verge-rss

Image: The Verge

OpenAI is announcing its much-anticipated entry into the search market, SearchGPT, an AI-powered search engine with real-time access to information across the internet.
The search engine starts with a large textbox that asks the user “What are you looking for?” But rather than returning a plain list of links, SearchGPT tries to organize and make sense of them. In one example from OpenAI, the search engine summarizes its findings on music festivals and then presents short descriptions of the events followed by an attribution link.
In another example, it explains when to plant tomatoes before breaking down different varieties of the plant. After the results appear, you can ask follow-up questions or click the sidebar to open other relevant links. There’s also a feature called “visual answers,” but OpenAI didn’t get back to The Verge before publication on exactly how this works.

Image: OpenAI
An example of SearchGPT query.

SearchGPT is just a “prototype” for now. The service is powered by the GPT-4 family of models and will only be accessible to 10,000 test users at launch, OpenAI spokesperson Kayla Wood tells The Verge. Wood says that OpenAI is working with third-party partners and using direct content feeds to build its search results. The goal is to eventually integrate the search features directly into ChatGPT.
It’s the start of what could become a meaningful threat to Google, which has rushed to bake in AI features across its search engine, fearing that users will flock to competing products that offer the tools first. It also puts OpenAI in more direct competition with the startup Perplexity, which bills itself as an AI “answer” engine. Perplexity has recently come under criticism for an AI summaries feature that publishers claimed was directly ripping off their work.

Image: OpenAI
SearchGPT’s “visual answers” feature showcases an AI-generated video from OpenAI’s Sora, via YouTube.

OpenAI seems to have taken note of the blowback and says it’s taking a markedly different approach. In a blog post, the company emphasized that SearchGPT was developed in collaboration with various news partners, which include organizations like the owners of The Wall Street Journal, The Associated Press, and Vox Media, the parent company of The Verge. “News partners gave valuable feedback, and we continue to seek their input,” Wood says.
Publishers will have a way to “manage how they appear in OpenAI search features,” the company writes. They can opt out of having their content used to train OpenAI’s models and still be surfaced in search.
“Responses have clear, in-line, named attribution and links
“SearchGPT is designed to help users connect with publishers by prominently citing and linking to them in searches,” according to OpenAI’s blog post. “Responses have clear, in-line, named attribution and links so users know where information is coming from and can quickly engage with even more results in a sidebar with source links.”
Releasing its search engine as a prototype helps OpenAI in a few different ways. First, if SearchGPT’s results are wildly incorrect — like when Google rolled out AI Overviews and told us to put glue on our pizza — it’s easier to say, well, it’s a prototype! There’s also potential for getting attributions wrong or maybe wholesale ripping off articles like Perplexity was accused of doing.
This new product has been whispered about for months now, with The Information reporting about its development in February, then Bloomberg reporting more in May. We reported at the same time that OpenAI had been aggressively trying to poach Google employees for a search team. Some X users also noticed a new website OpenAI has been working on that hinted toward the move.
OpenAI has slowly been bringing ChatGPT more in touch with the real-time web. When GPT-3.5 was released, the AI model was already months out of date. Last September, OpenAI released a way for ChatGPT to browse the internet, called Browse with Bing, but it appears a lot more rudimentary than SearchGPT.
The rapid advancements by OpenAI have won ChatGPT millions of users, but the company’s costs are adding up. The Information reported this week that OpenAI’s AI training and inference costs could reach $7 billion this year, with the millions of users on the free version of ChatGPT only further driving up compute costs. SearchGPT will be free during its initial launch, and since the feature appears to have no ads right now, it’s clear the company will have to figure out monetization soon.

Image: The Verge

OpenAI is announcing its much-anticipated entry into the search market, SearchGPT, an AI-powered search engine with real-time access to information across the internet.

The search engine starts with a large textbox that asks the user “What are you looking for?” But rather than returning a plain list of links, SearchGPT tries to organize and make sense of them. In one example from OpenAI, the search engine summarizes its findings on music festivals and then presents short descriptions of the events followed by an attribution link.

In another example, it explains when to plant tomatoes before breaking down different varieties of the plant. After the results appear, you can ask follow-up questions or click the sidebar to open other relevant links. There’s also a feature called “visual answers,” but OpenAI didn’t get back to The Verge before publication on exactly how this works.

Image: OpenAI
An example of SearchGPT query.

SearchGPT is just a “prototype” for now. The service is powered by the GPT-4 family of models and will only be accessible to 10,000 test users at launch, OpenAI spokesperson Kayla Wood tells The Verge. Wood says that OpenAI is working with third-party partners and using direct content feeds to build its search results. The goal is to eventually integrate the search features directly into ChatGPT.

It’s the start of what could become a meaningful threat to Google, which has rushed to bake in AI features across its search engine, fearing that users will flock to competing products that offer the tools first. It also puts OpenAI in more direct competition with the startup Perplexity, which bills itself as an AI “answer” engine. Perplexity has recently come under criticism for an AI summaries feature that publishers claimed was directly ripping off their work.

Image: OpenAI
SearchGPT’s “visual answers” feature showcases an AI-generated video from OpenAI’s Sora, via YouTube.

OpenAI seems to have taken note of the blowback and says it’s taking a markedly different approach. In a blog post, the company emphasized that SearchGPT was developed in collaboration with various news partners, which include organizations like the owners of The Wall Street Journal, The Associated Press, and Vox Media, the parent company of The Verge. “News partners gave valuable feedback, and we continue to seek their input,” Wood says.

Publishers will have a way to “manage how they appear in OpenAI search features,” the company writes. They can opt out of having their content used to train OpenAI’s models and still be surfaced in search.

“SearchGPT is designed to help users connect with publishers by prominently citing and linking to them in searches,” according to OpenAI’s blog post. “Responses have clear, in-line, named attribution and links so users know where information is coming from and can quickly engage with even more results in a sidebar with source links.”

Releasing its search engine as a prototype helps OpenAI in a few different ways. First, if SearchGPT’s results are wildly incorrect — like when Google rolled out AI Overviews and told us to put glue on our pizza — it’s easier to say, well, it’s a prototype! There’s also potential for getting attributions wrong or maybe wholesale ripping off articles like Perplexity was accused of doing.

This new product has been whispered about for months now, with The Information reporting about its development in February, then Bloomberg reporting more in May. We reported at the same time that OpenAI had been aggressively trying to poach Google employees for a search team. Some X users also noticed a new website OpenAI has been working on that hinted toward the move.

OpenAI has slowly been bringing ChatGPT more in touch with the real-time web. When GPT-3.5 was released, the AI model was already months out of date. Last September, OpenAI released a way for ChatGPT to browse the internet, called Browse with Bing, but it appears a lot more rudimentary than SearchGPT.

The rapid advancements by OpenAI have won ChatGPT millions of users, but the company’s costs are adding up. The Information reported this week that OpenAI’s AI training and inference costs could reach $7 billion this year, with the millions of users on the free version of ChatGPT only further driving up compute costs. SearchGPT will be free during its initial launch, and since the feature appears to have no ads right now, it’s clear the company will have to figure out monetization soon.

Read More 

Illustration: The Verge

Meta CEO Mark Zuckerberg announced today on his WhatsApp channel that WhatsApp now has 100 million daily active users in the US as of July 25th.
According to WhatsApp, Los Angeles, New York, Miami, and Seattle are among the fastest-growing markets in the country. However, the company says it’s experienced “significant growth” in Southern states as well, with more than 10 million daily users in Texas.
“This is the first time data pertaining to users has been released since Meta’s acquisition of the messaging app and demonstrates how the messaging platform, already the biggest messaging app in the world, has experienced tremendous growth in the U.S.,” WhatsApp spokesperson Ashley O’Reilly wrote in a press release.
Since purchasing WhatsApp for $16 billion in 2014, Meta has put a lot of effort into making the app as popular in the US as it is internationally. One such way is by advertising WhatsApp’s availability on iOS and Android, particularly how easy it is to message users across platforms. (Because iPhones don’t need to switch from iMessage to SMS, messages from Androids don’t show up as green bubble with lower-quality images, removed encryption, and shortened messages.) Recently, WhatsApp captured attention when it reunited the cast of Modern Family for an ad promoting this very feature to families.

Yet, while WhatsApp now boasts more than 2 billion users in over 180 countries, it’s still less popular in the US than rivals like Apple’s iMessage. While it might have experienced growth in the US, it’s going to be even harder for WhatsApp to compete once Apple replaces Rich Communication Services (RCS) as the communication protocol between Android and iOS devices. RCS support will be coming to iOS 18 this fall, which means iPhone users won’t necessarily need to switch group chats over to WhatsApp — and Meta really needs to start rethinking its marketing strategy.

Illustration: The Verge

Meta CEO Mark Zuckerberg announced today on his WhatsApp channel that WhatsApp now has 100 million daily active users in the US as of July 25th.

According to WhatsApp, Los Angeles, New York, Miami, and Seattle are among the fastest-growing markets in the country. However, the company says it’s experienced “significant growth” in Southern states as well, with more than 10 million daily users in Texas.

“This is the first time data pertaining to users has been released since Meta’s acquisition of the messaging app and demonstrates how the messaging platform, already the biggest messaging app in the world, has experienced tremendous growth in the U.S.,” WhatsApp spokesperson Ashley O’Reilly wrote in a press release.

Since purchasing WhatsApp for $16 billion in 2014, Meta has put a lot of effort into making the app as popular in the US as it is internationally. One such way is by advertising WhatsApp’s availability on iOS and Android, particularly how easy it is to message users across platforms. (Because iPhones don’t need to switch from iMessage to SMS, messages from Androids don’t show up as green bubble with lower-quality images, removed encryption, and shortened messages.) Recently, WhatsApp captured attention when it reunited the cast of Modern Family for an ad promoting this very feature to families.

Yet, while WhatsApp now boasts more than 2 billion users in over 180 countries, it’s still less popular in the US than rivals like Apple’s iMessage. While it might have experienced growth in the US, it’s going to be even harder for WhatsApp to compete once Apple replaces Rich Communication Services (RCS) as the communication protocol between Android and iOS devices. RCS support will be coming to iOS 18 this fall, which means iPhone users won’t necessarily need to switch group chats over to WhatsApp — and Meta really needs to start rethinking its marketing strategy.

Read More 

Image: Cath Virginia / The Verge

The push to childproof the internet is reaching a tipping point. The Kids Online Safety Act (KOSA) was first introduced in 2023; over a year later, with the August recess looming, the Senate is poised to vote on the bill. With 70 cosponsors, the bill appears likely to pass.
Proponents of the bill believe that the law is necessary to safeguard children from harm that could result from the platforms’ relentless quest for user attention. Critics argue that KOSA not only erodes internet freedoms but could also prevent minors — particularly LGBTQ minors — from accessing potentially lifesaving information.
Indeed, lead cosponsor Sen. Marsha Blackburn (R-TN) has publicly justified KOSA on the basis that “we should be protecting minor children from the transgender in this culture.”
The current text of KOSA creates a duty of care for platforms, requiring them to take reasonable steps to mitigate a specific list of harms to minors. Those include things like cyberbullying, sexual exploitation, and eating disorders.
On top of that, it would mandate certain kinds of parental tools, require the highest level of privacy settings to be on for kids by default, and let young users have a say in whether they get personalized recommendations, like through algorithmic feeds.
Because the House of Representatives is adjourning ahead of schedule next week, the fate of KOSA is very much up in the air.
Here’s all the news about KOSA:

Image: Cath Virginia / The Verge

The push to childproof the internet is reaching a tipping point.

The Kids Online Safety Act (KOSA) was first introduced in 2023; over a year later, with the August recess looming, the Senate is poised to vote on the bill. With 70 cosponsors, the bill appears likely to pass.

Proponents of the bill believe that the law is necessary to safeguard children from harm that could result from the platforms’ relentless quest for user attention. Critics argue that KOSA not only erodes internet freedoms but could also prevent minors — particularly LGBTQ minors — from accessing potentially lifesaving information.

Indeed, lead cosponsor Sen. Marsha Blackburn (R-TN) has publicly justified KOSA on the basis that “we should be protecting minor children from the transgender in this culture.”

The current text of KOSA creates a duty of care for platforms, requiring them to take reasonable steps to mitigate a specific list of harms to minors. Those include things like cyberbullying, sexual exploitation, and eating disorders.

On top of that, it would mandate certain kinds of parental tools, require the highest level of privacy settings to be on for kids by default, and let young users have a say in whether they get personalized recommendations, like through algorithmic feeds.

Because the House of Representatives is adjourning ahead of schedule next week, the fate of KOSA is very much up in the air.

Here’s all the news about KOSA:

Read More 

Image: Hugo Herrera / The Verge

Global temperatures are shattering records this week. And the United Nations is making an urgent “call to action” on extreme heat that’s killing workers and putting more people at risk with climate change.
The first three days of this week have been the hottest on record for the planet, according to preliminary data from the European Union’s Copernicus Climate Change Service. And it’s now looking likely that 2024 could beat last year to become the hottest year on record. Against that backdrop, United Nations (UN) agencies released a new global analysis of heat stress on workers today, along with a roadmap for how to protect the most vulnerable.
“Let’s face facts: extreme temperatures are no longer a one-day, one-week, or one-month phenomenon,” said UN Secretary-General António Guterres.
“Extreme temperatures are no longer a one-day, one-week, or one-month phenomenon.”
Sunday likely set a record for the highest global average temperature recorded since at least 1940, the start of Copernicus’ data set. That record fell by Monday when the global average temperature reached a new high of 17.15 degrees Celsius (62.87 degrees Fahrenheit). The following day came very close to being as hot, according to preliminary data. That means that July 21st through 23rd have probably been the three hottest days on record — at least so far. Heat-trapping greenhouse gas emissions from fossil fuels are raising global average temperatures and making heatwaves more frequent and more intense.
A tourist suffered third-degree burns on his feet and had to be rescued after either losing or breaking his flip-flops in California’s Death Valley, where temperatures reached around 50.55 degrees Celsius (123 degrees Fahrenheit). Heat has made travel more treacherous in other parts of the world. At least 10 tourists from other countries were found dead or missing along hiking trails in Greece last month. At least 1,300 people died amid scorching temperatures in Saudi Arabia this June during the annual pilgrimage to Mecca.
Workers are routinely at risk from heat stress around the world, a report released today by the International Labour Organization (ILO) shows. Heat leads to nearly 23 million injuries and 18,970 deaths globally among workers each year, according to the ILO. More than 70 percent of the global workforce is vulnerable to extreme heat on the job, it says, particularly those who work outdoors or indoors without proper ventilation or cooling.
A separate report released today by the secretary-general and 10 different UN agencies issues a “call to action on extreme heat.” It lays out policy recommendations for governments, including stronger protections for workers. “Sensible occupational safety and health measures” would save economies $361 billion a year, according to the ILO. Labor productivity drops by 50 percent when daily temperatures surpass 34 degrees Celsius (93.2 degrees Fahrenheit).

Cities and buildings can also be designed to stay cooler and trap less heat. Early warning systems for heat (similar to those for storms) can also save lives, the UN says. Heat kills more people around the world than tropical cyclones, but those deaths can be prevented by helping people find a safe place to cool down. More than 98,300 heat deaths could be prevented each year with adequate warning systems in place in just 57 countries, according to the report.
Addressing the root cause of the problem, however, depends on transitioning away from the fossil fuels causing global warming. “Countries must phase out fossil fuels — fast and fairly,” Guterres said today. “The world must rise to the challenge of rising temperatures.”

Image: Hugo Herrera / The Verge

Global temperatures are shattering records this week. And the United Nations is making an urgent “call to action” on extreme heat that’s killing workers and putting more people at risk with climate change.

The first three days of this week have been the hottest on record for the planet, according to preliminary data from the European Union’s Copernicus Climate Change Service. And it’s now looking likely that 2024 could beat last year to become the hottest year on record. Against that backdrop, United Nations (UN) agencies released a new global analysis of heat stress on workers today, along with a roadmap for how to protect the most vulnerable.

“Let’s face facts: extreme temperatures are no longer a one-day, one-week, or one-month phenomenon,” said UN Secretary-General António Guterres.

Sunday likely set a record for the highest global average temperature recorded since at least 1940, the start of Copernicus’ data set. That record fell by Monday when the global average temperature reached a new high of 17.15 degrees Celsius (62.87 degrees Fahrenheit). The following day came very close to being as hot, according to preliminary data. That means that July 21st through 23rd have probably been the three hottest days on record — at least so far. Heat-trapping greenhouse gas emissions from fossil fuels are raising global average temperatures and making heatwaves more frequent and more intense.

A tourist suffered third-degree burns on his feet and had to be rescued after either losing or breaking his flip-flops in California’s Death Valley, where temperatures reached around 50.55 degrees Celsius (123 degrees Fahrenheit). Heat has made travel more treacherous in other parts of the world. At least 10 tourists from other countries were found dead or missing along hiking trails in Greece last month. At least 1,300 people died amid scorching temperatures in Saudi Arabia this June during the annual pilgrimage to Mecca.

Workers are routinely at risk from heat stress around the world, a report released today by the International Labour Organization (ILO) shows. Heat leads to nearly 23 million injuries and 18,970 deaths globally among workers each year, according to the ILO. More than 70 percent of the global workforce is vulnerable to extreme heat on the job, it says, particularly those who work outdoors or indoors without proper ventilation or cooling.

A separate report released today by the secretary-general and 10 different UN agencies issues a “call to action on extreme heat.” It lays out policy recommendations for governments, including stronger protections for workers. “Sensible occupational safety and health measures” would save economies $361 billion a year, according to the ILO. Labor productivity drops by 50 percent when daily temperatures surpass 34 degrees Celsius (93.2 degrees Fahrenheit).

Cities and buildings can also be designed to stay cooler and trap less heat. Early warning systems for heat (similar to those for storms) can also save lives, the UN says. Heat kills more people around the world than tropical cyclones, but those deaths can be prevented by helping people find a safe place to cool down. More than 98,300 heat deaths could be prevented each year with adequate warning systems in place in just 57 countries, according to the report.

Addressing the root cause of the problem, however, depends on transitioning away from the fossil fuels causing global warming. “Countries must phase out fossil fuels — fast and fairly,” Guterres said today. “The world must rise to the challenge of rising temperatures.”

Read More 

Photo by Alex Wong / Getty Images

The Associated Press has apparently retracted a fact-check published yesterday with the headline, “No, JD Vance did not have sex with a couch.” As of Thursday morning, the article page displayed a “page unavailable” error message.
The fact-check references a few joke social media posts that have been circulating that claim the Republican vice presidential nominee wrote in his book Hillbilly Elegy about having sex with a couch.
One tweet reads, “In his dreadful novel, ‘Hillbilly Elegy,’ JD Vance described having sex with a rubber glove secured between cushions on his couch. Republicans chose him to be one heartbeat away from becoming POTUS. Voters in NC, the U.S. furniture capital, should be particularly horrified.” (Fact-check: Hillbilly Elegy is a memoir, not a novel.)
Another post: “Can’t say for sure but he might be the first VP pick to have admitted in a ny times bestseller to fucking an inside-out latex glove shoved between two couch cushions (vance, hillbilly elegy, pp. 179-181).”
To be clear, Vance did not write in Hillbilly Elegy about having sexual relations with living room furniture, and a Snopes debunk remains live about the viral joke. The archived version of AP’s fact-check also notes that there’s no such passage in the PDF of the book. But there’s also technically no proof that Vance didn’t have sex with a couch — there’s no way a journalist could truly know that. He just didn’t write about it.

Oh my God pic.twitter.com/G8beJlERn4— UAE Exotic Falconry & Finance (@FalconryFinance) July 24, 2024

“The story, which did not go out on the wire to our customers, didn’t go through our standard editing process. We are looking into how that happened,” AP spokesperson Nicole Meir told The Verge in an email.
News reports (and fact-checks specifically) are often worded in a way that carefully threads a needle — there’s a difference between saying something definitively didn’t happen versus saying there’s no evidence of it. My guess is that the AP headline was the problem here because it claims to debunk something that is unknowable. A headline like, “No, JD Vance didn’t write about fucking a couch” perhaps would have been more accurate.
And when you really peel back the layers (turn over the cushions?), there’s really no way for any of us to prove we didn’t have sex with a couch.
Correction, July 25th: A previous version of the piece misstated that the AP article was removed as of Friday morning. It was Thursday morning.

Photo by Alex Wong / Getty Images

The Associated Press has apparently retracted a fact-check published yesterday with the headline, “No, JD Vance did not have sex with a couch.” As of Thursday morning, the article page displayed a “page unavailable” error message.

The fact-check references a few joke social media posts that have been circulating that claim the Republican vice presidential nominee wrote in his book Hillbilly Elegy about having sex with a couch.

One tweet reads, “In his dreadful novel, ‘Hillbilly Elegy,’ JD Vance described having sex with a rubber glove secured between cushions on his couch. Republicans chose him to be one heartbeat away from becoming POTUS. Voters in NC, the U.S. furniture capital, should be particularly horrified.” (Fact-check: Hillbilly Elegy is a memoir, not a novel.)

Another post: “Can’t say for sure but he might be the first VP pick to have admitted in a ny times bestseller to fucking an inside-out latex glove shoved between two couch cushions (vance, hillbilly elegy, pp. 179-181).”

To be clear, Vance did not write in Hillbilly Elegy about having sexual relations with living room furniture, and a Snopes debunk remains live about the viral joke. The archived version of AP’s fact-check also notes that there’s no such passage in the PDF of the book. But there’s also technically no proof that Vance didn’t have sex with a couch — there’s no way a journalist could truly know that. He just didn’t write about it.

“The story, which did not go out on the wire to our customers, didn’t go through our standard editing process. We are looking into how that happened,” AP spokesperson Nicole Meir told The Verge in an email.

News reports (and fact-checks specifically) are often worded in a way that carefully threads a needle — there’s a difference between saying something definitively didn’t happen versus saying there’s no evidence of it. My guess is that the AP headline was the problem here because it claims to debunk something that is unknowable. A headline like, “No, JD Vance didn’t write about fucking a couch” perhaps would have been more accurate.

And when you really peel back the layers (turn over the cushions?), there’s really no way for any of us to prove we didn’t have sex with a couch.

Correction, July 25th: A previous version of the piece misstated that the AP article was removed as of Friday morning. It was Thursday morning.

Read More 

Illustration: The Verge

Google’s Gemini AI chatbot will be able to respond to you more quickly and process more content in prompts thanks to an upgrade to the company’s Gemini 1.5 Flash AI model.
With the upgrade to 1.5 Flash, you’ll see “across-the-board improvements in quality and latency, with especially noticeable improvements in reasoning and image understanding,” Amar Subramanya, VP of engineering for Gemini, says in a blog post. Gemini’s context window, or the amount of chunks of text that an AI model can process at once, is also getting quadrupled to 32K tokens, Subramanya says.
1.5 Flash, which was announced at Google I/O in May, will be available in the free version of Gemini on the web and mobile starting today, according to Subramanya.
Google is announcing a few other pieces of Gemini news, too. To help you find factual information on topics you’re researching, the company is going to show links to related content in Gemini for “fact-seeking prompts” starting today, Subramanya says. You can see the links by clicking on a gray arrow at the end of a paragraph.

GIF: Google

In addition, Google will start rolling out Gemini “gradually” in Google Messages to users in the European Economic Area, UK, and Switzerland. And Gemini for teens will be available in over 40 languages “in the coming week.”

Illustration: The Verge

Google’s Gemini AI chatbot will be able to respond to you more quickly and process more content in prompts thanks to an upgrade to the company’s Gemini 1.5 Flash AI model.

With the upgrade to 1.5 Flash, you’ll see “across-the-board improvements in quality and latency, with especially noticeable improvements in reasoning and image understanding,” Amar Subramanya, VP of engineering for Gemini, says in a blog post. Gemini’s context window, or the amount of chunks of text that an AI model can process at once, is also getting quadrupled to 32K tokens, Subramanya says.

1.5 Flash, which was announced at Google I/O in May, will be available in the free version of Gemini on the web and mobile starting today, according to Subramanya.

Google is announcing a few other pieces of Gemini news, too. To help you find factual information on topics you’re researching, the company is going to show links to related content in Gemini for “fact-seeking prompts” starting today, Subramanya says. You can see the links by clicking on a gray arrow at the end of a paragraph.

GIF: Google

In addition, Google will start rolling out Gemini “gradually” in Google Messages to users in the European Economic Area, UK, and Switzerland. And Gemini for teens will be available in over 40 languages “in the coming week.”

Read More 

The set perfectly recreates the pixelated 16-bit Mario and Yoshi sprites from Super Mario World. | Image: Lego

Lego has announced its latest collaboration with Nintendo at San Diego Comic-Con: a pixelated recreation of Mario and Yoshi from the 16-bit Super Nintendo classic, Super Mario World. Like Lego’s Art sets, it’s a 2D creation meant for display but also includes hidden moving mechanisms, bringing Yoshi to life.
The 1,215-piece set stands over 15 inches tall when assembled and features a yellow-caped Mario riding Yoshi atop a grass-covered hill. The front of the set is assembled from Lego tiles and looks to be a very accurate recreation of the 16-bit sprites used in Super Mario World.

Image: Lego
The mechanisms powering the display piece’s animations are all left exposed on the back.

But when you turn the set around, it becomes very obvious why Lego is positioning this as a display piece. Lego’s designers incorporated a pair of mechanisms to recreate the game’s sprite animations. Turning a crank attached to the base makes Yoshi’s legs run in an endless loop, while a knob on the back opens the dinosaur’s mouth and makes his bright red tongue pop in and out.
The Lego Super Mario World: Mario & Yoshi set is officially available for preorder starting today through Lego’s online store for $129.99 and will ship starting on October 1st. It will be the fifth adult-focused set released as part of Lego and Nintendo’s ongoing collaboration, joining the 2,646-piece NES console, 2,807-piece Mighty Bowser, 540-piece Piranha Plant, and 2,064-piece transforming question mark block.

The set perfectly recreates the pixelated 16-bit Mario and Yoshi sprites from Super Mario World. | Image: Lego

Lego has announced its latest collaboration with Nintendo at San Diego Comic-Con: a pixelated recreation of Mario and Yoshi from the 16-bit Super Nintendo classic, Super Mario World. Like Lego’s Art sets, it’s a 2D creation meant for display but also includes hidden moving mechanisms, bringing Yoshi to life.

The 1,215-piece set stands over 15 inches tall when assembled and features a yellow-caped Mario riding Yoshi atop a grass-covered hill. The front of the set is assembled from Lego tiles and looks to be a very accurate recreation of the 16-bit sprites used in Super Mario World.

But when you turn the set around, it becomes very obvious why Lego is positioning this as a display piece. Lego’s designers incorporated a pair of mechanisms to recreate the game’s sprite animations. Turning a crank attached to the base makes Yoshi’s legs run in an endless loop, while a knob on the back opens the dinosaur’s mouth and makes his bright red tongue pop in and out.

The Lego Super Mario World: Mario & Yoshi set is officially available for preorder starting today through Lego’s online store for $129.99 and will ship starting on October 1st. It will be the fifth adult-focused set released as part of Lego and Nintendo’s ongoing collaboration, joining the 2,646-piece NES console, 2,807-piece Mighty Bowser, 540-piece Piranha Plant, and 2,064-piece transforming question mark block.

Read More 

Image: Android Headlines

Google’s Pixel Buds Pro 2 could bring back the wing tip design we saw on the first generation of the Pixel Buds. A series of images shared with Android Headlines by reliable leaker OnLeaks show the buds with a slight bulge opposite the ear tip.
The wing isn’t as prominent as the so-called “stabilizer arc” that protruded from the original Pixel Buds. Our review of those buds from 2020 mentioned that the arc could cause discomfort after wearing them for long periods of time. Hopefully, the more subdued wing on the Buds Pro 2 won’t do the same.

Image: OnLeaks via Android Headlines

Other than the wing tip, the Pixel Buds Pro 2 appear to trade their predecessor’s black grills for larger ones that match the color of the buds. Android Headlines also obtained images of what looks like the Pixel Buds Pro 2 in several colors: gray, white, green, and pink.
Meanwhile, the case looks largely the same, featuring the same egg shape and LED light on the front. Leaked images from the accessory maker Spigen show that the Pixel Buds Pro 2 case may also come with a small hole next to the USB-C port. This could be a speaker that emits a sound when using Google’s Find My Device feature, but its purpose still remains unclear.

Image: OnLeaks via Android Headlines

The refreshed Pixel Buds Pro isn’t the only device that’s been leaked ahead of Google’s August 13th hardware event. Earlier today, we saw marketing materials containing what could be the specs for the Pixel 9, Pixel 9 Pro, and Pixel 9 Fold. We’ll probably see even more leaks in the weeks leading up to the event — if Google doesn’t fully reveal its devices first.

Image: Android Headlines

Google’s Pixel Buds Pro 2 could bring back the wing tip design we saw on the first generation of the Pixel Buds. A series of images shared with Android Headlines by reliable leaker OnLeaks show the buds with a slight bulge opposite the ear tip.

The wing isn’t as prominent as the so-called “stabilizer arc” that protruded from the original Pixel Buds. Our review of those buds from 2020 mentioned that the arc could cause discomfort after wearing them for long periods of time. Hopefully, the more subdued wing on the Buds Pro 2 won’t do the same.

Image: OnLeaks via Android Headlines

Other than the wing tip, the Pixel Buds Pro 2 appear to trade their predecessor’s black grills for larger ones that match the color of the buds. Android Headlines also obtained images of what looks like the Pixel Buds Pro 2 in several colors: gray, white, green, and pink.

Meanwhile, the case looks largely the same, featuring the same egg shape and LED light on the front. Leaked images from the accessory maker Spigen show that the Pixel Buds Pro 2 case may also come with a small hole next to the USB-C port. This could be a speaker that emits a sound when using Google’s Find My Device feature, but its purpose still remains unclear.

Image: OnLeaks via Android Headlines

The refreshed Pixel Buds Pro isn’t the only device that’s been leaked ahead of Google’s August 13th hardware event. Earlier today, we saw marketing materials containing what could be the specs for the Pixel 9, Pixel 9 Pro, and Pixel 9 Fold. We’ll probably see even more leaks in the weeks leading up to the event — if Google doesn’t fully reveal its devices first.

Read More 

Photo by Gary Hershorn/Getty Images

Southwest Airlines is about to change how it handles customer seating by switching to assigned and premium ones, a significant change from the sit-anywhere-you-want setup it’s had for over 50 years. The transformation comes as Southwest reported some turbulence in its earnings, with a 46 percent drop in profits for the second quarter, as reported by WSJ.
In a press release, Southwest Airlines CEO Bob Jordan said the changes will be based on employee and customer feedback. Jordan called offering seats with premium legroom and assignments “transformational” but did not offer a timeline or how it would look. The Airline said it has conducted simulation and live trials that helped it make the decision, and that customer preferences have evolved as they’ve started taking longer flights.
The new model comes as Southwest made other changes, like adding faster Wi-Fi, power outlets, and larger overhead bins, and the airline is working to refresh its cabins and upgrade to more comfortable seats. It’s a series of promising upgrades for the airline — and will hopefully eliminate the mad dash for the best boarding group.

Photo by Gary Hershorn/Getty Images

Southwest Airlines is about to change how it handles customer seating by switching to assigned and premium ones, a significant change from the sit-anywhere-you-want setup it’s had for over 50 years. The transformation comes as Southwest reported some turbulence in its earnings, with a 46 percent drop in profits for the second quarter, as reported by WSJ.

In a press release, Southwest Airlines CEO Bob Jordan said the changes will be based on employee and customer feedback. Jordan called offering seats with premium legroom and assignments “transformational” but did not offer a timeline or how it would look. The Airline said it has conducted simulation and live trials that helped it make the decision, and that customer preferences have evolved as they’ve started taking longer flights.

The new model comes as Southwest made other changes, like adding faster Wi-Fi, power outlets, and larger overhead bins, and the airline is working to refresh its cabins and upgrade to more comfortable seats. It’s a series of promising upgrades for the airline — and will hopefully eliminate the mad dash for the best boarding group.

Read More 

Image: The Verge

The ClaudeBot web crawler that Anthropic uses to scrape training data for AI models like Claude has hammered iFixit’s website almost a million times in a 24-hour period, seemingly violating the repair company’s Terms of Use in the process.
“If any of those requests accessed our terms of service, they would have told you that use of our content expressly forbidden. But don’t ask me, ask Claude!” said iFixit CEO Kyle Wiens on X, posting images that show Anthropic’s chatbot acknowledging that iFixit’s content was off limits. “You’re not only taking our content without paying, you’re tying up our devops resources. If you want to have a conversation about licensing our content for commercial use, we’re right here.”

Hey @AnthropicAI: I get you’re hungry for data. Claude is really smart! But do you really need to hit our servers a million times in 24 hours?You’re not only taking our content without paying, you’re tying up our devops resources. Not cool.— Kyle Wiens (@kwiens) July 24, 2024

iFixit’s Terms of Use policy states that “reproducing, copying or distributing” any content from the website is “strictly prohibited without the express prior written permission” from the company, with specific inclusion of “training a machine learning or AI model.” When Anthropic was questioned on this by 404 Media, however, the AI company linked back to an FAQ page that says its crawler can only be blocked via a robots.txt file extension.
Wiens says iFixit has since added the crawl-delay extension to its robots.txt. We have asked Wiens and Anthropic for comment and will update this story if we hear back.
iFixit doesn’t seem to be alone, with Read the Docs co-founder Eric Holscher and Freelancer.com CEO Matt Barrie saying in Wiens’ thread that their site had also been aggressively scraped by Anthropic’s crawler. This also doesn’t seem to be new behavior for ClaudeBot, with several months-old Reddit threads reporting a dramatic increase in Anthropic’s web scraping. In April this year, the Linux Mint web forum attributed a site outage to strain caused by ClaudeBot’s scraping activities.
Disallowing crawlers via robots.txt files is also the opt-out method of choice for many other AI companies like OpenAI, but it doesn’t provide website owners with any flexibility to denote what scraping is and isn’t permitted. Another AI company, Perplexity, has been known to ignore robots.txt exclusions entirely. Still, it is one of the few options available for companies to keep their data out of AI training materials, which Reddit has applied in its recent crackdown on web crawlers.

Image: The Verge

The ClaudeBot web crawler that Anthropic uses to scrape training data for AI models like Claude has hammered iFixit’s website almost a million times in a 24-hour period, seemingly violating the repair company’s Terms of Use in the process.

“If any of those requests accessed our terms of service, they would have told you that use of our content expressly forbidden. But don’t ask me, ask Claude!” said iFixit CEO Kyle Wiens on X, posting images that show Anthropic’s chatbot acknowledging that iFixit’s content was off limits. “You’re not only taking our content without paying, you’re tying up our devops resources. If you want to have a conversation about licensing our content for commercial use, we’re right here.”

iFixit’s Terms of Use policy states that “reproducing, copying or distributing” any content from the website is “strictly prohibited without the express prior written permission” from the company, with specific inclusion of “training a machine learning or AI model.” When Anthropic was questioned on this by 404 Media, however, the AI company linked back to an FAQ page that says its crawler can only be blocked via a robots.txt file extension.

Wiens says iFixit has since added the crawl-delay extension to its robots.txt. We have asked Wiens and Anthropic for comment and will update this story if we hear back.

iFixit doesn’t seem to be alone, with Read the Docs co-founder Eric Holscher and Freelancer.com CEO Matt Barrie saying in Wiens’ thread that their site had also been aggressively scraped by Anthropic’s crawler. This also doesn’t seem to be new behavior for ClaudeBot, with several months-old Reddit threads reporting a dramatic increase in Anthropic’s web scraping. In April this year, the Linux Mint web forum attributed a site outage to strain caused by ClaudeBot’s scraping activities.

Disallowing crawlers via robots.txt files is also the opt-out method of choice for many other AI companies like OpenAI, but it doesn’t provide website owners with any flexibility to denote what scraping is and isn’t permitted. Another AI company, Perplexity, has been known to ignore robots.txt exclusions entirely. Still, it is one of the few options available for companies to keep their data out of AI training materials, which Reddit has applied in its recent crackdown on web crawlers.

Read More