going to happen for more than 25 years,” Manzello, who leads the agency’s coral bleaching project, told Vox in April. A build-up of carbon emissions is the underlying cause of warming, which has also supercharged more temporary drivers of ocean heat, including El Niño (which has recently weakened).

Yet coral reefs were collapsing well before the current bleaching crisis. A study published in 2021 estimated that coral “has declined by half” since the mid-20th century. In some places, like the Florida Keys, nearly 90 percent of the live corals have been lost. Past bleaching events are one source of destruction, as are other threats linked to climate change, including ocean acidification.

The past and current state of corals raises an important but challenging question: If the planet continues to warm, is there a future for these iconic ecosystems? What’s become increasingly clear is that climate change doesn’t just deal a temporary blow to these animals — it will bring about the end of reefs as we know them.

Will there be coral reefs 100 years from now?

In the next few decades, a lot of coral will die. That’s pretty much a given. And to be clear, this reality is absolutely devastating. Regardless of whether snorkeling is your thing, reefs are essential to human well-being: Coral reefs dampen waves that hit the shore, support commercial fisheries, and drive coastal tourism around the world. They’re also home to an incredible diversity of life that inspires wonder.

“I’m pretty sure that we will not see the large surface area of current reefs surviving into the future,” said Hans-Otto Pörtner, who was involved in the landmark 2018 report, led by the Intergovernmental Panel on Climate Change (IPCC), that predicted the downfall of tropical reefs at 1.5°C warming. “Every year is going to be worse.”

 NOAA
A map of coral bleaching “alerts,” which indicate where the ocean is unusually warm and bleaching is likely to occur. Red areas have a risk of reef-wide bleaching; magenta and purple regions are at risk of coral death.

But even as many corals die, reefs won’t exactly disappear.

The 3D formation of a typical reef is made of hard corals, such as brain corals, which comprise colonies of polyps that produce skeleton-like structures. When those polyps die, they leave their skeletons behind.

Animals that eat live coral, such as butterfly fish and certain marine snails, will likely vanish. Plenty of other fish and crabs, meanwhile, will stick around because they can hide among those skeletons. Algae will dominate on ailing reefs, as will “weedy” kinds of coral, like sea fans, that don’t typically build the reef’s structure.

Simply put, dead reefs aren’t so much lifeless as they are home to a new community of less sensitive (and often more common) species.

“Reefs in the future will look very different,” said Jean-Pierre Gattuso, a leading marine scientist who’s also involved with the IPCC. “Restoring coral reefs to what they were prior to mass bleaching events is impossible. That is a fact.”

On the timescale of decades, even much of the reef rubble will fade away, as there will be no (or few) live hard corals to build new skeletons and plenty of forces to erode the ones that remain. Remarkably, about 30 percent of the carbon dioxide that we pump into the atmosphere is absorbed by the oceans. When all that CO2 reacts with water, it makes the ocean more acidic, hastening the erosion of coral skeletons and other biological structures made of calcium carbonate, like snail shells.

A scuba diver swims through an underwater cluster of staghorn coral, which resemble floating trees with branches similar to antlers. Jennifer Adler for Vox
Bleached staghorn coral in a nursery run by the Coral Restoration Foundation off the coast of the Florida Keys in September 2023.

There’s only one real solution

For decades now, hard-working and passionate scientists have been trying to reverse this downward trend, largely through restoration. They’ve been “planting” pieces of coral on damaged reefs, much like you might plant saplings in a logged forest. In reef restoration, many scientists and environmental advocates see hope and a future for coral reefs.

But these efforts come with one major limitation: If the oceans continue to grow hotter, many of those planted corals will die too. Last fall, I dived a handful of reefs in the Florida Keys where thousands of pieces of elkhorn and staghorn — iconic, reef-building corals — had been planted. Nearly all of them were bleached, dead, or dying.

“When are [we] going to stop pretending that coral reefs can be restored when sea temperatures continue to rise and spike at lethal levels?” Terry Hughes, one of Australia’s leading coral reef ecologists, wrote on X in April.

Ultimately, the only real solution is reducing carbon emissions. Period. Pretty much every marine scientist I’ve talked to agrees. “Without international cooperation to break our dependence on fossil fuels, coral bleaching events are only going to continue to increase in severity and frequency,” Manzello told me. Echoing his concern, Pörtner said: “We really have no choice but to stop climate change.”

From above, a group of bleached pieces of staghorn coral looks like a boneyard. Jennifer Adler for Vox
A collection of bleached “planted” staghorn coral on a reef in Florida in September 2023.

In the meantime, other stuff might help. Planting pieces of coral may help bring back reefs, at least temporarily, if those corals are more tolerant to threats like extreme heat or disease. And researchers are trying to breed more heat-resistant individuals or identify those that are naturally more tolerant to stress — not only heat, but disease. Even after a summer of extreme bleaching in Florida, many corals survived, according to Jason Spadaro, a restoration expert at Florida’s Mote Marine Laboratory.

Some scientists also see an urgent need to curb other, non-climate related threats, such as intensive fishing. “To give corals the best possible chance, we need to reduce every other stressor impacting reefs that we can control,” Manzello told Vox. Not all marine biologists agree, however, that limiting fishing will do much to help reefs if the world’s oceans continue to warm.

The forecast for this summer is once again grim. The ocean is still unusually hot, and scientists fear that bleaching in the Caribbean could be severe for a second year in a row.

“This [bleaching] event is still growing in size,” Manzello said on a NOAA press call Thursday. “El Nino is dissipating, but the ocean is still so anomalously hot that it won’t take much additional seasonal warming to push temperatures past the bleaching threshold, particularly in the Caribbean.”

This story appeared originally in Today, Explained, Vox’s flagship daily newsletter. Sign up here for future editions.

Update, May 16, 12:10 pm ET: This piece, originally published April 26, has been updated with new information released by NOAA about the ongoing bleaching crisis.

26 Apr 15:26

Tesla’s Autopilot and Full Self-Driving linked to hundreds of crashes, dozens of deaths

by Andrew J. Hawkins
Left side of Tesla Model 3 main screen showing a computer-generated image of an intersection with cars parked on the sides and the Model 3 following another car
Image: Owen Grove / The Verge

In March 2023, a North Carolina student was stepping off a school bus when he was struck by a Tesla Model Y traveling at “highway speeds,” according to a federal investigation that published today. The Tesla driver was using Autopilot, the automaker’s advanced driver-assist feature that Elon Musk insists will eventually lead to fully autonomous cars.

The 17-year-old student who was struck was transported to a hospital by helicopter with life-threatening injuries. But what the investigation found after examining hundreds of similar crashes was a pattern of driver inattention, combined with the shortcomings of Tesla’s technology, resulting in hundreds of injuries and dozens of deaths.

Drivers using Autopilot or the system’s more advanced...

Continue reading…

26 Apr 15:25

Google CEO On Owning ‘Best’ AI, 1,000 New Cloud Products And Workspace Security

by mharanas@thechannelcompany.com (Mark Haranas)
Google CEO Sundar Pichai’s bold remarks during Google’s earnings report include touting Workspace as the leader in cybersecurity, Google Cloud and YouTube having a combined $100 billion run rate by the end of the year, and why a ‘strong partner program’ is needed to drive AI.
24 Apr 22:41

Tesla’s in its flop era

by Andrew J. Hawkins
Photo collage of Elon Musk.
Cath Virginia / The Verge | Photo by Grzegorz Wajda, Getty Images

When Tesla releases its first quarter earnings this afternoon, the company’s CEO Elon Musk will field the usual questions about new products, new factories, and progress toward its futuristic vision of self-driving cars and robot workers. But Musk will also face increasingly urgent questions about its current state of affairs — and why everything seems to be going to shit.

Earlier this month, the company reported its first year-over-year sales drop in four years, a sign of rougher waters ahead. Tesla’s stock has fallen more than 40 percent since the start of the year, including a 13 percent drop in the last week. The company laid off over 14,000 employees last week, 10 percent of its global workforce — which could end up being closer to...

Continue reading…

24 Apr 22:32

Neat Bar Pro Now Pairs with Shure’s Microflex Ecosystem

by James Stephen

Neat has partnered with Shure to improve audio-visual user experiences by enabling the seamless integration between the Neat Bar Pro and various Shure Microflex Ecosystem solutions.

By combining the video technology from the Neat Bar Pro with the audio capabilities of Shure’s Microflex Ecosystem, Neat believes users can gain the “optimum combination”.

Furthermore, the union of technologies can provide and engaging and equitable meeting experience for all meeting spaces, however complex they may be, ensuring can take part in the conversation.

Peter James, VP of Global Business Development at Shure, pointed to some of the benefits that the technological partnership can deliver:

“Our integrated solutions can now easily be installed where customers want a front-of-room, all-in-one device and extraordinary audio coverage throughout their space.

“We are thrilled to provide a first-of-its-kind combination for large conferencing environments that will provide better experiences and more reliable collaboration.”

‘Cutting-Edge Collaboration’

The video device designer, Neat, explains why this is such a powerful technology collaboration in a recent article by its Senior Director of Global Alliances, Sherri Pipala.

According to Pipala, the Neat Bar Pro includes the “most advanced” audio and video technology, including auto-framing, zoom, and speaker tracking no matter where they are situated in a room.

In fact, the Neat Symmetry feature can not only track who is talking but the gestures and expressions of other meeting participants at the same time.

Adding to this, a range of Shure Microflex solutions will enable the Neat Bar Pro to be deployed in more complex spaces, while delivering “crystal clear” audio for varying speaker configurations, helping to streamline audio-visual set ups.

Shure solutions can be connected to the Neat Bar Pro through a single USB-C port, reducing excess cables.

Neat and Shure Combos

The Shure Microsoflex Advance MXA902 Integrated Conferencing Ceiling Array and the Shure ANIUSB-MATRIX USB Audio Network interface are recommended as integration options for large boardrooms or where it is critical that one or more speakers can be seen and heard.

The Microflex Advance MXA920 Ceiling Array Microphone, and IntelliMix P300 Audio Conferencing Processor and MXN5W-C Networked Loudspeaker would be well suited for all -hands training rooms and divisible spaces.

The Neat Bar Pro, which has previously been reviewed by the UC Today Team, was launched in October 2020 and gained Teams certification in March 2022, adding to the interoperability it already had with Zoom.

Now, all these combination options have been certified for Zoom Rooms.

John Stearns, Global Head of Zoom Spaces at Zoom, believes this partnership represents real value for its customers: “Zoom customers are looking for exceptional video and audio experiences in all kinds of meeting spaces, which is why we’re excited about this partnership between Neat and Shure.

“Together, they provide a simple, easy-to-use solution that extends to the most challenging large spaces.”

Earlier this month, Neat unveiled the Neat Bar Generation 2, the latest iteration of its  all-in-one AV bar designed for premium meeting experiences.

UC Today’s George Malim recently uncovered the various benefits of the Shure Microflex Ecosystem that could make it the ideal collaboration reinforcement for every meeting space.

 

 

24 Apr 22:31

AT&T’s ‘Endgame’ Is To Boost Business Revenue Through 5G, Fiber, CEO Says

by gnarcisi@thechannelcompany.com (Gina Narcisi)
AT&T has a plan to boost business revenues through a mix of its core 5G and fiber-based connectivity services and legacy voice services decline, the telecom giant’s CEO John Stankey said during AT&T’s first-quarter 2024 earnings call.
20 Apr 23:59

Target confirms it’s all but completely ditching DVDs in physical stores

by Emma Roth
A photo showing the exterior of a Target in Texas
Image: Target

Target is scaling back the presence of physical media in its stores. A spokesperson for Target tells IGN that the company is “transitioning the limited assortment of DVDs” it sells in stores to its website.

“Moving forward, we’ll offer select DVDs in stores when they are newly released or during key times throughout the year when they are more popular, like for gift giving during the holidays,” Target says. That will make Target’s physical DVD section even smaller and only available during specific times. You’ll still be able to buy “thousands” of DVDs on Target’s website, but I have to wonder how broad its selection will be.

Reports about Target backing away from DVDs first emerged on Wednesday, when the X account @PhysicalMedia posted...

Continue reading…

20 Apr 03:11

Microsoft To Invest $1.5B In UAE AI Developer G42

by wmillward@thechannelcompany.com (Wade Tyler Millward)
Microsoft and UAE AI company G42 partner up with help from the Biden administration.
19 Apr 18:26

NEC Exits Premises-based UC

By Dave Michels
Though it exits UC, NEC’s UCaaS strategic alliance with Intermedia, which created UNIVERGE BLUE, continues.
19 Apr 05:28

The GOP Is Blocking A Last Ditch Effort To Bring Cheap Broadband To Poor Americans

by Karl Bode

The FCC’s Affordable Connectivity Program (ACP), part of the 2021 infrastructure bill, currently provides 23+ million low-income Americans a $30 broadband discount every month. But those 23 million Americans are poised to soon lose the discount because key Republicans — who routinely dole out billions of dollars on far dumber fare — refuse to fund a $4-$7 billion extension.

As a result, the FCC is informing struggling Americans that their broadband bills are all about to jump significantly as the program starts to wind down. There’s a last ditch effort to save the program, but it’s unclear if it has traction in one of the least productive Congress’ in U.S. history:

“Speaking at an event hosted by USTelecom on Thursday, FCC Commissioner Geoffrey Starks called the ACP “the most effective program we’ve ever had” for incentivizing low-cost internet plans, and he mourned the fact that its time could be running out.”

The ACP’s death is an interesting case because the telecom industry generally supports it, many Republicans support it, and many Democrats support it.

Big ISPs support it because it basically involves throwing money at them to temporarily reduce broadband prices that wouldn’t be high in the first place if big ISPs hadn’t spent the last 40 years lobbying to crush competition and regulatory oversight (usually with the help of corrupt centrist Democrats and Republicans).

But a handful of Republicans, nervous that the popular program could help Democrats during an election season, are engaging in obstructionism. House Speaker Mike Johnson appears to be slow walking the emergency funding proposal to death.

Most “both sides” news orgs, always wary of upsetting sources and advertisers, lack the courage to inform readers the program is being specifically killed by Republicans. And given these kinds of bread and butter issues don’t generate clicks and ad engagement in the attention infotainment economy, many outlets just won’t cover it at all.

Republicans claim their concern is about costs but it’s not, really; you’ll recall the Trump tax cuts doled out $42 billion to AT&T alone in exchange for a bunch of layoffs. These same Republicans routinely throw absurd gobs of money at all sorts of ideological dipshittery and badly managed corporate handouts. The fiscal conservatism is a performance the press is happy to help parrot to the public.

Some ISPs, like Verizon, say they’ll continue to offer discounts to poor people for about six months even after the program ends. At which point poor Americans are just shit out of luck, and their monthly bill will return to its usual expensive state, potentially driving them offline.

The $42.5 billion in infrastructure bill broadband subsidies flowing to the states try to at least nudge big ISPs to offer a cheaper, low-income broadband plan if they take taxpayer money, but they’re lobbying to have those requirements killed.

Ideally you wouldn’t need such requirements or programs like the ACP (that temporarily and artificially lower rates) if we had policymakers with the political courage to take direct aim at the real problem: highly concentrated monopoly power and the corruption and regulatory capture that protects it.

That’s what causes broadband competitive market failure and high prices across much of the U.S., but Congress and the FCC can’t even admit it exists, much less propose a solution that might offend telecom giants closely tethered to our domestic surveillance systems. The ACP was a backup plan to do the bare minimum for the least fortunate; and Congress couldn’t even accomplish that.

18 Apr 23:54

How to tame notifications on your Android phone

by Allison Johnson
Hand holding Android phone against illustrated background
Quiet, please. | Illustration by Samar Haddad / The Verge

Notifications are the soundtrack to modern digital life. When they’re out of control, they make it hard to hear anything else going on. But if you manage them well, you can turn them into a subtle background noise. App developers want you to opt in to every notification possible, but thankfully, Android provides a lot of tools to help you tune in to the helpful stuff and tune out the distractions. It takes a little work upfront, but the payoff is worth it.

I used a Google Pixel 8 Pro running Android 14 and a Samsung Galaxy Z Flip 5 running One UI 6.1 to write the following guide — steps may be slightly different on other Android phones.

Notifications in the settings menu

On either a Pixel or Galaxy phone, a good place to start is by...

Continue reading…

18 Apr 23:16

Meta says Llama 3 beats most other models, including Gemini

by Emilia David
Image of Meta’s logo with a red and blue background.
Illustration by Nick Barclay / The Verge

The next generation of Meta’s large language model Llama, which releases today to cloud providers like AWS and to model libraries like Hugging Face soon, performs better than most current AI models, the company said in a blog post.

Llama 3 currently features two model weights, with 8B and 70B parameters. (The B is for billions and represents how complex a model is and how much of its training it understands.) It only offers text-based responses so far, but Meta says these are “a major leap” over the previous version. Llama 3 showed more diversity in answering prompts, had fewer false refusals where it declined to respond to questions, and could reason better. Meta also says Llama 3 understands more instructions and writes better code...

Continue reading…

18 Apr 23:14

Is It Too Late to Do Something About Dangerous “Forever Chemicals”?

by Mary Harris
For many, the EPA’s new regulations are a day late and a dollar short.
18 Apr 23:14

Meta launches Llama 3, wide availability with more versions on deck

by Larry Dignan
By Larry Dignan

Meta launched its Llama 3 open source large language model and said it will be available on AWS, Databricks, Google Cloud, Microsoft Azure, Snowflake, IBM WatsonX, Nvidia NIM and have support from enterprise hardware platforms.

Yes, we're in the age of weekly LLMs that leapfrog each other, but Llama 3, which will initial come in 8B and 70B parameters with more versions on deck, is of interest to enterprises.

Why? Companies are likely to look to capable open source LLMs and then look to fine tune them with enterprise data. Llama 3 represents a backdoor enterprise play for Meta.

In a blog post, Meta outlined the Llama 3 effort, which is a big leap over Llama 2. "Improvements in our post-training procedures substantially reduced false refusal rates, improved alignment, and increased diversity in model responses. We also saw greatly improved capabilities like reasoning, code generation, and instruction following making Llama 3 more steerable," said Meta.

Meta added that it evaluated Llama 3 based on prompts covering 12 use cases including brainstorming, advice, coding, creative writing, extraction and summarization to name a few. In this use case testing scenario, Llama 3 70B topped Claude Sonnet, Mistral Medium and Llama 2.

According to Meta, Llama 3 is being deployed across its applications including Facebook, Instagram and WhatsApp. Llama 3 is also available for a spin on the web.

Going forward, Meta said a 400B parameter model is training and it'll "release multiple models with new capabilities including multimodality, the ability to converse in multiple languages, a much longer context window, and stronger overall capabilities."

08 Apr 23:32

Microsoft’s new safety system can catch hallucinations in its customers’ AI apps

by Emilia David
Microsoft logo
Illustration: The Verge

Sarah Bird, Microsoft’s chief product officer of responsible AI, tells The Verge in an interview that her team has designed several new safety features that will be easy to use for Azure customers who aren’t hiring groups of red teamers to test the AI services they built. Microsoft says these LLM-powered tools can detect potential vulnerabilities, monitor for hallucinations “that are plausible yet unsupported,” and block malicious prompts in real time for Azure AI customers working with any model hosted on the platform.

“We know that customers don’t all have deep expertise in prompt injection attacks or hateful content, so the evaluation system generates the prompts needed to simulate these types of attacks. Customers can then get a...

Continue reading…

02 Apr 14:06

AI “agents” could do real work in the real world. That might not be a good thing.

by Kelsey Piper
An illustration shows a screen with an anthropomorphic robotic head attached. Speech and text bubbles appear to float out of the screen and into the air.
Malorny/Getty Images

Why AI agents that could book your vacation or pay your bills are the next frontier in artificial intelligence.

ChatGPT and its large language model (LLM) competitors that produce text on demand are very cool. So are the other fruits of the generative AI revolution: art generators, music generators, better automatic subtitles and translation.

They can do a lot (including claim that they’re conscious, not that we should believe them), but there’s one important respect in which AI models are unlike people: They are processes that are run only when a human triggers them and only to accomplish a specific result. And then they stop.

Now imagine that you took one of these programs — a really good chatbot, let’s say, but still just a chatbot — and you gave it the ability to write notes to itself, store a to-do list and the status of items on the to-do list, and delegate tasks to other copies of itself or other people. And instead of running only when a human prompted it, you had it work on an ongoing basis on these tasks — just like an actual human assistant.

At that point, without any new leaps in technology whatsoever — just some basic tools glued onto a standard language model — you’d have what is called an “AI agent,” or an AI that acts with independent agency to pursue its goals in the world.

AI agents have been called the “future of artificial intelligence” that will “reinvent the way we live and work,” the “next frontier of AI.” OpenAI is reportedly working on developing such agents, as are many different well-funded startups.

They may sound even more sci-fi than everything else you’ve already heard about AI, but AI agents are not nonsense, and if effective, could fundamentally change how we work.

That said, they currently don’t work very well, and they pose obvious challenges for AI safety. Here’s a quick primer on where we’re (maybe) headed, and why.

Why would you want one of these?

Today’s AI chatbots are fun to talk to and useful assistants — if you are willing to overlook a set of limitations that includes making things up. Such models have already found sizable and important economic niches, from art to audio and video transcription (which have been quietly revolutionized over the last few years) to assisting programmers with tools like Copilot. But the investors pouring hundreds of billions of dollars into AI are hoping for something more transformative than that.

Many people I talk to who use AI in their work describe it as like having a slightly scatterbrained but very fast intern. They do useful work, but you have to define each problem for them and carefully check their work, meaning that much of what you might gain in productivity is lost in oversight.

Much of the economic case for AI is that it could do more than that. The people at work on AI agents hope that their tools won’t just help software developers, but that the tools could be software developers. In this future, you wouldn’t just consult AI for trip planning ideas; instead, you could simply text it “plan a trip for me in Paris next summer,” as you might a really good executive assistant.

Today’s AI agents do not live up to that dream — yet. The problem is that you need a very high accuracy rate on each step of a multistep process, or very good error correction, to get anything valuable out of an agent that has to take lots of steps.

But there’s good reason to expect that future generation AI agents will be much better at what they do. First of all, the agents are built on increasingly powerful base models, which perform much better on a wide range of tasks, and which we can expect to continue to improve. Secondly, we’re also learning more about how to build agents themselves.

A year ago, the first publicly available AI agents — AutoGPT, for example, which was just a very simple agent based on ChatGPT — were basically useless. But a few weeks ago, the startup Cognition Labs released Devin, an AI software engineer that can build and deploy entire small web applications.

Devin is an impressive feat of engineering, and good enough to take some small gigs on Upwork and deliver working code. It had an almost 14 percent success rate on a benchmark that measures ability to resolve issues on the software developer platform GitHub.

That’s a big leap forward for which there’s surely an economic niche — but at best, it’s a very junior software engineer who’d need close supervision by a more senior one. Still, like most things AI, we can expect improvement in the future.

Should we make billions of AI agents?

Would it be cool for everyone in the world to have an AI personal assistant who could plan dinner, order groceries, buy a birthday present for your mom, plan a trip to the zoo for the kids, and pay your bills for you while notifying you of any unexpected ones? Yes, absolutely. Would it be incredibly economically valuable to have AI software engineers who can do the work of human software engineers? Yes, absolutely.

But: Is there something potentially worrying about creating agents that can reason and act independently, earn money independently, make copies of themselves independently, and do complex things without human oversight? Oh, definitely.

For one, there are questions of liability. It’d be just as easy to make “scammer” AIs that spend their time convincing the elderly to send them money as it would to make useful agents. Who would be responsible if that happens?

For another, as AI systems get more powerful, the moral quandaries they pose become more pressing. If Devin earns a lot of money as a software engineer, is there a sense that Devin, rather than the team that created him, is entitled to that money? What if Devin’s successors are created by a team that’s made up of hundreds of copies of Devin?

And for those who worry about humanity losing control of our future if we build extremely powerful AI systems without thinking about the consequences (I’m one of them), it’s pretty obvious why the idea of AIs with agency is nerve-racking.

The transition from systems that act only when users consult them to systems that go out and accomplish complex goals in the real world risks what leading AI scientist Yoshua Bengio calls “rogue AI”: “an autonomous AI system that could behave in ways that would be catastrophically harmful.”

Think of it this way: It’s hard to imagine how ChatGPT could kill us, or could even be the kind of thing that would want to. It’s easy to imagine how a hyper-competent AI executive assistant/scam caller/software engineer could.

For that reason, some researchers are trying to develop good tests of the capabilities of AI agents built off different language models, so that we’ll know in advance before we widely release ones that can make money, make copies of themselves, and function independently without oversight.

Others are working to try to set good regulatory policy in advance, including liability rules that might discourage unleashing an army of super-competent scammer-bots.

And while I hope that we have a few years to solve those technical and political challenges, I doubt we’ll have forever. The commercial incentives to make agent AIs are overwhelming, and they can genuinely be extremely useful. We just have to iron out their extraordinary implications — preferably before, rather than after, billions of them exist.

A version of this story originally appeared in the Future Perfect newsletter. Sign up here!

02 Apr 13:55

Food delivery fees have soared. How much of it goes to workers?

by Whizy Kim
Close-up photo of someone looking at a burger on a food delivery app on their phone.
Food delivery apps have recently added new fees in response to minimum pay rules in New York City and Seattle. | Getty Images/iStockphoto

Amid delivery discourse, a new report claims to shed some light on DoorDash’s delivery fees.

No one is happy about the delivery apps. Not the customers, who feel gouged by an avalanche of fees. Not restaurants, who feel gut-punched by the commission apps take from them. Certainly not delivery workers, who have long been rewarded with a pittance for doing a job that, in a city like New York, has a higher injury rate than that of construction workers.

Amid this dogpile of disgruntlement, the merry-go-round of debating the value of food delivery keeps spinning. After all, some people, especially those with disabilities, rely on such services — but then, it is difficult work, and everyone ought to tip well. Another faction argues that this isn’t fair, because it’s already so unaffordable. The delivery apps themselves recede somewhat into the background, as if their existence is a given. They’re merely fulfilling a demand in the market, naturally taking a cut for themselves — two plus two equals four. Our desire to consume is seen as the problem, the having-cake-and-eating-it-too mentality of expecting affordable convenience.

But we should give credit where it’s due. Delivery apps have expended a lot of effort (and money) making the case that we — restaurants, workers, and consumers — desperately need them. Unhappy about the state of things now? You’ll really be pulling your hair out if you try to force the apps to change. In New York City and Seattle, new minimum pay laws for delivery workers recently went into effect.

Immediately, additional “regulatory” fees were charged to customers, and restaurants and delivery workers complained that orders dropped, with Uber claiming in a blog post that they had dipped by 30 percent. Neither city’s minimum wage laws have forced delivery apps to tack on new fees, but both DoorDash and Uber Eats have introduced them nonetheless. (Grubhub did not.) The message is clear: If you try to mediate how the apps operate, things will just get worse.

Now, Sens. Elizabeth Warren (D-MA), Bob Casey (D-PA), and Ben Ray Luján (D-NM) have sent letters to DoorDash and Uber calling on the companies to stop charging junk fees. “When additional hidden fees nearly triple the price of an order, that is price gouging — plain and simple,” reads a copy of the letter sent to Vox.

The letter also requests answers to exactly what the fees cover, including how much of the fees have gone to delivery workers versus to executive pay, among other questions, by no later than May 15.

An Uber spokesperson told Vox that there were “consequences to bad regulations and we made these consequences clear in repeated testimony that both cities chose to disregard.” A DoorDash spokesperson wrote that its platform “has to work for everyone who uses it — Dashers, merchants, and customers alike — which is why we’ve opposed these extreme new rules.” They continued that the new laws “require platforms like DoorDash to pay well above the local minimum wages, not including additional pay for mileage and tips. Just as we warned, the increased costs created by these regulations have led to an alarming drop in work for Dashers and lost revenue for small businesses.”

“Grubhub is complying with the new pay standards in New York City and Seattle, and we have made adjustments to our platform to run a sustainable business given the added costs to operate in these markets,” a Grubhub spokesperson told Vox. “We warned that these ill-conceived policies would have immediate negative impacts on the people they were intending to help, and the data is showing that to be the case.” Grubhub supports increased earnings for workers, but has previously cautioned how pay laws could impact workers’ ability to choose when and how much they work.

Some headlines have already declared app-delivery regulations a failure; Seattle City Council President Sara Nelson has already proposed a new ordinance that lowers the pay rate. At the crisis point of consumers fed up with the cost of food delivery, companies like DoorDash, Uber Eats, and Grubhub — the three biggest in the US — are insisting on their irreplaceable value to the restaurants, consumers, and workers who have long complained about them.

“These guys are doing what I call a corporate tantrum”

Kimberly Wolfe, a delivery app driver in Seattle who fought for the wage law with an advocacy group called Working Washington, isn’t buying it. “These guys are doing what I call a corporate tantrum,” she tells Vox. “They’re just cutting off their nose to spite their face.”

What apps take from restaurants and customers

To be sure, delivery apps are convenient. For this ease of use, customers are painfully up-charged. Menu prices are almost always more expensive than ordering directly from restaurants. Then there are the line-item fees that appear on the receipt. There’s the delivery fee, but also the frustratingly generic “service fee” that could cover anything from keeping the apps’ servers up to paying their drivers. The letter sent by Sens. Warren, Casey and Luján notes that US lawmakers demanded more transparency on these fees last February too — but the responses from DoorDash and Uber provided little clarity, according to the new April 16 letter. The letter also points out how fees have ballooned alongside executive compensation: in 2020, DoorDash CEO Tony Xu was the highest paid CEO in Silicon Valley with a pay package worth $413 million.

DoorDash charges a 15 percent service fee that starts at a $3 minimum. Uber Eats charges an unspecified service fee that depends on basket size. Browsing Grubhub in Seattle, I loaded a sample $62 food order and was levied a $14 service fee. Then add the taxes and tip. For the privilege of having a meal delivered to your home — something pizza and Chinese restaurants have done for at least half a century — you might find yourself paying nearly double the cost of just the food.

For restaurants, there’s a price as well. For the privilege of being found in the apps’ centralized hubs, apps can swipe as much as 30 percent of an order’s subtotal from restaurants, even collecting a commission on pickup orders. That’s if diners choose them over the influx of ghost kitchens and promoted partners.

Much attention has been paid to the fact that delivery apps aren’t profitable, or were on a long road to becoming profitable — but that’s in large part because they chose to invest aggressively in growth over being in the black at the end of the year. Last year, DoorDash’s profit margin was nearly 49 percent. Even after deducting a bunch of its biggest expenses, including driver pay, Uber’s delivery segment pocketed $1.5 billion, an increase of 173 percent from 2022.

Out of $8.6 billion in revenue in 2023, DoorDash spent almost $2 billion on sales and marketing, and another billion on R&D. It also spent $750 million last year buying back its own stock, a move often used by corporations to boost stock value. Uber has also long poured money into sales and marketing, which includes things like promotions and discounts, as well as R&D, in order to grow. This year, the company is preparing to shell out a cool $7 billion on stock buybacks.

What (little) apps provide to delivery workers

While customers find themselves paying $9-plus service fees on a delivery order, the worker handing you the food might only get a few dollars, all while paying for their own vehicle and fuel.

Wolfe recalls how paltry some of the payouts were before the Seattle wage law, when she would see $2 to $3 for an order before tips. In May 2022, Working Washington aggregated data from over 400 delivery jobs in the Seattle area and found that restaurant delivery workers were making on average $8.71 per hour after deducting basic expenses such as gas, which was far below the city’s 2022 minimum hourly wage of $17.27. During a Working Washington protest at City Hall in 2022, paper bags with receipts showing how much a worker had made on a delivery order were put on display.

“There were quite a few that were negative,” says Wolfe. “Once you figured expenses and all that, you were basically paying them to deliver.”

A 2022 study from NYC’s Department of Consumer and Worker Protection (DCWP) found that, after expenses, food delivery workers in the city were making an average of $11.12 per hour — again, sub-minimum wages. Crucially, customer tips made up about half of a delivery driver’s total earnings before expenses. (Data from Solo, which makes software for app-based gig workers, shows that tips make up a similar proportion of pay in Seattle.) A more recent report on the adopted minimum pay projected that drivers’ annual earnings after expenses (and accounting for the common practice of working for multiple apps) would rise from $11,970 in 2021 to $32,500 by 2025. Yet this calculation relies on a key assumption: that customers would keep tipping about the same amount as before the wage law.

It’s hard to imagine that tipping rates in Seattle and NYC would stay the same given that the apps have added friction to the process. On both DoorDash and Uber Eats in these two cities, the tipping prompt now comes up after delivery, not at checkout, when diners are less likely to engage with the app. On GrubHub, the option to tip at checkout is still available, but many NYC-area restaurants on the platform now show lower default tipping options that max out at 12 percent. (Of course, a customer can still input a custom amount.)

“After the minimum wage started, I would be on the apps and after two hours it would lock me out”

It has also likely gone down because deliveries have gone down. While a spokesperson for the DCWP told StreetsBlog that “mass lockouts” were not occurring, some workers in NYC report that the apps are now locking them out, restricting the number of hours they work. Justice for App Workers, a coalition of rideshare and delivery workers, held a rally in front of New York’s city hall on March 27 to demand that the city address the lockouts. Food delivery workers are saying that they’re “unable to work for hours and days on end,” according to a statement released by the group.

Bimal Ghale, a delivery worker in New York who is part of the Justice for App Workers group, told Vox through an interpreter that he used to work five to six hours at a time. “After the minimum wage started, I would be on the apps and after two hours it would lock me out,” he says. “The apps claim the area isn’t busy.” But Ghale is still delivering in the same neighborhoods he did before the new pay law, and the DCWP has also stated that orders have “remained steady.”

An Uber spokesperson said that the city had known workers’ access to apps would become limited due to the new hourly pay rule. “Since the rule went into effect, nearly 6,000 couriers have lost access to the platform, nearly 20,000 people are on the waitlist to work on the app,” the spokesperson said.

Since last December, when the pay rule went into effect in NYC, at least 500 complaints have been lodged with the DCWP alleging that apps aren’t following it. A DCWP spokesperson told Vox that the department was monitoring compliance.

In Seattle, DoorDash has slapped a $4.99 regulatory fee on all orders, and in NYC it charges an extra $1.99. It’s unclear how these meaningfully differ from the catchall service fee, a portion of which can also cover worker pay — except that the labeling points the finger at the law for higher prices. DoorDash’s regulatory response fees are meant to cover the costs of new regulations. The DCWP estimates that if apps passed on only half of their labor costs to consumers, instead of all of it, they would still pocket $232 million a year in revenue. It’s not a given that the apps have to charge us more to pay their workers better.

A new report released by advocacy group Working Washington claims to shed some light on how these fees might be divvied up. By comparing customer receipts and driver pay for 31 delivery orders placed since Seattle’s new minimum pay law was implemented, the group estimated that only about half of the fees DoorDash charges customers and restaurants went to workers. The report alleges that the company could ax the $4.99 regulatory fee and still retain a 30 percent margin on each order.

DoorDash says that these calculations are inaccurate. “This is misinformation being sold as a ‘report’ — full of outrageous claims, exaggerations, and in many cases outright fabrications meant to deceive the public and policymakers,” a spokesperson told Vox, noting that the flat regulatory fee helps “offset only some of the costs associated with this burdensome law.”

“As we have long said, if costs can be reduced through compromise legislation, we will look at all ways to reduce costs for consumers,” the spokesperson continued.

Apps cry that their hands are tied

Not long after the pay law went into effect, DoorDash published a blog claiming that Seattle businesses had already lost over $1 million in revenue and that workers were making less because orders on the platform had dropped. Grubhub’s write-up on the law’s adverse effects claims that tips are down 26 percent, with no mention of the fact that many of its Seattle-area merchants now show a lower range of tipping options — a tactic the company has used before.

None of these tactics are new. Just look at what happened in California after the passage of a ballot initiative called Proposition 22 a few years ago, which allowed app-based gig work companies like Uber and DoorDash to classify their workers as independent contractors, saving them a lot of money. In exchange, they agreed to pay 120 percent of the minimum wage for every hour of trip time — as in, time spent logged on the app, waiting for a ride or for an order to appear, would not count. App companies spent hundreds of millions of dollars backing Prop 22, even threatening to pull out of California if it failed to pass. They also warned that, without Prop 22, prices would go up for customers. A month after the successful vote, delivery apps announced fee increases anyway.

The math doesn’t add up. On the one hand, delivery apps play up the fact that they’re just intermediaries helping facilitate the sale or delivery of a product — they’re not employers, who would be on the hook for far greater payroll taxes and other employment costs than what apps currently pay. On the other hand, they command a steep price from restaurants and customers for matchmaking, of which the workers only see a narrow slice. The apps don’t make the food taste better, or deliver faster, and it’s obviously not cheaper. So who, exactly, benefits from their existence? What do they really add to the tangle of relationships we call the economy? If app companies leave cities like Seattle and New York to avoid having to pay higher labor costs, who would lose?

Wolfe doesn’t seem worried. Her thinking is that if they can’t run a competent business, perhaps they shouldn’t be in business. “Don’t let the door hit you,” she says. “Because you want capitalism — baby, that’s capitalism.”

Update, April 26, 10:35 am ET: This story was originally published on April 2 and has been updated multiple times, most recently with a new report that claims to reveal how much of DoorDash’s fees goes to paying workers.

24 Mar 20:38

Victrola’s cheapest Sonos-ready turntable has fallen to yet another low

by Sheena Vasani
The Victrola Stream Onyx turntable sitting on a black desk.
The stylish Victrola Stream Onyx is on sale for $334.20 as part of Amazon’s Big Spring Sale. | Image by Jennifer Pattison Tuohy / The Verge

We’re two days into Amazon’s Big Spring Sale, and we’ve already seen some pretty good deals on personal audio gear and home theater equipment, including a new low on Apple’s latest AirPods Pro and a rare discount on the Sonos Ray and second-gen Sonos Beam. The sale has gotten even better since it kicked off, too, especially now that Amazon is selling the Victrola Stream Onyx for a new low of $334.20 ($266 off).

Victrola’s entry-level Works with Sonos turntable lets you stream all your records to your Sonos speakers like its pricier siblings — including the forthcoming Stream Sapphire and midrange Stream Carbon. What you don’t get is the Sapphire’s unique ability to stream records to practically anything or the Carbon’s high-end Ortofon...

Continue reading…

21 Mar 20:54

Threads’ fediverse beta opens to share your posts on Mastodon, too

by Emma Roth
An image showing the Threads logo
Image: The Verge

Threads is rolling out a beta of its fediverse integration in the US, Canada, and Japan. In a post on Thursday, Meta CEO Mark Zuckerberg announced that toggling on the feature will let you cross-post and view likes from other federated platforms, like Mastodon.

Threads previewed its fediverse integration earlier this week during the FediForum. As outlined on its support page, Meta says that you must have a public account to turn on fediverse sharing, which will allow users on other servers to “search for and follow your profile, view your posts, interact with your content, and share your content to anyone on or off their server.”

Image: Meta

There are still a few limitations, though. The beta currently doesn’t let...

Continue reading…

21 Mar 20:52

The lock-in problem at the heart of the Apple monopoly lawsuit

by Victoria Song
Illustration of an iPhone surrounded by green and blue message bubbles.
Green bubbles are actually a big part of the DOJ’s case against Apple. | Illustration by Cath Virginia / The Verge

It’s no secret that Apple products work best if you stick with an iPhone. It turns out that’s a big reason why Apple landed in hot water today with the US Department of Justice, which alleges that the company went too far in locking down messaging, smartwatches, and digital wallets to intentionally hobble its rivals.

This won’t be a surprise to most consumers. We’ve all known for years about green bubbles and that you can’t bring your Apple Watch to an Android phone. What the DOJ is saying is that, altogether, this series of protective policies makes it extremely difficult for an iPhone user to leave its walled garden, limiting competition so much that it breaks the law.

Messaging

Green bubbles make a key appearance in the lawsuit. It’s...

Continue reading…

21 Mar 20:51

Reinventing Service Provider Voice

By Dave Michels
Service providers can still be disruptors in voice, but doing so likely means taking their own advice and moving to the cloud.
21 Mar 16:16

Meet the EPA’s new Choose Your Own Adventure! regulation for car pollution

by Umair Irfan
A new Ford Transit Custom Plug-in Hybrid van which is connected to FordLive is displayed during a vehicle show on September 2, 2021, in Birmingham, England.
The EPA has finalized new emissions rules for light-duty vehicles like cars and medium-duty vehicles like transit vans that will push automakers to make cleaner vehicles like electrics and plug-in hybrids. | John Keeble/Getty Images

Here’s what the federal rules mean for car companies, the climate, and you.

The Environmental Protection Agency has officially cemented new pollution rules for cars, pickup trucks, vans, and SUVs that the Biden administration called the US’s strongest-ever clean vehicle regulations. The EPA says the new rules will avert 7 billion tons of greenhouse gas emissions and provide close to $100 billion in savings per year across the country in the form of fuel costs, lower maintenance needs, and health benefits.

The challenge for the government and carmakers, though, will be actually getting people to buy enough of these cleaner cars to move the needle. And that may be harder than regulators and the industry thought.

Because of that concern — and some feedback from car companies that the technology and demand wouldn’t materialize fast enough — these new standards are notably more lenient than those the EPA proposed almost a year ago. The EPA slowed the pace at which companies would need to lower pollution outputs and offered more options to meet the targets, which kick in for model year 2027 and run through 2032.

The regulations are not specifically designed to promote battery electric vehicles and instead allow carmakers to pursue a variety of options, like more efficient gasoline engines, plug-in hybrids, and hydrogen-powered cars.

“We are providing a little bit more lead time so that these investments can occur, not just in the automobiles themselves, but the infrastructure,” EPA administrator Michael Regan told Vox.

Car buyers will see even more fuel-efficient and electric models in showrooms going forward, but the most gas-guzzling cars and trucks won’t be available for sale much longer. Some of these cleaner cars may have a higher sticker price at first, but the EPA says that the new rules will save drivers money over time and that prices will fall as manufacturers scale up.

That’s a key question: While EVs are more popular than ever, demand has fallen short of what many manufacturers predicted. A Yahoo Finance/Ipsos poll last year found that one in three prospective car buyers said it was likely their next car would be an EV. Still, in 2023, only 7.6 percent of new cars sold were electric, due in part to concerns about price, performance, and spotty charging infrastructure.

Some companies are scaling back their electric offerings, while others are renewing their bets on hybrid-electric cars.

This underscores why decarbonizing transportation is one of the toughest climate challenges. It’s not simply about imposing rules or developing technology; it requires winning over fickle consumers who are worried about many other things in addition to the environment.

The EPA says the new vehicle regulations would protect health, the climate, and wallets

Transportation is the largest source of heat-trapping gases emitted by the US, the majority of which come from road vehicles burning gasoline and diesel. For the US to meet the climate change targets set by the Biden administration — cutting emissions in half by 2030 relative to 2005 levels — it’s essential to curb greenhouse gases from cars and trucks.

These vehicles also spew pollutants that have immediate health harms — nitrogen oxides, soot, volatile organic compounds — so reducing their output will have widespread air quality benefits as well.

The EPA this week also changed the way it compares the fuel economy of electric cars to those powered by gasoline, giving manufacturers much less credit for EVs. Electric cars obviously don’t burn gasoline, but they still have an environmental footprint depending on the emissions of the electricity used to charge them and their overall efficiency. And in order to calculate the average fuel economy for a fleet of vehicles that includes electrics, the EPA has devised a formula to figure out roughly how electrons translate into hydrocarbons.

For instance, the Ford F-150 Lightning electric pickup truck used to get 238 miles per gallon, but under the new formula, it gets 83 mpg. That means Ford will have to sell a lot more electrics or dramatically increase the fuel efficiency of its conventional cars to meet mileage requirements.

Combined, these regulations give manufacturers a strong push to not just build cleaner vehicles, but also promote them to their customers.

Prices for cleaner cars need to drop alongside emissions

The new EPA regulations don’t map out a specific route to their destination, but if carmakers choose the electrification path, the agency estimates that between 30 and 56 percent of new light-duty and 20 to 32 percent of medium-duty vehicles sold from model years 2030 to 2032 would have to be fully electric.

“Sold” is the key word here. It’s not enough for the government to set the standard and for automakers to build the car — someone has to actually buy it. And right now, the industry isn’t making enough of the cars that consumers want or can afford to purchase that line up with the new regulations.

EVs have also broken promises to drivers. Many buyers say models have failed to live up to range estimates, have quality and reliability problems, and cost more to repair than anticipated. Other owners have complained about the availability and reliability of public charging infrastructure. Some car dealers have been reluctant to keep EVs on their lots.

Some of these issues are just growing pains that come with a new technology, but changing consumer perception will be critical to keeping EV growth humming.

That’s created an opening for hybrid cars, which offer many of the same strengths as gasoline and electric vehicles in one package. For companies like Toyota that have a long history building hybrids but have been slower to produce EVs, the final regulations offer them an easier path forward. “They will tell you selling a large number of hybrids is far better than selling a moderate or small number of electric vehicles,” said Alan Baum, principal of Baum and Associates, a market research firm focused on the auto industry. That’s likely true given that close to 80 percent of car trips are less than 10 miles, so several plug-in hybrids that run fully electric, as they can for shorter distances, could displace more greenhouse gas emissions with a lower-range battery capacity than a single car that runs solely on batteries for hundreds of miles.

But it’s also important to remember that the average car in the US has been on the road for 12.5 years. Many vehicles keep rolling for two decades, or even longer. So, shifting gears to cleaner vehicles will have the greatest impact on climate change if it happens at the starting grid rather than the finish line.

20 Mar 17:11

Could a major lawsuit against realtors mean lower home prices?

by Whizy Kim
A graphic of homes tumbling down an arrow that’s pointing at the ground
A realtors’ trade group has agreed to a major settlement that could mean lower fees for home sellers. | Getty Images

What the National Association of Realtors settlement means for buyers and sellers.

Are home prices about to fall?

That’s the question many of us asked in March, after the National Association of Realtors, the trade group representing the industry, agreed to cough up $418 million as part of an antitrust lawsuit alleging that the group had artificially inflated realtor commissions that home sellers pay — which, in turn, helped inflate home prices. Now, a federal judge has approved the settlement, though a final approval will need to go through in November.

Until now, home sellers paid about 6 percent of the sale price toward a fee that would be split between their own agent and the buyer’s agent. Experts are divided on exactly how much impact this will have on home buyers, who will now likely have to start paying their agents themselves. The median sale price of homes as of late 2023 was about $417,700 — 6 percent of that amounts to a little over $25,000.

As Business Insider’s James Rodriguez noted, lower fees don’t automatically mean homes will be cheaper. In certain cases, it’s possible that sellers might list their home for the same price they would have before the settlement, and pocket more of the sale. But lower commission fees can also encourage more homeowners to list their property on the market, which could lower house prices overall.

The fact is, this real estate settlement is still too new for anyone to know for sure what the ripple effects will be. But one potential winner is tech companies in the real estate space, such as Zillow and Redfin, which have made it more feasible for people to start the home-buying process on their own instead of with a real estate agent. Vox spoke to Sonia Gilbukh, a real estate professor at City University of New York, Baruch College, to explore some of the possible outcomes.

The following conversation has been edited for length and clarity.

What was the problem with the old way realtor commissions worked? And how does this settlement change that?

It used to be that when a seller hired their agent to list a property for sale, they were paying the full commission for the transaction, which was approximately 6 percent — sometimes 5 and a half. The selling agent would then offer about half of that commission to the buyer’s side. Then the buyer’s agent will bring their clients to show all the properties, and if they end up buying the house, [the buyer’s agent] would be entitled to that commission that the seller agent was advertising for the property.

There were several rules that were part of the NAR settlement. Can you explain the new rule that sellers can’t advertise buyer agents’ commissions on the multiple listing service, or MLS, the portal that many realtors subscribe to in order to share and receive information about for-sale homes?

Yes, so the settlement is that they can no longer say, “I’m going to offer the buyer agent 3 percent,” for example, or 2.5 percent. So now, what happens is that the buyer’s agent basically would have no way to know whether they’re going to be paid for the work that they do. So something will have to change. Most likely, the buyer agents will have to directly negotiate with the buyer on the commission that they’re going to receive on a transaction.

Is it still possible that the seller’s agent would pay the buyer agent’s fee?

I think if they really wanted to, they could still post it on their website — there are ways to communicate that. But I think it would be harder to sell that as an industry standard, to the seller. Because the way it worked before is that the selling agent would say, “If you want to sell your house, we have to offer the buyer agent 3 percent, the industry standard. If we don’t, then the buyer agents are not going to show your house to their clients and you’re not going to be able to sell.” Now I feel like it would be harder to make that argument.

I’m guessing that new ways of compensating buyer agents will emerge — maybe some flat fee services, or they’ll negotiate to get paid a percentage of the deal but out of the buyer’s pocket. I don’t think they’re going to be able to keep the status quo.

I’ve been seeing in various reports that the old system, of the seller paying both agents, incentivized a practice called “steering.” Can you explain what that is, and is it really common?

Steering is a practice where the buying agent will not show, or discourage their buyers from properties that offer lower commissions.

Maisy Wong, Panle Jia Barwick, and Parag Pathak have a paper called Conflicts of Interest and Steering in Residential Brokerage, and they show that when buyer agents are offered less than the industry standard, the homes have more trouble selling. That’s basically their conclusion, that the buyer agents are steering their clients away from homes that offer lower commissions to them. I think there’s some potentially alternative explanations — if you offer less commission than the standard, maybe you’re particularly hard to deal with, difficult to negotiate with. But we certainly do see that in the data, that if you’re offering less than the standard, you were potentially jeopardizing your sale outcomes.

The plaintiffs for this lawsuit were home sellers. Beyond lower fees, what does this mean for sellers? Are there other benefits for them?

Well, we don’t know what’s going to happen, but let’s say that they’re no longer responsible for the buyer commission, then the sellers are going to be paying a 3 percent transaction cost. Now, of course, most people who sell their house also then buy a different house — so they’re still going to be paying the buyer commission on the new house that they buy.

I think what’s going to come out of this decoupling of the commission — that the buyer is going to pay for their agent, the seller’s going to pay for their agent — is that the commissions are going to become more negotiable.

And what will happen for buyers? Will some of them forgo hiring a realtor at all? Will the process of searching for a home look different?

I was talking to my mother-in-law, who is a real estate agent, and she actually owned a brokerage before. She was telling me that she views buyers to be in one of two categories: Either you’re a first-time buyer, or you’re somebody who’s selling their house and also buying something else. Those who are selling and then buying, they probably have a relationship with their agents, they probably want their agents to help them buy. So it could be a similar scenario of the status quo for them, with the possibility of maybe shaving a little bit more off the commission.

For new buyers, I think the option of paying a flat fee is going to be more attractive, because it’s going to be cheaper for them to pay a flat fee of, say, $2,000 for you to help me navigate the paperwork or something like that.

Will this mean that home prices fall?

I think eventually, if the transaction costs are going to fall, because the commissions are going to become cheaper and more negotiable. That will put a downward pressure on houses — I also think that will bring more people to sell their homes, because the transaction fee falls, people are going to be more likely to move.

I see. But you said “eventually,” so it’s not necessarily something we might see right away.

Yeah, I think it’s hard to know what’s going to happen — how buyer agents are going to be compensated, and [if] we still have buyer agents at all. We’re in this period of murky transition. For now, it’s pretty easy to sell because there’s just not a lot of inventory. But there’s not a lot of transactions actually happening.

I’m curious why we used this structure in the first place. Why have sellers typically paid both selling and buying agents?

It became the industry standard [in a period when] we had no information out there. We didn’t have Zillow. So buyer agents had a monopoly on information; if I’m not compensated as a buyer agent, or if my compensation is uncertain, then I’m going to only show [clients] the listings where I’m also the seller agent. When the commission structure changed, it improved the cooperation between agents, so they ended up showing their clients listings from other agencies. So that was actually really good.

But of course, now we have Zillow. And the potential for [buyer agents] to steer their clients only to their listings is very limited right now. There’s sort of no need for this system anymore.

Since commissions have historically been paid as a percentage of the sale, did that incentivize agents to show more expensive listings?

For the selling side, they have the incentive to sell at the highest price, essentially. But when you talk to agents, their main objective is to have the transaction happen in the first place. If they put the price too high, they risk the transaction not happening at all, then it’s not really a good trade-off. There’s also this thinking that the big houses sort of subsidize the salaries of the agents, who then also work with cheaper homes.

Some experts seem to think that this settlement will mean some real estate agents exit the industry. Do you think that’s likely? And if there are fewer realtors, is that good or bad for home buyers?

I think that’s very likely. I think most new people who come into the profession start out as buying agents, so if their compensation is going to fall, it’s not going to be worth it for them to enter anymore.

I do think it’s a good thing overall. I actually have a paper, with my co-author Paul Goldsmith-Pinkham, about the experience of real estate agents, and we find that over a quarter of all agents in the market have no experience at all. I think those are the people most likely to exit. As a result, we’re going to have more experienced real estate intermediaries, and more competitive pricing. So I do think it’s overall a good thing for consumers.

What’s the housing market like right now? Is it a seller’s market or a buyer’s market?

I think it’s still a seller’s market, but it’s sort of artificial, because we still have pretty low inventory. So yes, houses are selling quickly, but mostly because there aren’t a lot of homes for sale. Once we’re past this lock-in period — right now, most of the homes have been sold on really low mortgage rates, so it’s hard for sellers to sell and buy something new, because mortgage rates are so much higher. But eventually people will start moving, and eventually they’ll be paying off their loans. So maybe eventually the [mortgage] rates will also drop.

What else is possible in terms of reform and change in the real estate industry?

They could just straight-up outlaw sellers paying buyer commissions — but the current settlement essentially all but does that.

Are there reasons other than the long-term possibility of lower home prices for sellers and buyers to get excited about this settlement? Just how important is it?

I think it’s important. I think there’s going to be more experienced agents out there to represent buyers and sellers. I think the prices are going to drop — a little or a lot, we don’t know yet — but I think they’ll have to adjust. I think there’s going to be more people willing to move homes because the transaction cost of doing that is going to be lower.

The point you make about more homes just being on the market — that seems huge, because as you said before, one of the biggest roadblocks we’re facing is low inventory.

Yes, yeah.

I do want to say that, even though I’ve done extensive research on inexperienced agents, I do think that experienced professionals are really valuable. People should seek help, because [buying a property] is the most important transaction in their lives, probably.

Update, April 25, 9:43 am ET: This story, originally published on March 20, has been updated to reflect that a judge has approved the settlement.

20 Mar 17:11

Beeper couldn’t bring iMessage to Android — but it can still make a great chat app

by Emma Roth
An image showing the Beeper logo on a colorful background
Image: Beeper

Beeper needed a fresh start. Late last year, the company bet big on a hacked-together iMessage app for Android, only to be shut down by Apple within days. Beeper tried to fight back, but it ultimately threw in the towel just weeks later in a widely covered defeat.

“December was a wild ride,” Eric Migicovsky, the co-founder of Beeper, tells The Verge.

With dreams of an iMessage app dead, Beeper went back to square one to figure out its next steps. And the answer was to stop fighting Apple and return to its original goal: bringing every chat app — iMessage excepted — into a single place. “We’ve refocused on our primary mission,” Migicovsky says, “which is to build the best chat app on earth.”

The new Beeper app launched in beta last week,...

Continue reading…

19 Mar 22:49

Meta just showed off Threads’ fediverse integration for the very first time

by Emma Roth
An image showing the Threads logo
Illustration: The Verge

Threads is coming to the fediverse — and we just got our first official look at how that might work from Meta itself. During the FediForum conference on Tuesday, Meta’s Peter Cottle showed off a brief demo of how users will eventually be able to connect their accounts and posts to the fediverse. The integration will let users share their posts across different platforms through Threads, letting them reach multiple audiences at once. Meta is just one of the many platforms aiming to join the fediverse, a group of decentralized social networks aiming to become interoperable with one another.

As you can see in the video below, which FediForum shared with The Verge, Cottle can navigate to his Threads account settings and toggle on an option...

Continue reading…

19 Mar 20:03

Supreme Court Seems Skeptical Of The Claims That The Federal Government Coerced Social Media To Moderate

by Mike Masnick

With the standard caveat that reading too much into oral arguments is dangerous, I will note that I am cautiously optimistic after listening to the oral arguments in the Murthy v. Missouri case at the Supreme Court this morning. There are very real concerns about where the line is between government coercion (not allowed) and government persuasion (allowed) on speech, as we’ve discussed repeatedly. Sometimes this is referred to as the legality of “jawboning.”

I tend to think that government officials are all too often allowed to get away with jawboning that I think crosses the line. And this is a bipartisan problem, as government officials on both sides of the aisle seek to silence or punish speech they dislike. It would be great to have the Supreme Court set out a clear rule for how to deal with such things, and to establish where the line is.

There are some cases, such as Bantam Books at the Supreme Court, the Okwedy case at the 2nd Circuit and the Backpage case at the 7th Circuit that address some of this, but none of them clearly lay out an easily followed test. And that’s part of the reason this case has been such a mess so far. But the other, much larger reason why is that the folks bringing this case (Missouri, Louisiana, and a bunch of nonsense peddlers who were mad their content was moderated) basically filed a bunch of half-truths, lies, misleading, out of context, conspiracy theories, and more. And that has made the specific record of this case a complete mess of lies and nonsense.

Now, it’s possible that you could still get a useful ruling even with all of that. The Court obviously should make it clear that the federal government cannot seek to threaten, coerce, or intimidate social media websites into removing speech. That would be a clear First Amendment violation. But they should be able to express their opinion or to provide useful information regarding the truth or falsity of some information, or about if they’ve found evidence of (for example) foreign influence campaigns.

The problem of this particular case is that those who brought the case presented situations of the latter (providing useful information) as the former (coercing removals) and the judge at the district court inflated it even further. The judges at the 5th Circuit then took that and trimmed back some of the hyperbole by the lower court, but still (mostly) accepted as true the arguments that the sharing of information was inherently coercive. This creates some real problems, especially if the government wants to again share useful information with social media companies (which they’ve all but stopped doing because of this lawsuit).

Thankfully, the Justices (on both sides of the ideological spectrum) seemed to recognize the problematic nature of the record in this case, and the extreme claims of the states (and the lower courts). Justices Brett Kavanaugh and Sonia Sotomayor both seemed the most vocal in calling out the weird posture of the states and nonsense peddlers. Justices Amy Coney Barrett and Elena Kagan seemed similarly troubled by what the states were claiming. Justice Roberts asked good questions though it’s unclear where he is leaning exactly. Justice Ketanji Brown Jackson seemed to go too far towards arguing that the government should have more leeway in pushing companies to take down speech.

Justice Alito clearly seemed to think that the states had the stronger argument, but he may have been alone. As Louisiana’s Solicitor General (a former Alito clerk) flailed around while the other Justices battered him with hypotheticals that proved his position ridiculous, Alito stepped into try to rescue him (trying to bring him back to his arguments) and even that didn’t work very well. Justice Thomas is likely with Alito but didn’t ask very much. Gorsuch asked some odd questions, but wasn’t as clear on what he was thinking.

The key to me was seeing that Justices Kavanaugh and Barrett seemed pretty clear on the fact that of course the government should be able to express its opinion on certain content. At one point, Alito insisted that it would be crazy for anyone in the federal government to, for example, berate the press (whom he equated to social media companies) in trying to get them not to publish a story. Kavanaugh, having actually worked in the federal government (unlike Alito), pointed out that this actually happens “regularly” based on his “own experience.” He even noted that those calls often involve anger from the federal government directed at media employees:

JUSTICE KAVANAUGH: Do you think on the anger point, I guess I had assumed, thought, experienced government press people throughout the federal government who regularly call up the media and — and berate them. Is that — I mean, is that not —

MR. FLETCHER: I — I — I don’t want

JUSTICE KAVANAUGH: — your understanding? You said the anger here was unusual. I guess I wasn’t —

MR. FLETCHER: So that —

JUSTICE KAVANAUGH: — wasn’t entirely clear on that from my own experience.

Later on, he returns to this point and highlights that the federal government has plenty of perfectly good reasons to reach out to the media and warn them of harm they’re causing:

Then, on the killing people hypothetical or — not hypothetical — the statement, I mean, that raises kind of national security analogies. I don’t know what your experience is or if you’ve looked into this, but it’s probably not uncommon for government officials to protest an upcoming story on surveillance or detention policy and say, you know, if you run that, it’s going to harm the war effort and put Americans at, you know, risk.

And, from there, he cuts to the heart of the issue: just saying “doing this will kill people” is not the same as saying “and if you do so, we will punish you” and the US gov’t agrees with him on that. That is the line between persuasion and coercion:

JUSTICE KAVANAUGH: But if they tack onto that: And if you publish the story, we’re going to pursue antitrust action against you?

MR. FLETCHER: A huge problem, yeah.

Kavanaugh then repeats how the federal government regularly complains about speech in the media:

JUSTICE KAVANAUGH: You’re speaking on behalf of the United States. Again, my experience is the United States, in all its manifestations, has regular communications with the media to talk about things they don’t like or don’t want to see or are complaining about factual inaccuracies.

Justice Kagan, who also worked in the federal government, backed Kavanaugh up on this point in pushing back on Louisiana’s Solicitor General, who was making the arguments for the states/nonsense peddlers:

JUSTICE KAGAN: I mean, can I just understand because it seems like an extremely expansive argument, I must say, encouraging people basically to suppress their own speech. So, like Justice Kavanaugh, I’ve had some experience encouraging press to suppress their own speech.

You just wrote about editorial. Here are the five reasons you shouldn’t write another one. You just wrote a story that’s filled with factual errors. Here are the 10 reasons why you shouldn’t do that again.

I mean, this happens literally thousands of times a day in the federal government.

Soon after, Chief Justice Roberts also chimed in on this point. He first noted (jokingly) that unlike Kagan and Kavanaugh he doesn’t have experience trying to get the media to drop a story. But he said that just having a government official complain about speech doesn’t seem like enough to be coercion:

CHIEF JUSTICE ROBERTS: I was just going to say, first, I have no experience coercing anybody.

(Laughter.)

CHIEF JUSTICE ROBERTS: But — but, second, I mean, the government is not monolithic either. I suspect, when there’s pressure put on one of the platforms or certainly one of the other media outlets, they have people they go to, probably in the government, to say: Hey, they’re trying to get me to do this, and that person may disagree with what the government’s trying to do. It’s not monolithic. And that has to dilute the concept of coercion significantly, doesn’t it?

As for Justice Barrett, it was good to see her call out the faulty record in the lower courts and how problematic that makes the outcome.

JUSTICE BARRETT: Okay. My other question is about the findings of fact and clear error. So you were pretty insistent with Justice Kagan that we really, to address the standing point, don’t have to review any of the district court’s factual findings for clear error.

I just want to make sure that that’s right because I’m thinking about things you talked about with — I think it was Justice Alito, the interchange with the expletives, you know, we’re getting mad, we want answers now, you know, are you, whatever, serious?

MR. FLETCHER: Yeah.

JUSTICE BARRETT: And that was actually about his own Facebook account. Or there was another change that was — exchange that was actually about somebody impersonating the President’s granddaughter on Twitter.

MR. FLETCHER: Yeah.

JUSTICE BARRETT: So, if the lower courts, which I think they did, kind of conflated some of those threats with threats that were designed to be — threats related to the pandemic and that kind of suppression, wouldn’t that then be clear error, or do you think that’s application of facts to law or what?

Later on, during the states’ arguments, there was a lot of pushback on Louisiana’s Solicitor General. I noted some of it above in the exchanges with Roberts, but Kagan went after him for suggesting that the government shouldn’t complain about certain speech. She uses the example of terrorist speech, which is mostly protected speech, though Louisiana’s Solicitor General doesn’t seem to realize that kind of important point:

JUSTICE KAGAN: So, I mean, what about that? I mean, you know, take a — an example where — I mean, these platforms, they’re compilers of speech, and some part of the government, let’s call it part of the law enforcement arm of the government, says you might not realize it, but you are hosting a lot of terrorist speech, which is going to increase the chances that there’s going to be some terrible harm that’s going to take place, and we want to give you this information, we want to try to persuade you to take it down.

Are — are — the government can’t do that?

MR. AGUINAGA: The government can absolutely do that, Justice Kagan.

JUSTICE KAGAN: They’re taking —

MR. AGUINAGA: Terrorist activity, criminal —

JUSTICE KAGAN: — they’re — they’re asking them to take down the speech.

MR. AGUINAGA: Terrorist activity, criminal activity, that is not protected speech. Absolutely, the government can inform the — the

JUSTICE KAGAN: Well, that might — might be protected speech. I mean, terrorists engage in, you know, things that come under the First Amendment. I mean, let’s say they’re just recruiting people for their organizations.

MR. AGUINAGA: Your Honor, if it’s First Amendment speech, protected speech, then I think we’re in an entirely different world.

After some more back and forth on this, Kagan again points out how frequently this happens:

JUSTICE KAGAN: So back in — this — this — this still happens now — decades ago, it happened all the time, which is somebody from the White House got in touch with somebody from The Washington Post and said this will — this will just harm national security, and The Washington Post said, okay, whatever you say.

I mean, that was all — we didn’t know enough, but that was — that was coercion?

Eventually, she sorta gets him to admit that that’s okay, which sort of left him trapped as the standard he presented earlier wouldn’t allow that:

JUSTICE KAGAN: I guess what I’m just trying to suggest is that there’s all kinds of things that can appear on these platforms that do all kinds of different harms, and — and the inability of government that you’re suggesting to — to reach out to these platforms and say we want to give you information that you might not know about on this, and we want to give you our perspective on what harms this is doing, and — and, you know, we want to be able to answer questions that you have because we really do think that it would be a good thing if you on your own chose to take this speech down.

MR. AGUINAGA: And, Your Honor, if those were the facts in this case, then I think it would be a much harder case for me. I think

JUSTICE KAGAN: Well, now I don’t know what your standard is. You just told me that that was — that was good enough for you.

MR. AGUINAGA: No —

JUSTICE KAGAN: That was coercion.

And as he tried to dig himself out of the hole by arguing that there was clear evidence of actual coercion in this case, Justice Sotomayor came out swinging about how there is nothing really in the record to support any of that (as some of us have been noting). She asks him to support these claims, and as he brings up a few examples, she pushes back on each of them, explaining how they don’t actually show what he pretends they show. And then:

JUSTICE SOTOMAYOR: You know, I — I have such a problem with — with your brief, counselor. You omit information that changes the context of some of your claims. You attribute things to people who it didn’t happen to. At least in one of the defendants, it was her brother that something happened to, not her. I don’t know what to make of all this because you’re — you have a — I’m not sure how we get to prove direct injury in any way.

There were some other situations where the problematic nature of the record got called out. There was a lot of talk of “traceability.” That is, is there any actual evidence that social media companies acted in a specific way due to government pressure. And the examples given didn’t fare well.

MR. AGUINAGA: Your Honor, I think the clearest way — and if I understand — so let me answer your question directly, Your Honor. The way — the link that I was drawing there was a temporal one. If you look at JA 715 to 717, that’s a May 2021 e-mail. Two months later after that e-mail, calls were targeting health groups just like Jill Hines’s group. She experiences the first example of that kind of group being —

JUSTICE KAGAN: Yeah, so in two months, I mean, a lot of things can happen in two months. So that decision two months later could have been caused by the government’s e-mail, or that government e-mail might have been long since forgotten, because, you know, there are a thousand other communications that platform employees have had with each other, that — a thousand other things that platform employees have read in the newspaper.

I mean, why would we point to one e-mail two months earlier and say it was that e-mail that made all the difference?

MR. AGUINAGA: Your Honor — and I would say a thousand other e-mails between the White House and Facebook in those two months. I mean, that’s the volume of this interaction, this back and forth, between the platform and the government. And it’s all —

JUSTICE KAGAN: The specific —

MR. AGUINAGA: — about the same topic.

JUSTICE KAGAN: But if it’s encouragement — I mean, let’s even take that this was something that the — that the government was continually pressing the — encouraging the platforms to do. I mean, until you can show that there’s something about — overbearing the platform’s will, which, you know, seems sort of hard to overbear Facebook’s work — will from what I can gather from the world, but, you know, how do you say it’s the government rather than Facebook?

Eventually, Kagan points out that nothing in the record supports actual traceability:

JUSTICE KAGAN: And you could say that about pretty much everything that’s in your brief, that there’s just nothing where you can say, okay, the government said take down that communication. The government is making some broad statements about the kinds of communications it thinks harmful. Facebook has a lot of opinions on its own about various kinds of communications it thinks harmful.

I guess if you’re going to use standard ideas about traceability and redressability, I guess what I’m suggesting is I don’t see a single item in your briefs that would satisfy our normal tests.

The US’s representative points out that the record is even worse than Louisiana suggests:

Justice Kagan, you pointed out that her groups were moderated at least two and four months after the relevant exchange between Facebook and the government. But it’s actually worse than that. The May 2021 e-mail from Facebook to the government says we’ve already taken action on health groups to remove them from our recommendation feature. It wasn’t reporting on something it would do in the future. It was reporting on something that was already done. And it’s even not clear from the e-mail that Facebook was doing that because of any request from the government. It was a report of its own action.

Justice Barrett pointed to a hypothetical. She asked if Louisiana’s Solicitor General would feel that, if his own info had been doxed (and yes, she used that term), and the FBI reached out to social media to tell them about it, would he still feel that’s a First Amendment violation:

JUSTICE BARRETT: Well, let me just — let me just ask you then, let me give you a hypothetical. Let’s say that you get doxed and so do numerous other members in Louisiana state government. You’re doxed, and somebody is posting online about how people should really rally and do something about this. People should rally and you should be harmed, okay?

The FBI sees these posts and calls the social media outlet, like X, Facebook, whatever, and says we really encourage you to take these down because these are significantly threatening and we see some people may be responding to them.

That’s — that’s a problem?

That leaves him in a deep hole that he has trouble getting out of:

MR. AGUINAGA: So my first question, Your Honor, is whether that would be protected speech, those tweets would be protected speech, Your Honor, under this Court’s —

JUSTICE BARRETT: Okay. Let’s just assume — let’s assume that everything that’s said, I was trying to make it so that they —

MR. AGUINAGA: Yes, they are.

JUSTICE BARRETT: — stop short of actually being illegal in and of themselves.

MR. AGUINAGA: Your Honor, so I think, you know, as I say, I’m a purist on the First Amendment, so my answer would be yeah, like, that —

JUSTICE BARRETT: So the FBI can’t make — do you know how often the FBI makes those kinds of calls?

To be honest, I don’t think the FBI actually makes those calls very often. I would imagine, however, that the U.S. Marshals does make those calls pretty regularly about info on judges (ask me how I know), but it is making the same basic point. There are times that the federal government and law enforcement may have pretty good reasons for reaching out to social media and pointing out potential very real harms that may occur if information is left up, so long as the platforms still get to make the final call on whether or not to leave it up.

Indeed, the fact that the platforms regularly kept up information the government informed them about was brought up repeatedly. Many of the Justices remarked on it as well, including Kavanaugh, and how it argued against believing that the speech was coercive.

Justice Kavanaugh also had trouble with the argument that the government can’t communicate potentially problematic info to social media companies:

JUSTICE KAVANAUGH: And one thing that I think I want to square up with you is if someone calls and — or contacts the social media company and says what you have there, this post, has factually erroneous information, so not a viewpoint that we disagree with, factually erroneous information, and the social media company says, we’ll take a look at that and — and you still think that’s significant encouragement that qualifies as coercion, if they take it down in response to concluding that it, in fact, is factually erroneous?

MR. AGUINAGA: No, Your Honor. If there’s no ask from the government, if the government’s just saying here’s our view of the statement —

JUSTICE KAVANAUGH: Okay. And we think it should be — it should be taken down, it’s up to you, but we think it should be taken down.

MR. AGUINAGA: I think that’s a harder case for me. I guess, you know, if you think it is a close case decide it under the First Amendment.

Kavanaugh asks why that’s harder, and Aguinaga suggests that it’s because of the specifics of the situation and denies (as some commenters here regularly argue) that any communication from the government is inherently threatening. So Kavanaugh asks him about more specifics, including if someone is publishing information on US troop movements. And all Aguinaga can come up with is “well, the government can respond.”

Perhaps the strangest, and arguably the most troubling, comments came from Justice Jackson, who seemed to think that there were situations in which the federal government should be able to coerce sites to take down speech. She started out on the right foot, worrying how the states’ position could be seen as barring the ability of the government to provider useful info:

JUSTICE JACKSON: I mean, I’m a little worried about the Respondents’ — what I think could be taken away from their view, which is that in situations in which the government has information that may be unique to the government’s knowledge but that it feels important for the public to have, that that somehow becomes prohibited if, as a result of that information, these companies decide they’re going to do something different with respect to content moderation.

But then she goes much further at times:

JUSTICE JACKSON: So I think — I think that part of the reason why you might be running into all of these difficulties with respect to the different factual circumstances is because you’re not focusing on the fact that there are times in which the government can, depending on the circumstances, encourage, perhaps even coerce, because they have a compelling interest in doing so.

And so that’s why I keep coming back to the actual underlying First Amendment issue, which we can isolate in this case and just talk about — about coercion, but I think there — that you have to admit that there are certain circumstances in which the government can provide information, encourage the platforms to take it down, tell them to take it down.

I mean, what about — what about the hypo of someone posting classified information? They say it’s my free speech right, I believe that I — you know, I got access to this information and I want to post it.

Are you suggesting that the government couldn’t say to the platforms, we need to take that down?

I mean… isn’t that just the Pentagon Papers story? And in that case the court said of course the newspapers could publish that info.

I think she’s trying to say that even if the First Amendment applies, there may be cases where under strict scrutiny, the government still has the right to pressure companies to remove protected speech, but that seems to be a pretty far reach and not even an issue in this particular case.

To that point, though, she is using this point to argue that the arguments of the states/nonsense peddlers that the federal government can never even encourage removals clearly goes too far:

But you just seemed to suggest that as a blanket matter, the government doesn’t have the ability to, you know, encourage or require this kind of censorship. And I don’t know that that’s the case.

She goes on to talk about a “teen challenge” that puts kids in danger, and whether or not the government can pressure sites to take that down. Again, I think she’s going way too far here. I think the gov’t can raise the issue, highlight how such things can lead to harm, but that they cannot coerce sites to take such info down. They need to decide based on their own policies:

JUSTICE JACKSON: Suppose someone started posting about a new teen challenge that involved teens jumping out of windows at increasing elevations. This is the challenge. And kids all over the country start doing this. There’s an epidemic, children are seriously injuring or even killing themselves in situations. Is it your view that the government authorities could not declare those circumstances a public emergency and encourage social media platforms to take down the information that is instigating this problem?

Even Roberts seems a bit surprised at how far Louisiana’s SG is willing to go on this argument:

CHIEF JUSTICE ROBERTS: Well, you don’t think — well, do you think that simply Justice Jackson’s hypotheticals ended by saying we encourage you to take it down, is that rise to the level of coercion that you think is problematic?

MR. AGUINAGA: Your Honor, if the test is coercion and that’s the test that this Court applies, I think I might have a harder case saying that’s coercion. I think it’s — by its definition, it’s maybe easier addressed as a substantial encouragement case.

But if — you know, whether — as I said earlier, regardless of the label that you apply, whether it’s coercion, whether it’s encouragement, or joint participation and conspiracy, at the end of the day, if what the government is trying to do is to eliminate viewpoints from public discourse, that I think

CHIEF JUSTICE ROBERTS: Well, again, under my colleague’s hypothetical, it was not necessarily eliminate viewpoints, it was to eliminate instructions, let’s say, about how to engage in some game that is seriously harming children around — around the country, and they say we — we encourage you to stop that.

I mean, is it — that violates the Constitution?

MR. AGUINAGA: Your Honor, I agree as a policy matter, it might be great for the government to be able to do that, but the moment that the government identifies an entire category of content that it wishes to not be in the modern public sphere, that is a First Amendment problem.

There’s more in the full transcript, but it sure sounds like this could be a 6-3 ruling with Kavanaugh, Barrett, Roberts, Sotomayor, Kagan, and Jackson all somehow overturning the 5th Circuit. We’ll find out where it ends up in a few months. But, the hearing didn’t go nearly as far off the rails as I feared it might.

19 Mar 19:58

Google DeepMind co-founder joins Microsoft as CEO of its new AI division

by Emma Roth
A photo showing Mustafa Suleyman during the World Economic Forum 2024
Image: Getty

Microsoft has hired Google DeepMind co-founder Mustafa Suleyman. In a post on X, Suleyman announced that he’s joining Microsoft as the CEO of a new team that handles the company’s consumer-facing AI products, including Copilot, Bing, and Edge.

Suleyman will also serve as executive vice president of Microsoft AI and join the company’s senior leadership team that reports directly to CEO Satya Nadella. Suleyman co-founded the AI lab DeepMind in 2010, which was later acquired by Google in 2014.

DeepMind has remained a pioneering AI force within Google. However, Suleyman hasn’t been part of the division in many years. He was placed on leave in 2019 over controversy surrounding some of the projects he led, according to a report from Bloomberg....

Continue reading…