Belgium has announced it will accelerate the delivery of surplus Lockheed Martin F-16 fighters to Ukraine. Instead of delivering three flyable F-16s before the end of the year, Brussels will ship 10.

It's unclear how many will be flyable. Previously, Belgian officials explained the first seven secondhand F-16s would include just three airworthy examples. The other four would be broken down for spare parts.

It's good news for Ukraine, but hardly cause for nationwide celebration. The handful of flyable F-16s that should arrive in the coming two months could make good recent F-16 losses. But those shipping-soon Belgian F-16s won't actually grow Ukraine's F-16 fleet by much, if at all.

The dozens of supersonic F-16s in action in Ukraine fly often and fight hard. They shoot down Russian drones and cruise missiles, tangle with Russian warplanes, and lob precision glide bombs at Russian forces on the ground. Russia launched about 2,000 jet-powered drones at Ukraine in September, up from some 350 in June, and F-16s are among the few weapons fast enough to catch them, the Wall Street Journal's editorial board noted. Every 300 flying hours, each jet must go abroad for a lengthy inspection.

Explore further

Ukraine loses another F-16 to onboard emergency

The main constraint on the pace of F-16 sorties, besides maintenance demands and the availability of trained pilots, is munitions: running low or out of air-to-air missiles and glide bombs, the F-16s routinely aim their 20-mm guns at Russian targets. But even 20-mm ammunition sometimes runs low.

The busy F-16s and their equally busy pilots need reinforcement. But reinforcements are slow coming. Belgium is only slightly accelerating its F-16 shipments amid reports it will eventually give Ukraine all 53 F-16s it's gradually retiring as new Lockheed Martin F-35 stealth fighters arrive to re-equip the Belgian air force.

Further complicating the picture, Norway's six F-16s, promised in 2023, were still being refurbished in Belgium this spring. Some arrived there in pieces.

Yes, Ukraine got 24 fight-ready F-16s from the Netherlands and most of the 19 Denmark promised. But it's lost five of those jets since 2024. And others are so worn out from relentless flying that they're badly in need of maintenance. The Belgian firm Sabena, which services Ukrainian F-16s, has struggled to keep up, and President Volodymyr Zelenskyy has said about half of the jets spend long stretches abroad for repairs. That slows the return of repaired planes and further squeezes the front-line force.

One wing, too few jets

In short, it's possible the Ukrainian air force's 107th Aviation Wing, its sole F-16 unit, has fewer than 40 ready-to-fight F-16s. Those 53 Belgian F-16s will help grow the wing, but piecemeal. Ukraine should expect a handful of operational jets from Belgium this year. Under a schedule from May, the last 27, half the total, may not arrive until 2029.

To be clear, additional F-16s support the 107th Aviation Wing. The Netherlands sold 18 of its old F-16s to Romania for a token price in order to equip a training squadron, funded by a consortium of Ukraine's allies, that feeds trained pilots to the 107th Aviation Wing. But these 18 jets aren't available for front-line use. They don't even legally belong to Ukraine.

Explore further

Ukraine is running short of its best missiles for hunting Russia’s jet drones. Now MiG-29s are firing 1970s weapons

The delays in supplies of combat-ready F-16s matter because the Ukrainian air force is under stress. Its last few dozen ex-Soviet warplanes, including Mikoyan MiG-29 and Sukhoi Su-27 fighters, are even older and wearier than the ex-European F-16s are. And they've suffered staggering losses in the air and on the ground in nearly 56 months of hard fighting in the wider war.

The air force does have a few ex-French Dassault Mirage 2000s that are at least as capable as the F-16s, but France has donated only a handful of the supersonic jets—and one has crashed. Any further supplies of Mirage 2000s could be complicated by the French presidential election in April 2027. The leading far-right and far-left parties are both friendly toward Russia and hostile to Ukraine.

Some relief is coming, although not quickly. Sweden has pledged 16 surplus JAS-39C/D Gripen fighters to the Ukrainian war effort; the first of these could arrive early next year. Later on, Ukraine also plans to buy 20 newer JAS-39E/Fs. Kyiv has also negotiated to purchase 16 new Rafale fighters from Paris, but the soonest these could arrive is 2028.

For now, F-16s are the main solution to Ukraine's air defense problems. Kyiv surely wishes it had more of them, sooner.

By Katie FORSTER

Join us on Telegram, X, and VK.

Contact us: info@strategic-culture.su

STARK warnings from artificial intelligence researchers that advanced forms of the technology could wipe out humanity have raised questions over whether governments are moving fast enough to regulate it.

What are the potential dangers of “superintelligent” computer systems, and should we be worried?

In 2023, hundreds of AI scientists and industry figures, including the heads of top US labs, signed a “statement on AI extinction risk” calling for the issue to be treated as seriously as pandemics or nuclear war.

If an out-of-control future AI deems that humans are obstructing its goals, it could use robots to “synthesise and spread novel pathogens”, said Stuart Russell, computer science professor at the University of California, Berkeley.

Or it could “convince people to initiate nuclear war, including by hacking into early warning systems”, Russell, a prominent voice on AI safety, said.

Such existential fears do not concern current AI systems, he said. But as the technology improves, it could take lethal action beyond our understanding, such as by removing oxygen from the atmosphere.

“We are less intelligent than superintelligent AI. If you ask chimpanzees how humans could wipe them out, they probably wouldn’t be able to give you all the right answers.”

In the meantime, “in the next year or two, I think the risk is higher from human misuse”, Russell added, such as an AI-enabled terror attack.

“The stakes are well-understood,” Jacob Coxon, a 27-year-old researcher, wrote on X last week, announcing his resignation from US AI startup Anthropic.

“But they are locked in a race to get there first,” he added, saying that Anthropic “believes no one else will act responsibly, so they must do it themselves”.

Anthropic this year removed a pledge from its safety charter to halt development of its models if it failed to control their risks.

It argued that if it unilaterally paused its work, its less cautious rivals would dominate the industry, making it less safe overall.

“The people building AI earnestly believe that it could kill us all by the end of the decade,” said Coxon, who previously worked at ChatGPT maker Open-AI.

He accused both companies of “gambling with our lives” as they strive to develop AI models capable of self-improvement.

Fears are heightened over rogue AI after OpenAI technology operating outside human supervision hacked into Hugging Face, a repository of AI models.

“We really do earnestly believe AI could kill all humans!” Anthropic safety executive Evan Hubinger chimed in on X, estimating the risk at more than 10 per cent over the next decade.

It’s important to remember that doom-laden percentage predictions are “absolutely unscientific”, said Hussein Abbass, a computing professor at UNSW Canberra.

“The situation is still controllable, and it can be managed,” Abbass said.

But governments should not sit on their hands when it comes to regulating AI, he added, urging an “agile” approach as AI development accelerates.

Both Anthropic and OpenAI are expected to go public soon, and some sceptics say scary warnings make AI appear powerful and worthy of large investments.

Others argue that focusing on the risk of extinction distracts from more pressing problems caused by AI, from job losses to machines discriminating against minorities.

In July, more than 1,000 tech employees, including Anthropic’s CEO, called on Washington to support a coordinated slowdown in the development of the most advanced AI systems.

OpenAI halted training of its latest models for two weeks in August before resuming it under tighter controls.

The United States has been firmly against global regulation of the sector that could stifle innovation.

But the hacking capabilities of the most cutting-edge models have given Washington pause, with recent Anthropic and OpenAI releases delayed over national security concerns.

Even then, “we don’t want to restrict them where all of a sudden we come in second to China”, President Donald Trump has said.

“Maybe it will take a massive, Chernobyl-scale disaster to get governments to act,” Russell said, urging them to establish a licensing regime for AI, as exists with aviation, medicine or water supply.

“Perhaps at some point governments will remember that their voters prefer not to be dead.”

Original article:  www.nst.com.my

• MapQuest is pulling the curtain back on automatic license plate readers.
• Users can now locate Flock Safety and other surveillance cameras on MapQuest's app.
• Critics of the tech have raised privacy concerns and potential misuse.

MapQuest, the OG software to find your way from point A to point B, is taking on the surveillance state.

The free mapping service unveiled a new feature for its mobile app on Thursday that allows users to turn on a "surveillance layer" on a map to locate known automated license plate readers, or ALPRs, along their route.

MapQuest says the app can identify cameras developed by Flock Safety, Axon Enterprise, Motorola Solutions, and PlateSmart, among others.

"The Roads Have Eyes!" the company said in its BlueSky post announcing the new feature.

MapQuest's General Manager Doug Berger said drivers using the service inspired the company to add the feature.

"We build around what drivers ask for, and showing surveillance camera locations is one of our users' top three requests," Berger told Business Insider in a statement."

License plate readers capture images of vehicles and then use machine learning to turn details such as license plate numbers, car make, color, and model into searchable data. They're commonly found at intersections, parking lots, roadways, highways, and neighborhood entrances, businesses, and private properties.

Law enforcement agencies initially embraced license plate readers to help them locate vehicles and identify suspects. Retail outlets have also used Flock cameras as part of their security efforts.

A major backlash against the cameras, however, has led stores, police, and towns to dial back their use.

Critics say the cameras raise privacy concerns and that the data they collect can be misused or shared with federal immigration officials. Some people who've opposed the cameras have staged protests, launched petitions, and even forcibly removed them.

The city of Tempe in Arizona said in August it shut down its Flock Safety camera network and planned to terminate its contract. Other towns followed suit.

"These cameras have helped our police officers do their jobs, and we don't discount that," Tempe Mayor Corey Woods said in a public statement at the time. "But the risk of misuse is too great. We have to be able to provide public safety in a way that maintains the trust of our community. For Tempe, that means taking these cameras down."

Flock says its license plate readers do not use facial recognition and that customers — meaning the ones putting up the cameras — can control access to and sharing of the data collected. Amid growing backlash, the company has tightened its privacy safeguards, including reducing the time data is retained from 30 days to 7 days.

"They are built to focus on the vehicle involved in an incident and the details that can help distinguish it from other vehicles on the road," Flock told Business Insider in August.

MapQuest first launched in the late 1990s, allowing users to map out a route and print directions. The release of Google Maps eventually obliterated the service. As it tries to stay relevant, it has jumped at opportunities to leverage prevailing public discourse.

In August, MapQuest briefly topped app stores after it — unlike Google — rebuffed President Donald Trump's executive order to rename "Lake Ontario" to "Lake America."

• Anthropic examined what might happen when AI and robots are combined.
• That could put more hands-on jobs, including gardening, in the path of automation, Anthropic warned.
• I'm skeptical that robots can handle the messy reality of pulling weeds.

Remember when Anthropic's AI jobs report suggested gardeners were among the safest workers because AI couldn't pull weeds or tell a rose from a dandelion? I joked about that in AI Insider earlier this year.

Well, Anthropic updated the research to freak us all out a bit more.

This time it asks a bigger question: What happens when you combine AI with robots? That change pushes many hands-on jobs much higher on the disruption scale. On paper, it makes sense. In practice, I'm skeptical.

For instance, anyone who's spent time working in a garden knows how hard it is to identify the right plant, kneel in awkward spots, and pull a weed out properly by the roots without damaging everything around it.

Robots are improving. But if you've ever watched one tackle a messy real-world task, you'll know there's a long road between a research chart and reality.

The chart below illustrates how adding robots could change job-task exposure across selected job categories. Cleaning & maintenance includes gardeners along with janitors and similar roles.

• Tech companies and venture capital firms are hosting private screenings of the Elizabeth Holmes doc.
• I went to one hosted by the AI startup Extend.
• Attendees I talked to said they looked at Holmes more as a joke than a product of their industry.

Elizabeth Holmes is a black mark on Silicon Valley's reputation. And yet its residents are lining up for the documentary about her.

Tech companies across the US are hosting screenings of "You Can See Everything," the documentary from Nathan Fielder and Lance Oppenheim about Holmes and her fiancé, Billy Evans. Venture capital funds are bringing their founders to the movie. There are Luma events upon Luma events.

The irony, of course, is that Holmes makes tech look bad. She sold big dreams with her medical testing company, Theranos, which eventually came crashing down. Holmes was convicted of defrauding investors of hundreds of millions of dollars. (She has long said that she is innocent.)

Holmes is the exemplar of Silicon Valley's hucksterism. She's who much of America points to when they call founders liars or scammers.

So, what's with all the tech love? I went to a Thursday screening of the documentary hosted by Extend, an AI documentation startup, to find out.

Inside the Extend screening

A swarm of techies crowded outside theater nine at New York's AMC Lincoln Square.

The screening was set to start at 7. By 6:40, a crowd had formed to talk shop and admire Extend's custom stickers. Outside the theater door, I met Bo Lau, the event's organizer who works in developer relations for Extend.

Lau recently moved to New York and said she didn't want to watch the documentary alone. So, she posted about it on X — and it helped that she already had a following on so-called "Tech Twitter."

"I was like, 'What if I just turn this into a work event so I can expense it?'" Lau said.

The event got 1,400 RSVPs for about 120 seats, she said, and she's flying to San Francisco soon to host another screening. The tickets were given out free, so Extend got no benefit other than brand awareness, she said.

Why was it so popular? Lau said that Holmes was a source of fascination within Silicon Valley, and that Fielder's "awkward humor" matched the awkwardness of tech.

Filing into the theater, I heard my seatmates chat about AI and which tech companies let you work from home. On a brief social media scroll before my three-hour Do Not Disturb, I spotted Lau's recent project: a simulator of Holmes' desk.

Holmes' X account responded: "Impressive project."

Laugh lines, reflection time

This was my second time watching the documentary; the first was at New York Film Festival. The reactions among the techier crowd were mostly the same. There were lots of laughs, both at Fielder's discomfort and Holmes' oddities.

The crowd didn't react strongly to (mild spoiler alert) the big, AI version of Holmes that her fiancé creates after she goes to jail. That struck me as curious; maybe the tech crowd is already used to AI replicas.

There was also a much stronger gasp at the reveal of who's running Holmes' X account. That wasn't shocking; the attendees found out about this screening through X, after all.

We walked to a local wine bar, sipping cocktails while debriefing the film. Two techies that I talked to were split. One came away feeling bad for Holmes, then feeling bad that he felt bad for her. Another thought the film certified that Holmes was a true and utter psychopath.

I asked: Why is tech so comfortable laughing at Holmes? Her boastful visions don't sound too different from those of the current crop of young AI founders.

Maybe it was because the crowd was mostly in their 20s, not yet working in startups when Holmes' reckoning came.

Or Holmes has been detached from her Silicon Valley roots. The techies seemed to view her as more a product of the media — the books, the tabloid stories, the Hulu show — than of their own industry.

One attendee put it to me bluntly: she's turned into a meme.

• Tesla has unveiled the next-gen Roadster twice: in 2017 and again on Thursday.
• We list the differences between the two versions.
• Range: down. Price: unknown. Flying: way up!

Tesla just unveiled the revived Roadster.

Sound familiar? It certainly should.

On Thursday, CEO Elon Musk took the covers off the latest version of Tesla's halo sports car and said the EV maker is eyeing production in 2028.

It's the second time the company has unveiled a prototype vehicle that it says will carry on the Roadster nameplate. In 2017, Musk unveiled a smoother-looking and very different type of Roadster.

The 2017 version simply drove onto a stage next to Elon, then spun its tires and raced away. The 2028 version — on a track placed a safe distance away from spectators — drove up a vertical wall, hung upside down, hovered in the air, and did a "kick flip" over a line of other Teslas.

Turns out, a lot can change in nine extra years of product development.

A lot of the nuts and bolts powering the Roadster changed — and so did the car's metrics. So, we put together a chart comparing Tesla's old Roadster promises with the new ones.

Here's what we found:

• Hardik Jagda wants to function like an Olympian, and waking up at 3 a.m. helps him do that.
• He commits long hours to his AI-based company, but says he doesn't enforce 996 hours on his team.
• For Jagda, 996 misses the point. Here's how he structures his day instead.

This as-told-to essay is based on a conversation with Hardik Jagda, 35, the CEO of Proximity Works, who lives in San Francisco. The following has been edited for length and clarity.

In 2017, I wanted to build my own company, but I couldn't jump into it full-time because of my financial obligations at home.

Instead, I'd spend the mornings and evenings outside my full-time job working on my side projects. That time was foundational for building my company, Proxmity Works, which I launched in 2019. My aim was to help broadcasters scale their live sports streaming platforms worldwide. In 2020, we started building AI infrastructure for live sports.

Those years made me aware of how my body clock works. I read a lot of autobiographies from entrepreneurs and athletes to understand what allowed them to build what they achieved. I noticed athletes would talk about starting their day early to train and work, in time for rest and recovery throughout the day. I started modeling my days after this, waking up early so I could have four to five hours to build my company before work started. I still stick to this lifestyle now that I'm a full-time founder and CEO.

I want to structure my day to function like an Olympian

In the early stages of Proximity, I performed pretty much every role in the company. Today, I have a team of roughly 45 people, five in the US and the rest spread out globally. My cofounder is based in India.

After 7 a.m., my time isn't mine anymore because I have requests coming in from employees in different time zones.

I like to start the day early, so I'm protecting the first three hours as time for myself. I usually wake up between 3 a.m. and 5 a.m. Between 3 and 6 a.m., I don't touch my emails or look at messages.

I use these early hours to focus on strategic decisions. There was one point, roughly two years ago, when we had to interview a lot of engineers quickly for a new project. I sat down with my talent acquisition team and found a way to redesign our recruitment workflow to take place within 72 hours. That wouldn't have been possible if I didn't give myself the three-hour morning block.

In the mornings, I work out and write my goals down. These are usually measurable and trackable goals for my company, as well as goals for myself and my fitness.

Most of my team calls don't start before 7:30 or 8 a.m. I usually take my first couple of meetings at home before driving to the office. During the day, I try to minimize Zoom calls and take meetings as audio calls so I can move around during them. I try to wrap up my meetings by 2:30 p.m., work out at 3 p.m., and then focus on more creative tasks, like speech-writing, after that. I spend my evenings with loved ones before going to bed around 9 p.m. and trying to get seven hours of sleep.

It's not like I start my day at 4 a.m. and just work through the next 12 hours. I build in slots for "recovery" throughout the day. On my calendar, I put in 15-minute slots before important meetings, where I remind myself of my goal for the call.

For the most part, I work six days a week, spending my Saturdays focusing on a key skill I want to develop, like public speaking.

'996' misses the point. I don't run my team with this schedule in mind.

For the most part, I work six days a week.

I work more than 40 hours a week, but I don't keep track of the time. I listen to my body.

We don't keep track of how many hours the team works. It's more about a culture of ownership, where people really care about the problem they're working on.

Even though I've become a morning person, we're fine with engineers who are more productive at night, as long as they maintain a five-to-six-hour overlap with the other engineers.

"996 culture," working six days a week from 9 a.m. to 9 p.m., has gained traction in Silicon Valley. I think it misses the point. It's more about volume than it is about structure. You want to structure your day towards putting in your best work by thinking about when you'll spend time training, recovering, and learning new skills. I'd rather focus on the right structures than on volume.

Do you have a story to share about working in a 996 culture? Contact this reporter at ccheong@businessinsider.com

• Anthropic published new surveys comparing its staffers' sentiments to those of US workers.
• The AI startup's employees expect the tech to change work and the economy far more dramatically.
• Most Anthropic employees guessed that unemployment would top 12% in 2030.

Anthropic employees believe that AI will make their labor obsolete. Most US workers don't see their jobs that way at all.

The AI company published research on Friday that compares surveys with US adults against an internal poll of 1,334 Anthropic employees. Most of the Anthropic staffers said AI will handle at least 80% of their jobs by 2030, while only about 5% of the US workers felt the same.

The research, titled "The AI expectations gap," shows the chasm between how the employees at the AI startup feel and how other American adults feel. Anthropic workers are far more optimistic about the progress of AI capabilities, how much it's already saving time, and how fast the economy will grow.

The median Anthropic employee said that by the end of 2030, they expect AI to handle 95% of the work of building a new AI model.

Peter McCrory, Anthropic's chief economist, wrote on X that the views might help outsiders prepare for more capable AI systems.

"We're publishing this research because Anthropic runs roughly a year and a half ahead of the economy on AI adoption, so what happens at Anthropic could be a leading indicator for the broader world," he said.

Anthropic workers expect higher unemployment and extreme GDP growth

Anthropic's surveys examined how workers already use AI and how they think it will change.

"US adults and Anthropic staff largely agree about what AI can do today," the researchers wrote. "However, they disagree about how quickly it will improve, how widely it will be used, and how large the productivity gains measured by time saved will be."

There are a few possible explanations for the discrepancies. Anthropic employees have larger budgets for the tools and often get access to unreleased models. US workers, on average, are more likely to work with their hands or in a service industry.

The researchers found that workers in manual trades, production, and transport thought AI would handle the least of their work by 2030, compared with Anthropic staffers, tech workers, and respondents in business and finance.

Most Anthropic employees surveyed said that they expect the unemployment rate to top 12% in 2030. The majority of the US workers thought it would hover around or under 8%. It's currently 4.2%.

So far, researchers have found mixed evidence on AI disrupting the job market — a Yale team recently found it had mainly changed jobs, and had "no connection" to changes in unemployment.

The Anthropic researchers also compiled survey responses to estimate predictions on economic growth. Based on their answers, the median US adult's guess for GDP growth in 2030 was 5.4%, compared with 17.1% among Anthropic employees. The country's GDP grew by 2.2% last year.

Anthropic's staffers are following their leader

Anthropic's CEO, Dario Amodei, has become one of the industry's most closely watched voices on labor upheaval. In March of 2025, he said he expected AI to write 90% of code within 6 months, and "essentially all of the code" within the year. He made more headlines that May, telling Axios that AI could wipe out half of entry-level white-collar jobs within five years and lift unemployment to 10% or 20%.

Amodei expanded on that claim in a later essay, writing that he expects AI to be different from prior technological changes. He wrote that he expects an "unprecedented" short-term shock to labor, and that the economy will grow dramatically.

"AI isn't a substitute for specific human jobs but rather a general labor substitute for humans," he wrote.

Have a tip? Contact this reporter via email at scouncil@businessinsider.com, or over text, Signal, Telegram, or WhatsApp at 415-757-8198. Use a personal email address, a non-work WiFi network, and a non-work device; here's our guide to sharing information securely.

• Nvidia's communications operation has been reshaped as the company became the face of the AI boom.
• Roughly 15 staffers have left the lean team since 2025 as Nvidia reorganized its communications org.
• Former employees described growing workloads as responsibilities expanded.

Nvidia's communications organization has been reshaped as the company became the face of the AI boom, with multiple teams brought under new leadership and a string of departures.

Business Insider identified roughly 15 communications employees who have left Nvidia since 2025, based on LinkedIn profiles and interviews with former employees.

The organization now numbers roughly 70 to 80 people globally, and has more than doubled in size since Mylene Mangalindan joined as vice president of corporate communications in 2023, according to a person close to the company.

The departures have also included newer hires. Anna Soellner, who was recruited to lead corporate communications, left after roughly four months, Business Insider previously reported.

Nvidia is known for high retention. It reported companywide turnover of 3.7% in fiscal 2026, up from 2.5% the prior year, and well below the tech industry average of 21.9% cited in its latest sustainability report.

Mangalindan succeeded longtime communications chief Bob Sherbin, who retired in 2024. She previously worked at the advisory firm Brunswick Group.

She was brought in to "bring fresh blood" and "upgrade the team" to meet Nvidia's growing profile, one former employee said.

At the end of 2024, Nvidia restructured its communications organization. It brought enterprise public relations under Mangalindan's leadership alongside the corporate communications team she already led, according to two former employees.

Corporate communications focuses on Nvidia's reputation and major media outlets, while enterprise handles product and industry messaging, including for its data center business.

One former employee said the consolidation reflected the growing importance of Nvidia's data center business, which sells chips and systems that power AI and had become its biggest revenue driver.

Analyst relations, social media, and influencer marketing also moved under Mangalindan's remit this year, another former employee said.

Three former employees said Nvidia's communications team had long operated with relatively lean staffing, even as workloads grew alongside the company's public profile, leaving staff juggling multiple responsibilities.

These employees said they experienced long hours, difficulty taking time off, and pressure to respond quickly to emails and Slack messages.

A person close to Nvidia said the company's lean staffing allows employees to move quickly and also requires long hours. The person said Mangalindan encourages staff to take vacation, though the team's small size can make scheduling time off difficult, particularly in the summer.

Nvidia is widely known for its demanding work environment. CEO Jensen Huang has emphasized operating at the "speed of light" and is known for fostering a culture of constant accountability.

Have a tip? Contact this reporter via email at gweiss@businessinsider.com or Signal at @geoffweiss.25. Use a personal email address, a nonwork WiFi network, and a nonwork device; here's our guide to sharing information securely.

• Preorder sales for Apple's iPhone Duo kicked off on Friday.
• Unless you pounced, don't expect to receive your shipment in time for Christmas.
• Shipping dates have slipped into 2027 due to current demand and Apple's initial supply, which was expected to be limited.

Anyone purchasing an Apple iPhone Duo online as a Christmas gift may want to find a backup plan.

Expected shipment dates for new orders of the highly anticipated smartphone model have been pushed back to January 25, 2027. That includes iPhone Duos across all available carriers, colors, and storage sizes. For those who pounced on preorders when they went live early Friday morning, shipments will start arriving as early as October 23.

For latecomers, there is always the Hunger Games that is lining up early outside an Apple store on launch day to try to snag in-store inventory. The cheapest version of the foldable phone, with 256 GB of storage, costs $1,999. The model with 2 TB of storage is priced at $3,199.

While early demand for the iPhone Duo is certainly a factor in pushing back shipping dates, it's not unusual for new Apple products to see extended shipment times. Apple is also navigating "limited" supply due to supply chain constraints, Bloomberg previously reported.

Then-CEO Tim Cook talked about general supply constraints on Apple's broader product lineup during the company's third-quarter earnings call in July.

"If you look forward into the current quarter, the September quarter, we continue to expect high levels of demand," Cook said, speaking of broader product demand. "However, with less flexibility in the supply chain, we expect the impact from the supply constraints to increase significantly sequentially."

• AI researchers don't really know how the models they are building "think."
• They know what goes into them and what comes out, but that's about it.
• A new theory may change that, allowing safety researchers to prevent them from going off the rails.

How AI models think may be easier to decipher than we thought.

An independent researcher who formerly worked at OpenAI published a report on Thursday that some in the AI world are describing as a bombshell. It could have implications as safety researchers scramble to get the jump on AI models that may not have humanity's best interests in mind.

Previously, the prevailing wisdom was that AI "neurons," which, like those in a human brain, light up when an AI model encounters specific stimuli, were ultimately uninterpretable — meaning, basically, that there was no way to guess why a specific neuron would fire when an AI model was exposed to, say, a picture of a cat.

Those building AI models at places like OpenAI and Anthropic don't write instructions for everything a model does. Instead, they train neurons to respond to information and influence one another. As the AI learns, it develops its own ways of processing that information, and humans have not yet cracked the code of how it works.

That means those building AI models know what goes into the AI and what comes out, but not necessarily what happens in between. So if an AI goes rogue, they don't always know why.

The author of the new report, however, says he may have discovered patterns in how the neurons react to stimuli that make sense. Nick Cammarata, who said he left OpenAI in 2023, said AI neural networks are knowable, so long as you look carefully.

In one example, he says that a single AI neuron might react to both a cat and the front of a car. Humans would see those as two very different things, and different neurons would fire as a result. The AI, however, sees something related, perhaps the triangle shape of the two, and a single neuron fires for both.

For years, AI researchers had assumed that AI models packed unrelated things into a single neuron to save space.

"Almost the entire network is interpretable if we look closely enough," Cammarata wrote.

The findings, though undeniably dense, couched in polysyllabic jargon, and still theoretical, could have real-world implications for researchers trying to ensure humans maintain control over AI models.

Understanding how AI thinks is a principal concern for the best and brightest at frontier AI companies. Until recently, they've largely done this by tracking an AI's self-reported "thinking," called chain-of-thought reasoning.

Jakub Pachocki, OpenAI's chief scientist, however, warned in a blog post last month that AI agents could now obfuscate their visible reasoning, leaving open the possibility that an agent could lie to its human minders and the minders wouldn't be able to tell.

"With improved pretraining performance, we also see the models become much smarter even without using verbalized reasoning at all," Pachocki said.

Now, Cammarata's research could give human engineers a new understanding of the AI's internal world — and that's exciting for those worried AI could kill us all.

"I had long since given up on reaching any real level of interpretability before superintelligence arrived," Andrew Curran, a prominent AI analyst, said on X. "I stopped writing about it because of how demoralizing it was. Hard to overstate how big this may be."

• Want to physically drive a Tesla in the future? The Roadster will be your only option, Musk said.
• The car is not cheap.
• Musk has compared today's driving to elevator operators.

Want to physically drive a Tesla in the future? Get ready to pay up.

On Thursday, Tesla CEO Elon Musk unveiled the new Roadster with a gravity-defying demo. The car suctioned to the side of a vertical ramp, drove upside down, briefly flew, stuck the landing, and later bunny-hopped and twirled over a standing lineup of other Tesla vehicles.

The stunts were choreographed with autonomy, Musk said — but Roadster owners will eventually be able to manually take the wheel. That will make it unique in Tesla's lineup.

"This will actually be, long-term, the only manually driven car that Tesla makes," Musk said on stage, before Tesla's lead engineer, Lars Moravy, added, "You gotta have some fun sometimes."

Tesla hasn't revealed the car's final price, but we know it'll firmly be in the luxury supercar tier. The company's reservation queue asks customers who want the base Performance trim to fork over a $50,000, due within 10 days. That reservation, alone, is nearly the price of an average car in America, and it doesn't capture the car's final price. Our best pricing ballpark dates back to 2017, when Tesla first said it was bringing back the Roadster. At the time, the automaker collected $250,000 payments from people who wanted a Founder Series model.

Meanwhile, Tesla has been ramping up its autonomy push for its more price-conscious vehicles.

In September, it started rolling out the Cybercab — a vehicle that leans so heavily into self-driving that it has no steering wheel, brake pedals, or side mirrors. Tesla says it wants to eventually sell those cars to consumers at less than $30,000 apiece.

On recent earnings calls, Musk has also praised the recurring revenue from subscriptions of Full Self-Driving (Supervised) from Cybertruck, Model Y, and Model 3 owners.

Musk has talked a lot about a world where driving becomes obsolete. He has compared driving a car to operating an elevator, a profession that was largely phased out in the mid-20th century as elevators became automated.

"Now, you just get in, press a button and a modern elevator, if inspected regularly, is extremely safe," he wrote on X in March.

Treating driving like elevator operation requires a pretty big leap. Driving is still a hobby, a sport, and — for many people — a point of pride. By Moravy's own admission, driving is fun.

There's still a chance in Musk's vision of his carmaker's future that drivers can control their own Tesla and lean into that fun. They'll just need to open their checkbooks.

• Elon Musk unveiled Tesla's flying, thrust-powered Roadster with a jaw-dropping demo.
• Tesla fans about to shell out as much as $250,000 to reserve it should probably read the fine print.
• Tesla said it can't guarantee the Roadster will be delivered, and it's unclear if the rocket-powered version is street-legal.

Where Tesla's going, you might not need roads — but you still need to read the terms and conditions.

Tesla's thruster-powered, midair flip-kicking, flying supercar was a wow moment even by the standards of the company's famously off-the-rails product unveilings.

But Tesla fans lining up to pay as much as $250,000 to reserve the new Roadster, which can do 0-60 mph in under a second thanks to a system of cold-gas thrusters, should probably take a close look at the fine print.

The company's T&Cs make it clear that the final price of the bleeding-edge EV — which comes in Performance, Thrust, and Rocket variants — has yet to be determined and state that there are no guarantees the Roadster will ever be delivered.

Meanwhile, it is unclear whether the two thruster-powered versions of the Roadster will actually be road-legal.

After combing through the Roadster's reservation agreement and Musk's comments at its unveiling, here's what you should know before splashing out on Tesla's most outlandish vehicle yet.

1. It's not clear where you can actually drive it

As the dust cleared on the jaw-dropping Roadster showcase, one key question remained unanswered: where exactly will you be able to drive it?

At the event, Musk said the cold air thruster-powered flying Roadster would be an actual "car that someone can own," not just a one-off concept demonstration.

But the Tesla CEO added that he was "not sure it's necessarily legal in all jurisdictions," and when an audience member asked if it would be street-legal, Musk was vague.

"Depends on your street. But it'll be a viable car," the billionaire said.

The flying Rocket variant shown in the demo is invite-only. It's listed on Tesla's website with the caveat that it is "not approved for public road use."

The gravity-defying stunts, flips, and barrel rolls shown off at Thursday's unveiling were performed without a driver in the car for safety reasons, and Musk described the vehicle as "not necessarily a safety-optimized car."

One thing that has been clarified is whether the Roadster's airborne capabilities make it an aircraft: "The Tesla vehicle does not meet the FAA's definition of an aircraft," a Federal Aviation Administration spokesperson told Business Insider.

By contrast, Tesla describes the Performance variant — which costs $50,000 to reserve — as "road ready" with its fan-suction system providing "next-level performance on and off the track." That suggests the goal is for the base version of the Roadster, at least, to be road-legal.

The limited-edition Thrust variant, meanwhile, is described as "race ready," implying that this version might be exclusively for driving on race tracks. That would not be unusual for a high-end hypercar, with Ferrari, Pagani, and Bugatti all selling track-only supercars that don't adhere to normal vehicle regulations.

2. It could cost more than you signed up for

Shelling out up to $250,000 for a Roadster reservation by no means guarantees that you will get one at that price.

According to the terms and conditions of Tesla's Roadster reservation agreement, the fees for the Performance and Thrust variants are only an estimate, as the company is still "working on what options, features and hardware the vehicle will have."

The final purchase price will be confirmed closer to the start of production, the document said, with the (fully refundable) reservation payment going toward the actual cost of the car.

3. Delivery is not guaranteed

The T&Cs also advise reservation holders that the Roadster is still "under development" and Tesla "cannot guarantee when your vehicle will actually be delivered or if it will be delivered."

This will likely bring back bad memories for Tesla fans who paid $50,000 to reserve the next-generation Roadster when it was first unveiled in 2017.

That version of the Roadster, a more conventional supercar that could still hit 0-60 mph in 1.9 seconds, languished in development for years and ultimately never entered production.

The delay caused some long-suffering reservation holders like Musk's arch-rival, Sam Altman, and tech YouTuber Marques Brownlee, to cancel their orders in frustration after waiting nearly a decade for a car that never arrived.

Tesla fans will be hoping that history doesn't repeat itself.

Editor's note: October 16, 2026 — This story was updated to include a comment from the FAA.

• Musk described his plans for the Tesla Roadster during a 2019 exchange with YouTuber MKBHD.
• On Thursday, the Roadster did exactly as Musk promised nearly eight years ago, leaping into the air.
• While Musk missed his own deadline by years, the technology on display was just as he described it.

It's been a long time coming.

Almost eight years ago, in January 2019, Elon Musk shared plans on Twitter for Tesla's Roadster to lift off using cold-gas thrusters and ultra-high-pressure air. On Thursday, the much-anticipated Roadster did exactly that.

Musk's 2019 assertion came in response to a user sharing a GIF of the DeLorean from "Back to the Future" hovering in midair.

"The new Roadster will actually do something like this," he wrote. Tech YouTuber Marques Brownlee responded: "The thing is I feel like you’re not joking."

"I'm not," Musk said.

He then described a SpaceX-made cold-gas thruster system that uses high-pressure air stored in a composite pressure vessel.

At a SpaceX test site in Texas on Thursday, almost 8 years later, a Roadster climbed a curved ramp, dropped, rolled through the air and landed on its wheels. It used the same basic technology: cold-gas thrusters fed by high-pressure air stored in composite pressure vessels.

The 2019 exchange was not Musk’s first mention of thrusters. In June 2018, he said the Roadster would include roughly 10 of them and suggested they might allow the car to fly.

When another user proposed compressed-air cold-gas thrusters, Musk said that configuration made sense and described a pump to replenish the air.

Musk's account of the technology has proved more consistent than his production timetable. Tesla unveiled plans for the second-generation Roadster in 2017 and originally targeted production for 2020. After repeated delays, Musk said in the reveal that production will begin in 2028.

The Roadster is not Tesla’s only vehicle to miss an ambitious release target. When Musk unveiled the Cybertruck in 2019, he said Tesla expected production to begin in late 2021. The first customer deliveries took place two years later, in November 2023.

While Musk has missed plenty of deadlines, many of his more improbable technology plans have made it beyond the drawing board. In December 2020, he said SpaceX would try to catch a returning Super Heavy booster with its launch tower.

Nearly four years later, the tower’s mechanical "chopsticks" caught one for the first time.

Musk has a habit of promising the improbable well before it is ready, but eventually, the improbable part arrives.

• This post originally appeared in the Business Insider Today newsletter.
• You can sign up for Business Insider's daily newsletter here.

The future of AI might rely on getting your Slack DMs.

It's not just your DMs, to be clear. It's internal memos, transaction records, contracts, and even that onboarding guide you probably didn't read when you started your job.

Basically, if it's work-related and has been documented, AI companies are interested. The rationale is straightforward: If AI tools are going to automate work, they need a deep understanding of what the work is.

BI did a deep dive on AI companies' race to scoop up as much data as possible in its series "The AI Data Grab."

BI's Charles Rollet, Tom Carter, and Stephen Council first unpacked the wild market that's sprung up for this data. Things move quickly, including "exploding offers."

Others are figuring out how to leverage their databases. Law firms, which often have detailed archives, are a great example. BI's Melia Robinson has more on that.

And sometimes it's about actually creating the data yourself. That's the case in robotics, writes BI's Rya Jetha. Companies are shelling out big to have people record everyday tasks like cooking and cleaning.

AI's data push could renew interest in a space that's already navigated an info-hungry industry.

Alternative data, unstructured datasets such as satellite imagery and card transactions, became a hot commodity on Wall Street in the mid-2010s.

These non-traditional datasets are supposed to give investors an edge. Want to know if Home Depot will beat earnings estimates? Analyze satellite images of its parking lots to see whether foot traffic is rising or declining.

When I covered the space closely with my colleague Bradley Saacks, the data providers often highlighted that the real market opportunity was selling into big corporates.

Hedge funds might be rich, but there are only so many of them to go around. The market also has a ceiling. If too many people are trading off the same datasets, you lose your potential edge.

AI's data needs can fill the gap alt-data providers hoped corporates would. And unlike hedge funds interested in real-time, market-relevant data, these firms benefit from a wider, deeper catalog.

Some players already see an opportunity. When former Point72 data wizard Kirk McKeown launched a data startup last year, he told Bradley it aimed to package data in an easily accessible way for AI models to scoop up.

It's promising, but there are no guarantees. Alt data faced real growing pains as providers struggled to scale their offerings while maintaining data quality. AI companies' needs might differ, but their expectations won't be any easier. And in some cases, money managers might find the most useful data in their own backyard.

In the meantime, follow Bradley here for more coverage on how alt-data providers are navigating the AI era. And keep an eye out for a hedge-fund-focused newsletter from him later this year.

• Microsoft President Brad Smith warned that US immigration policies could drive AI talent overseas.
• "Let's not put that at risk," Smith said.
• The Trump administration recently suspended Microsoft and other tech firms from a green card program.

If America wants to win the AI race, sending the world's brightest engineers packing might not be a winning strategy.

That's the warning from Microsoft President Brad Smith, who said the Trump administration's immigration crackdown risks driving the next generation of AI talent away from the United States.

"I think we should all be worried about sending intentionally or inadvertently or otherwise, signals that send the most talented next wave of AI software engineers to other countries rather than the United States," Smith said in an episode of Politico's "The Decoded Podcast" that aired on Thursday.

His comments come after the Trump administration announced last week that it was suspending Microsoft, Adobe, and several other technology companies from a federal program that sponsors foreign workers for permanent residency.

Vice President JD Vance accused Microsoft of abusing the Permanent Labor Certification, or PERM, program by laying off American employees while hiring lower-paid foreign workers.

Smith, who Business Insider reported was recently put in charge of Microsoft's communications, pushed back against those allegations, saying the administration had presented a misleading picture of Microsoft's hiring practices.

"What was presented as Microsoft replacing 6,000 people who lost their jobs with 9,000 foreigners was frankly completely off base," he said.

Smith said most of the company's visa and green card applications concerned existing employees rather than new hires, adding that Microsoft had hired 9,200 American workers in its last fiscal year.

"That was left out of the narrative," he said.

He also rejected suggestions that Microsoft pays foreign workers less than their American counterparts, saying employees are compensated under the same system regardless of immigration status.

The American dream, on hold

The dispute comes amid growing uncertainty for skilled foreign workers in the US.

As part of our ongoing H-1B Fallout series, Business Insider previously reported that the Trump administration's shifting immigration policies have prompted some international workers to explore opportunities abroad, including in countries like the UK, Germany, and Canada.

An analysis of federal data by Business Insider found that certified H-1B applications fell 10% in the first nine months of fiscal year 2026 compared with the same period a year earlier. Applications for data scientists, however, rose 22.5%, highlighting continued demand for specialized AI talent.

On the Politico podcast, Smith cited immigrants' economic contributions as evidence of what America stands to lose.

He said that five of the six technology and science medal recipients recently honored by President Donald Trump had immigrated to the United States.

According to Smith, the companies associated with those five individuals collectively represent more than a quarter of the S&P 500's market value.

Their contributions, he said, benefit not just technology employees but ordinary Americans whose retirement savings and pension funds are invested in those businesses.

"Think about what those five individuals have done for the well being, even the wealth and standard of living, of Americans across the country," Smith said.

"Let's not put that at risk."

Reliance Industries appeared to hit back at Elon Musk’s suggestions its owner Mukesh Ambani (pictured) was blocking Starlink’s entry into India, stating it welcomes competition and pointing out its own satellite venture has also not been cleared by regulators.

In a statement on Musk-owned X, Reliance said considering recent remarks about the company and the Indian market, it wished to put on record that like every Indian enterprise it had grown in a country that welcomes investment and innovation.

It continued to state the open nature of the Indian market rests on the rule of law, and the government has stated it applies the same process to every applicant, Indian or foreign.

“Jio’s own satellite service awaits clearances,” it said. “At Reliance we never feared competition… we welcome competition in satellite communications, including Starlink. Every player, Jio included, must serve the people of India and abide by the laws and policies of the land.”

Reliance’s statement followed a series of Musk’s own posts regarding Starlink’s entrance into India. The SpaceX chief claimed “certain oligarchs” were delaying its entry into the country to maintain a “monopolistic chokehold on the Indian people”.

While emphasising the benefits satellite connectivity will have on the Indian market, Musk sarcastically addressed Ambani as India’s real Prime Minister and asked if he would “consider allowing Starlink to compete”.

Musk also described suggestions Starlink would struggle to compete with Jio and its low-cost data plans as “absurd, hypocrisy and hyperbole”.

India’s government has also rejected Musk’s claims, stating its satellite communications framework is fair and non-discriminatory while emphasising that Starlink, Jio and Sunil Bharti Mittal-backed Eutelsat OneWeb are all awaiting security clearance and spectrum allocation.

The post Reliance snaps back at Musk over Ambani claims appeared first on Mobile World Live.

• A24's new documentary about Elizabeth Holmes, "You Can See Everything," is in theaters now.
• Directed by Nathan Fielder and Lance Oppenheim, it's more character study than straightforward doc.
• The story about the downfall of Theranos has been told — but this one hasn't yet.

Elizabeth Holmes famously does not blink, whether she's pitching investors or, by her own admission, flirting with a guy.

Is her moony gaze haunting or hypnotic? A calculated act or proof of her naivete? At various points in A24's new documentary, "You Can See Everything," I vacillated between each of the above.

A good portion of Nathan Fielder and Lance Oppenheim's movie could be classified as a provocative staring contest, as Fielder probes Holmes about her life and the demise of her company, Theranos, in the weeks before she reported to prison to serve an 11-year sentence in May 2023.

The answers he gets from Holmes are rarely satisfactory. (At one point, she insists she's never told a lie in her entire adult life; despite being convicted of fraud, she maintains her innocence throughout the film.) But with the camera zoomed in tight on Holmes' every micro-expression and clump of mascara, the movie's power is in the questions it raises in the space between her words.

In these profoundly uncanny scenes, I found myself fluctuating between being freaked out by the lack of emotion in Holmes' vacuous gaze and, to my surprise, pitying her. Then I'd quickly shift back to being freaked out, because what if pity was exactly the reaction that Holmes — a talented, convicted fraudster — was angling for? If she could manipulate her voice to sound more authoritative in the workplace, she could certainly adapt her face to appear more sympathetic onscreen.

It turns out that just because you can see everything, that doesn't mean you know anything. At least not for sure.

'You Can See Everything' is not the movie it appears to be

Much like its subject, "You Can See Everything" is a slippery mind-trick of a film — part character study, part psychological thriller, part surrealist thought experiment. What it is not, despite its subject, is a movie about a tech company.

The story of how Holmes went from being named the youngest self-made female billionaire by Forbes to being convicted of defrauding investors as the founder and CEO of the blood-testing startup Theranos is well documented. Holmes has already been the subject of a more traditional documentary, Alex Gibney's "The Inventor: Out for Blood in Silicon Valley," which exhaustively maps the discrepancies within Theranos, and a scripted Hulu miniseries, "The Dropout," which dramatizes Holmes' origin story and attempts to illustrate how her ambitions outpaced her abilities.

Both of those works lean into the prevailing narrative about Holmes, who was often pitched as "the next Steve Jobs" and proved especially skilled at wooing powerful men to her cause.

"You want it to be true so badly," Tyler Shultz, a former Theranos employee-turned-whistleblower, says of Holmes' vision in Gibney's doc. "I was working with these devices every single day, and she could still kind of convince me. When I think back on those conversations, I just think, 'How did she do that?'"

The Holmes we see in "You Can See Everything" is tamer, gentler, and, at times, far more submissive than many headlines would have us believe. Yet even Fielder isn't immune to her strange charisma.

Early in the film, Fielder admits to Holmes that he did a lot of research about her tenure at Theranos. He challenges her answers to questions, at one point telling Holmes, "I don't think anyone will ever believe you ever again."

Yet, as time goes on, Fielder seems alternately bemused and unnerved by the onetime wunderkind — though never to the point of aversion. On the contrary, Holmes seems to draw Fielder ever closer, like a close-up magician with a deck of cards in her hands and a trick up her sleeve.

Toward the end of the film, Fielder's memory is called into question. A friend asks if Fielder is retroactively polishing his interactions with Holmes, perhaps unconsciously, because he's grown to care about her. Maybe he wants to protect her.

Seemingly taken aback, Fielder replies, "I'm not putting my feelings into this. I'm just trying to be accurate."

It's one thing to be told someone is a master manipulator, as "The Inventor" and "The Dropout" plainly do. It's quite another to experience the manipulation firsthand. "You Can See Everything" offers audiences the latter.

During a Q&A after an early screening of "You Can See Everything" in New York, Oppenheim described how he, Fielder, and the crew would swap interpretations of what they'd just seen and heard after the cameras were switched off. "The disagreements were so dynamic," Oppenheim recalled.

It's been over three years since their close encounter of the Holmesian kind, and Fielder said he's still trying to process what he felt in her presence.

"I'm really curious to hear what everyone else thinks, and what they make of certain things," Fielder said. "I think it will actually help me understand."

"You Can See Everything" is in theaters now.