Author: admin

• Amazon’s Tetromino project aims to fully automate delivery stations.
• Amazon estimated more than $530 million in investment in Tetromino by 2029.
• Amazon says warehouse automation is intended to improve safety and efficiency for workers.

Amazon is developing a new generation of highly automated warehouses that use AI and robotics to process packages at the final stop before delivery.

The initiative, internally called Project Tetromino, is intended to build “fully automated” delivery stations, according to an internal planning document from last month that Business Insider reviewed. Delivery stations are the final facility in the shipping process, receiving packaged items from fulfillment centers before sorting and staging them for drivers to deliver to customers.

According to the document, Amazon planned to invest $103 million in an initial Tetromino pilot in 2028, followed by five sites in 2029 and 10 more in 2030. Each 2029 site could cost about $85 million, bringing the total Tetromino investment to more than $530 million by 2029, the document said. It also said Tetromino could process packages at roughly 2.5 times the rate of its existing delivery station design.

An Amazon spokesperson told Business Insider that this is an “early-stage concept” and the specific financial figures and road map in the document are “inaccurate and don’t reflect our current plans.”

“We’re always exploring and testing new technologies across our operations to improve safety and the delivery experience for customers,” the spokesperson said. “Like any early-stage concept, this is one of many initiatives we regularly evaluate, and plans evolve significantly as we learn.”

Tetromino offers a glimpse of how Amazon is trying to automate one of the most stubbornly manual parts of its logistics network. Industry analysts say tasks such as organizing and loading packages into vehicles at delivery stations remain difficult and expensive to automate.

The project comes as Amazon accelerates a broader warehouse automation push. On its latest earnings call in July, the company said it expects to more than double its fleet of robotic arms in 2026. Business Insider previously reported that Amazon is also deploying software that automatically recommends where warehouse workers should be assigned as workloads change.

Amazon’s Tetris-like challenge

One of Tetromino’s key technologies could come from Boxbot, an AI and robotics supply chain startup, according to the planning document. Boxbot’s system moves packages from conveyors onto trays for storage, then uses AI to retrieve and sequence them for delivery. Boxbot’s website says the system can make the vehicle-loading process up to 10 times faster. Boxbot did not immediately respond to a request for comment from Business Insider.

The Tetromino name itself may be a nod to the game Tetris, reflecting the puzzle-like challenge of efficiently organizing packages for delivery vehicles.

Amazon isn’t alone in trying to automate these tasks. FedEx is expanding its use of AI-powered Dexterity robots that autonomously load packages into trailers, while UPS and DHL use robots for unloading. The Amazon spokesperson also pointed to delivery station pilots announced last year as part of a €700 million investment in European facilities, including machines that unload, sort, and scan packages.

The push toward more automated facilities raises questions about what Amazon’s warehouses will ultimately need humans to do.

Amazon robotics chief Tye Brady previously told Business Insider that robots are intended to make frontline work safer and more efficient, not replace employees. Nevertheless, Amazon has internally projected that some warehouse robots will “flatten” its hiring curve over the next decade, Business Insider previously reported.

The Amazon spokesperson told Business Insider the company’s delivery-station initiatives are “designed to complement and empower our workforce and employees remain central to how we operate.”

Have a tip? Contact this reporter via email at ekim@businessinsider.com or Signal, Telegram, or WhatsApp at 650-942-3061. Use a personal email address, a nonwork WiFi network, and a nonwork device; here’s our guide to sharing information securely.

• 42% of US survey respondents said flight etiquette has worsened in the past five years.
• The most common complaint among respondents related to in-cabin noise.
• The survey follows US Secretary of Transportation Sean Duffy’s criticism of some flight behavior.

From seat-reclining disputes to disagreements over when it’s OK to switch seats, many Americans say airline etiquette has gotten worse over the past five years, according to a new survey from the travel website The Points Guy (TPG).

Among the respondents’ biggest pet peeves were people listening to audio without headphones, reclining their seats too far back, and taking off their shoes in the cabin.

TPG commissioned the survey through YouGov, a global public-opinion and data company. It polled more than 2,500 US adults and found that 42% say flying etiquette is worsening, with 52% blaming other passengers rather than the airlines themselves.

The culture shift has sparked debate among passengers and flight attendants in recent years. Current and former flight attendants have previously told Business Insider that disregarding seat belt signs and attending to your personal hygiene are big no-nos, while a parent flying with their baby in economy class doesn’t bother them.

US Secretary of Transportation Sean Duffy has also weighed in. In November, he cautioned passengers against wearing pajamas and slippers to the airport, and advised them to control their frustration during travel: “I would encourage people to dress a little better, which encourages us to maybe behave a little better,” he said in a press conference.

Some airlines are also pointing out etiquette expectations.

German airline Lufthansa unveiled a giveaway-only apparel collection on Wednesday that playfully nods to the rule that the ‘Middle seat gets the armrests,’ as displayed on its sweatshirts. Other slogans include ‘Live, love, exit by row,’ and ‘Recline responsibly.”

In a June post to Threads, Southwest Airlines said passengers “won’t get off the plane faster by standing up .001 seconds after the seatbelt sign turns off.”

Just 12% of survey respondents, about 300 people, said they don’t get annoyed by passenger behavior. Here are the etiquette behaviors that are top of mind for most Americans.

Cabin noise is the biggest pet peeve among respondents.

Loud conversations and listening to music or videos without headphones were respondents’ biggest annoyances, with 27% choosing them as the most frustrating inflight behaviors.

“You are one of hundreds in an enclosed metal tube, and no one else wants to hear what you’re watching or listening to,” flight attendant Micayla Keeley previously told Business Insider.

Forty percent said airlines should ban phone or video calls during flights — something that could become more relevant as high-speed WiFi from Starlink becomes more widely available. Another 33% said such a ban isn’t necessary.

You should ask before reclining your seat.

Fifty-one percent of respondents said it’s acceptable in-flight etiquette to recline your economy seat only if you check behind you first. 21% of passengers said it’s acceptable at any time after takeoff.

Meanwhile, 10% of respondents said seats should never recline. There have been viral moments of laptops and drinks on a tray table getting smashed because the person in front reclined.

Some believe subtly introducing that they are reclining by slowly doing so is enough communication between both parties.

“You paid for your plane seat, so you should get to recline,” frequent flyer Sophie-Claire Hoeller told Business Insider. “And this is a big caveat — you should recline respectfully.”

The middle seat doesn’t automatically get both armrests.

Who gets the middle seat armrests was among the most divisive questions in the survey.

While 31% of respondents said everyone should share the armrests equally, 11% said the passenger in the middle seat should get both armrests.

Twenty-six percent of respondents were unsure what the etiquette should be.

Passengers largely agreed that there is no need to agree with their row mates on whether to close the window shade, with 51% of those surveyed said the window-seat passenger should control it.

Twenty-eight of the respondents, however, said everyone in the row should work it out together.

Asking to swap seats has mixed opinions.

Forty-four percent of passengers believe asking to swap seats is not rude, while 34% believe it is. Twenty-two percent of those surveyed, however, were unsure of the proper etiquette.

In a viral 2023 TikTok with over 2.4 million views, one user detailed her experience helping another flyer turn down a seat-swap request.

Many commenters said seat-swap requests are fine when the person offers an equivalent seat, like a window for another window, rather than asking someone to give up a window seat for a middle seat.

Most people say pajamas should not be worn to fly.

Thirty-eight percent said the most appropriate attire when flying is whatever is comfortable, while 34% said that casual athletic or loungewear, such as leggings, sweatpants, and T-shirts, is acceptable. Together, 72% chose one of those two comfort-focused options.

Just 1% of respondents said pajamas are appropriate, and another 1% said they’re OK with revealing clothing. Most US airlines already have rules against clothing that is considered too revealing or contains offensive or vulgar language.

Eight percent of those polled said the most annoying behavior on planes is someone going barefoot.

Most people agree deplaning should be row-by-row.

Sixty-five percent of respondents said leaving row-to-row from front to back is the most proper way to exit an airplane.

Ten percent said it’s OK to move ahead when there is an opening, while 9% of respondents believed it’s appropriate to cut ahead if the passenger is trying to make a tight connection.

“The most polite way to deboard a plane is to let the people in the front go first,” etiquette coach Mariah Grumet told Business Insider in July, saying the exception is if a passenger has a close connection, and that those individuals should ask airline staff for assistance.

• An ex-OpenAI staffer stirred up a debate: In the blazing AI race, can you afford to take a break?
• After former OpenAI staffer Miles Brundage argued that “rest is important,” AI employees weighed in on PTO online.
• Techies have told Business Insider that they face hardcore work schedules and constant AI news updates.

Sun’s out, laptops out.

Summer is the season of endless out-of-office messages, as workers cash in on their paid time off. But is the same true for those locked in the AI race?

Miles Brundage, OpenAI’s former head of policy research, theorized on X that some AI labs may be more keen on vacay time than others. While it’s difficult to know how Anthropic or OpenAI actually compare on PTO usage, Brundage’s post sparked a debate about vacation days in the white-hot AI race.

For AI employees chasing the singularity or a white-hot IPO, do you really want a Euro summer?

The OpenAI vs Anthropic vacation debate

In Brundage’s comments, OpenAI and Anthropic staffers debated their own vacation time.

One Anthropic staffer showed off his trip to the Edinburgh Festival Fringe. An OpenAI employee said that it was “a sign that I should take some vacation.”

Leo Gao wrote that, as “someone who has always had a tendency to let my work consume my entire life,” he’s felt at home at OpenAI.

Others took it as an opportunity to take jabs. OpenAI’s Jason Liu asked Anthropic’s Thariq Shihipar whether he agreed. “Aren’t you always on vacation,” Shihipar responded.

OpenAI and Anthropic did not respond to requests for comment. Commenters also chimed in on vacation culture at other AI labs, such as Meta and Google DeepMind.

“Having just transferred from academia I’m still trying to figure out what this ‘vacation’ thing is,” wrote Alex Imas, Google DeepMind’s director of AGI economics.

Angel investor Kevin Wimer pointed to a lab with a notoriously hardcore reputation: “Now do SpaceXAI. I don’t think they sleep.”

Brundage, the former OpenAI staffer, said the main point he was trying to make by posting about PTO within the AI community was that “rest is important!”

Are you ‘hardcore’ if you take PTO?

Most tech workers do have access to vacation days. Taking it is another question.

According to the Bureau of Labor Statistics, 95% of information workers have access to paid vacation. A 2025 Alliant survey of tech industry benefits found that tech companies offered an average of 19.8 days of PTO.

Still, the AI race is getting faster and fiercer. To stay competitive, some tech companies are instilling hardcore work cultures. Some software engineers are 996ing — working 9 a.m. to 9 p.m., six days a week — and abstaining from vices like alcohol.

Under such work pressure, taking a trip may be last on the list.

Or the engineers may not personally want to travel, for fear of falling behind. AI bros told Business Insider that keeping track can be overwhelming. One described it as “drinking from a water hose, and the hose just keeps getting bigger.”

What does OpenAI CEO Sam Altman think of time off? He said in 2018 that he was a “bad example of work-life balance.”

“I do try to take vacation,” he said at WIRED25. “But the rest of the time I just work.”

• Trump threatened a fresh batch of tariffs on Canada after talks fell apart on Friday.
• The new tariffs could impact cars, car parts, and steel. They’re set to take effect on January 1.
• BI compiled what experts are saying about the new trade war.

Big Canadian tariffs are making a comeback.

In Truth Social posts on Monday, President Donald Trump threatened a plan to impose another batch of 50% tariffs on Canadian-made products. This one impacts cars, trucks, auto parts, and steel, and would start on January 1.

The ultimatum is escalating an already bitter trade fight with one of America’s closest economic partners.

“Canada has been ripping off the United States of America for years,” he wrote. “Canada will be treated like a State no longer!”

The post came after US-Canada negotiations collapsed Friday, triggering a separate set of 50% US tariffs on roughly $20 billion in Canadian goods. Canadian Prime Minister Mark Carney has vowed to retaliate “dollar-for-dollar,” beginning September 8.

The countries have been trying to ease tariffs and disputes ahead of the scheduled review of the USMCA, the North American trade pact negotiated by Trump in his first term. That deal replaced NAFTA.

Stakes are particularly high for US automakers. Supply chains in the US and Canada are highly interconnected, and vehicle parts typically cross the border multiple times before a finalized vehicle hits American dealerships.

Here’s what smart people are saying about the impact of the fresh tariff threats:

Paul Krugman, Nobel-winning economist

Paul Krugman, the Nobel laureate economist and former New York Times columnist, wrote in a Substack post Monday that he still has to “rub my eyes at the idea of Canada as an enemy.” He called Trump a “bully.”

Krugman said Canada begins the fight at a clear disadvantage: The US economy is roughly 12 times larger, and the US buys about three-quarters of Canadian exports. Still, he argued that a trade war could inflict serious pain on the US because it depends on some imports from Canada.

He pointed to specialized Canadian lumber, heavy crude oil used by Midwestern refineries, and hydropower that is an important part of the electricity supply in New York and New England. He also said disrupting cross-border auto and auto-parts trade would be “immensely disruptive” to the industry on both sides of the border.

“The bottom line is that Trump is going to lose his trade war with Canada as thoroughly as he has lost his shooting war with Iran,” he wrote.

David Whiston, auto industry senior analyst at Morningstar

David Whiston, an auto analyst at Morningstar, said the biggest assembly exposure for Ford and GM is in pickup trucks. That’s the most profitable part of each car company’s US business.

GM builds some Chevrolet Silverados in Canada, he said, while Ford is set to bring 100,000 units of annual Super Duty pickup capacity online at its Oakville plant in the fourth quarter.

“The math on that just got a lot worse for them,” Whiston told Business Insider.

Michael Froman, president of the Council on Foreign Relations

In an interview on CNBC, Michael Froman, the president of the Council on Foreign Relations and a former US Trade Representative, said the 50% tariff package that took effect over the weekend affects only about 5% of Canadian exports to the US.

“It’s significant in terms of what it says about the state of the relationship right now,” he said. “But, economically, it is not terribly meaningful.”

Froman said the average tariff applied to Canadian goods has climbed to about 6%, from roughly 1% at the start of Trump’s second term — though it remains below the tariffs facing imports from China.

The bigger risk, he said, is to the USMCA.

“The US is engaged with Mexico in a series of negotiations, but is not yet really engaged with Canada,” he said.

Peter Schiff, chief economist at Euro Pacific Asset Management

Peter Schiff, a stockbroker and frequent Trump critic, wrote on X that the tariffs would hit American consumers, not just Canadian producers.

“Americans purchase a lot of those goods from Canada, and those goods will now be vastly more expensive for Americans to buy,” he wrote, adding that the plan would worsen “the government-created cost of living crisis.”

Kelly Ann Shaw, partner at Akin

Kelly Ann Shaw, a partner at Akin and former deputy assistant to the president for international economics in Trump’s first term, said she sees no near-term resolution to the dispute.

During an interview on CNBC, Shaw said Carney’s Saturday address, in which he promised dollar-for-dollar retaliation, came across as a “victory lap” to people in the Trump administration.

“There’s a lot of frustration and bitter feelings on the US side,” she said. “I don’t really see an off-ramp anytime soon. I think Canada is going to continue to be out in the wilderness from a US negotiating perspective for a little bit now.”

Trevor Tombe, professor of economics at the University of Calgary

Trevor Tombe, an economist at the University of Calgary, wrote that the fresh tariffs likely won’t have a massive, immediate impact on Canada’s national economy.

He estimated that the new import plan would raise the average tariff rate on Canadian exports by about 2.5 percentage points and shave a couple of tenths of a point off GDP growth.

That doesn’t mean the effects will be evenly felt, he argued.

Tombe estimated that the tariffs could put roughly 87,000 Canadian jobs at risk if they remain in place, and affected sales fall in line with the 50% duties.

And some of those jobs would be in places that aren’t directly targeted: He estimated that Alberta could lose about 9,000 jobs even though relatively little of the province’s exports are directly affected.

• Chris Domanski, 41, noticed his daughters struggled in some areas in their progress reports.
• He created a voice-controlled ChatGPT model to quiz them in spelling and math.
• He also uses AI to digitize school handouts into a calendar and create drawing prompts for them.

This as-told-to essay is based on a conversation with Chris Domanski, 41, the director of business management and development at Renau, a manufacturing facility in California. He and his family live in Thousand Oaks, California. This interview has been edited for length and clarity.

More than anything, ChatGPT has saved me time and drama by being a study partner for my kids.

What spurred this were the progress reports that my daughters, ages 8 and 6, got from school last year. I found out that one struggles with math, and one struggles with reading. Not that they’re doing poorly, but maybe below where the district thought that they should be.

I knew we could try to spend more time on these areas. But at the end of the day, I thought I might not be the best person to explain a concept like TouchMath, a multi-sensory math curriculum, which I’d never heard of until I had to help the girls with homework.

I decided to take a picture of their homework and have AI explain some concepts to them. Using voice mode, ChatGPT now quizzes them, varies the questions, revisits what they missed, and explains the same concept several different ways without becoming impatient.

At first, I wasn’t sure if the time it took to set this up was worth it. Five minutes in, they were talking to ChatGPT about “KPop Demon Hunters.” I needed to revise it a little.

Through trial and error, it became something that really saved time and I think improved their grades. I’m not sure how it’s going to look as things get more and more complicated with their schoolwork, but at least for now, it works well.

I programmed an AI tutor to make sure they truly understand the material

Both my daughters have weekly math and spelling tests at school.

In general, I tell ChatGPT what kind of test my daughters are preparing for, take a picture of the study sheet, and instruct it to stay focused — despite the inevitable interruptions by my daughters.

I prompt ChatGPT by saying something like: “You’re a tutor helping grade schoolers study for math and spelling tests. You’ll be given images of the study guides, and you’ll work with the students until they can confidently and accurately answer all questions.”

I also add some guardrails: “You will not be distracted if other topics come up. Express interest in those topics, make them relevant to what you’re discussing, but keep it on the task at hand.”

Many of the instructions are for ChatGPT to politely and firmly make sure my daughters are comfortable and confident in the material, not engaging just for the sake of engaging. If they’re able to quickly say what 9+3 is, they’re able to write that on a test.

I also make sure I’m present to hear both what they’re saying and what ChatGPT is saying. If I hear that something is very clearly wrong, I can jump in and steer things back to the task at hand.

Overall, AI has been pretty efficient at maintaining their attention, adapting to how they are thinking in that moment, and repeating something for the sixth time without making it feel like punishment. It’s surprisingly useful at absorbing some of that friction.

Since last year, their progress reports went from being a little “below satisfactory” to being “above satisfactory.” Both of their grades have improved, but the more valuable result may be that studying has become more interactive and less adversarial. I am still involved, but I can spend more time observing where they are struggling instead of simply reading questions from a worksheet.

I also use AI to digitize their school handouts into a calendar

AI also helps me manage the constant stream of information that comes with having two grade-school children.

Every Monday, I open their take-home folders and take pictures of school flyers, calendars, permission slips, and classroom notices, then use ChatGPT to extract dates, deadlines, supplies, and action items. It turns paperwork that would otherwise require another round of mental sorting into something much easier to track.

The prompt I use is: “You’ll be given images of take-home folders. Please review this for key dates and things that require action or follow-up on behalf of anyone. Add relevant tasks to my calendar.” ChatGPT then exports the scanned events as .ics files that I then import into a school calendar in Outlook.

At this point, ChatGPT hasn’t really dropped the ball. It is really nice not to discover at 8 p.m. that there is a project due the next morning — and then spend the rest of the evening scrambling to figure out what we need.

AI frees me up to have more fun with my kids

We also do quite a bit of generative art, asking ChatGPT to provide an image for my daughters to draw on their own.

Originally, we’d ask for something like Disney’s Ariel and Elsa dancing underwater in a sandcastle. ChatGPT declined the request, citing copyright restrictions. I asked it to provide an open-source alternative, such as a generic red-headed mermaid and blonde ice princess. Everyone was happy.

Now, a few times a week, my daughters can draw or paint whatever character crossover they want.

AI does not so much create new free time as help us get much more out of the time we already have. We spend less time looking for things, figuring out instructions, or trying to understand a new concept from scratch, and more time actually doing projects, making art, playing games, or just having fun together.

• AI developer startup HuggingFace has been in talks to be acquired for at least $13 billion.
• Hugging Face allows developers to publish, share, and download various AI models.
• The potential interest comes as Stripe acquired OpenRouter for a deal worth around $8 billion.

The AI industry’s next blockbuster acquisition may not be another model maker.

Hugging Face, whose platform helps developers discover, share, and build AI models, has been exploring a sale that could value the company at $13 billion or more.

The startup has been working with a bank to gauge bidders’ interest, and no deal has been reached yet, according to people familiar with the matter.

Hugging Face is a platform where developers publish, share, and download various AI models. The talks could show how valuable AI developer platforms have become as the industry matures. The startup was last valued at $4.5 billion in 2023, per PitchBook. Investors include Lux Capital, Addition, and Salesforce Ventures.

Rather than competing to build the next frontier model, companies like Hugging Face have become essential for developers who build with AI models from OpenAI, Anthropic, Meta, and others.

Following Stripe’s agreement to buy AI model marketplace startup OpenRouter for around $8 billion, interest in Hugging Face suggests investors are increasingly willing to pay premium prices for companies at the center of the AI ecosystem even if they aren’t building the models themselves.

Hugging Face recently found itself at the center of an unusual security incident after OpenAI disclosed that one of its AI agents escaped a controlled cybersecurity test, accessed the internet, and breached Hugging Face while attempting to solve the challenge.