• Elon Musk’s AI company xAI plans to allow Cursor to use some of its compute power for training.
• Cursor will train its AI model Composer 2.5 using tens of thousands of xAI GPUs, sources said.
• The arrangement marks a new strategy for xAI in a competitive AI landscape.

Elon Musk’s AI company, xAI, plans to put its stockpile of computing power to use in a new arrangement with coding startup Cursor, according to people familiar with the matter.

Cursor plans to train its latest AI coding model, Composer 2.5, on xAI infrastructure, the people said. Cursor will use tens of thousands of xAI’s graphic processing units (GPUs), the chips used to train AI models, they said.

The setup effectively turns xAI into a kind of cloud provider. By renting some of its GPUs to other companies, xAI could start generating revenue from its massive infrastructure while still developing its own AI models. The arrangement could help the company offset the costs of building and operating data centers, while also deepening ties with a startup that has access to valuable coding data.

Amazon, Microsoft, and Google, the largest cloud providers, own millions of chips and rent computing power out to thousands of companies and developers, generating huge profits. Newer players like CoreWeave and Lambda have built businesses around supplying GPUs to AI model developers. Access to computing power has become an increasingly competitive aspect of the AI arms race.

Representatives for xAI and Cursor did not respond to a request for comment.

It’s not the first time Cursor and xAI have overlapped. The startup hired two former Cursor product engineering leads in March, Andrew Milich and Jason Ginsburg. Ginsburg and Milich oversee xAI’s product team and report directly to Musk and xAI president Michael Nicolls, Business Insider previously reported.

xAI is one of many companies racing to build the best AI models, and it has one of the largest data center footprints. Musk said during an all-hands last December that xAI would beat competitors like OpenAI and Anthropic because it would have access to more power to train its models.

Over the past two years, xAI has rapidly expanded the footprint of its data centers, a project it has named Colossus. Last year, the company said it had around 200,000 Nvidia GPUs, and Musk has said it plans to expand to 1 million GPUs.

xAI’s infrastructure team has been experiencing a leadership shake-up. It lost its infrastructure lead, Heinrich Küttler, last week. The company moved Jake Palmer into a leadership role over the physical infrastructure team, and SpaceX’s Daniel Dueri took a leadership position over the compute infrastructure team last week, Business Insider previously reported.

In a memo to staff last week, Nicolls, xAI’s president, said the company’s model FLOPs Utilization (MFUs), a measure of how efficiently a GPU is used during AI training, was “embarrassingly low” at about 11%. Nicolls said he aims for the team to reach 50% in the next few months. For comparison, according to AI infrastructure company Lambda AI, most large-scale AI training operates between 35% to 45% MFU.

Cursor is in talks for a reported valuation of around $50 billion, Bloomberg reported last month. Meanwhile, it faces pressure as major AI startups like Anthropic and OpenAI push aggressively into building coding assistants.

In March, Cursor released Composer 2, a coding model designed to generate and edit code across large projects. Cursor built the model on top of an open-source AI model from Chinese startup Moonshot AI and fine-tuned it using its own data from its developer user base.

Do you work at xAI or have a tip? Contact this reporter via email at gkay@businessinsider.com or Signal at 248-894-6012. Use a personal email address, a non-work device, and non-work WiFi; here’s our guide to sharing information securely.

• Helping my dad plan for Medicaid showed me I had no estate plan of my own.
• I learned how easily a home and assets can be lost without proper preparation.
• I set up a will and trust to protect my family and avoid future complications.

I sat in a stiff chair next to my 86-year-old dad as a lawyer outlined the complicated rules of Medicaid law.

My dad furrowed his brow, trying to understand the legal complexities that could mean the difference between being able to leave his home — where he’d lived for nearly 60 years — for my sisters and me, rather than having it confiscated as collateral should he die in a long-term care facility.

A few months earlier, my dad — who was recently diagnosed with early-stage dementia — called me with an urgent request. “I want you to make sure they don’t take my house,” he said, obviously coming to terms with the possibility that extended memory care could be in his future.

My dad has no savings to cover long-term care

My dad grew up one of nine kids raised during the lean years of the Great Depression and World War II. He knew the value of owning a home, both monetarily and in the sense of pride he derived from knowing he’d at least leave that small piece of property to my sisters and me after he died.

But like many seniors, he lives on a fixed income and doesn’t have savings to cover the often astronomical expense of long-term care facilities, even with Medicare. Should he need long-term care — a possibility despite his age since he’s in great physical shape and had a mother who lived well into her 90s — the ownership of his home could be in jeopardy.

According to federal and many state laws, Medicaid is required to pursue “estate recovery” after a recipient dies, often placing liens on the decedent’s home or estate to recover care expenses. There are several ways to avoid this, including setting up trusts and a form of property co-ownership called a life estate. Ultimately, our lawyer advised we take advantage of a caretaker clause in Medicaid law, which would apply to my sister, who lives with our dad.

As we went through all these scenarios, while also helping my dad set up his will, healthcare power of attorney (who makes healthcare decisions on his behalf if he’s incapacitated), and durable power of attorney (who can make financial decisions on his behalf), I couldn’t help thinking about my own family.

It made me realize I didn’t have a will

My husband and I did not have a will, any power of attorney, or even beneficiaries on our bank accounts. We’re both cruising toward 50, and we have a young son, so I knew we needed to address our own estate planning.

Our bank offers free estate planning consultations, so I booked an appointment for us to go through the basics before speaking with an attorney who will charge for their services.

Though we’ve been married almost 15 years and faced some health crises, such as my breast cancer diagnosis nearly a decade ago, my husband and I had never discussed what would happen if one or both of us died. While these conversations are hard, we knew they were necessary to ensure our son is cared for physically and financially after we’re gone.

In our case, if one spouse died, the other would automatically inherit any assets and assume sole guardianship of our son. But if both of us were to die at the same time, things get more complicated.

Since we have an underage child, we set up a trust where any of our assets — our home, retirement accounts, etc. — would go to benefit our child once he reaches adulthood. Then we had to choose who would not only take guardianship of our son but also manage our estate, taking care of everything from ensuring our final tax returns were filed to managing the money left to our son.

Talking about death can be hard

Through these conversations, we learned that not having specific, legally verified instructions for your assets — even if it’s just a checking account or a car — can cause bureaucratic headaches for grieving loved ones. And not establishing specific instructions via a healthcare power of attorney — such as whether you want to be resuscitated or put on life support — can put family in the difficult position of making life-or-death choices in the midst of a crisis.

I know that we’ve done all we can to ensure my dad is taken care of for the remainder of his life, and that everything he has worked so hard for will go where he wants it. And I know my husband, son, or sisters won’t have to shoulder the burden of sorting out my affairs when I die.

Talking about my death, as well as the demise of my father and husband, was hard. No one wants to have these difficult conversations. But knowing that we’ve done the unpleasant work to make sure our wishes are respected and our loved ones are protected upon our deaths is well worth the discomfort.

Automotive IoT has moved from a niche capability embedded in premium vehicles to a foundational layer of modern mobility systems. As vehicles become increasingly connected, they generate and consume large volumes of data that influence everything from safety systems to fleet operations and customer experiences. This evolution is reshaping how vehicles are designed, operated, and monetized across the automotive and transportation industries.

For IoT decision makers and technology leaders, understanding Automotive IoT is no longer optional. It sits at the intersection of connectivity, cloud platforms, edge computing, and software-defined architectures. The convergence of these domains is driving a transition toward connected vehicles, advanced telematics, and software-defined mobility ecosystems.

Key Takeaways

• Automotive IoT connects vehicles to cloud, edge, and infrastructure systems, enabling real-time data exchange and control.
• Telematics platforms are central to fleet management, predictive maintenance, and usage-based services.
• Software-defined vehicle architectures decouple hardware and software, enabling continuous feature updates.
• Cellular connectivity, edge computing, and standardized protocols are key enablers of scalable deployments.
• Security, data governance, and system complexity remain major challenges in Automotive IoT ecosystems.

What is Automotive IoT?

Automotive IoT refers to the integration of connected sensors, embedded systems, and communication technologies within vehicles to enable data exchange with external systems such as cloud platforms, infrastructure, and other vehicles. It forms the backbone of connected vehicles, telematics solutions, and software-defined mobility services.

Within the broader IoT ecosystem, Automotive IoT acts as a mobile and distributed data platform. Vehicles are no longer isolated mechanical systems; they are networked endpoints capable of sensing, processing, and transmitting data in real time. This capability supports applications ranging from fleet optimization to advanced driver assistance and over-the-air software updates.

How Automotive IoT works

Automotive IoT systems rely on a layered architecture that combines in-vehicle hardware, connectivity networks, edge processing, and cloud-based platforms. At the core are embedded electronic control units (ECUs) and sensors that collect data related to vehicle performance, location, environment, and driver behavior.

Data generated within the vehicle is transmitted through a telematics control unit (TCU), which acts as a gateway between the vehicle and external networks. The TCU manages communication via cellular technologies such as LTE, LTE-M, NB-IoT, and increasingly 5G. In some cases, short-range communication technologies such as Wi-Fi or Bluetooth are also used for specific use cases.

Once transmitted, data is processed either at the edge or in the cloud. Edge computing capabilities within the vehicle or nearby infrastructure allow for low-latency decision-making, such as collision avoidance or real-time diagnostics. Cloud platforms, on the other hand, handle large-scale data aggregation, analytics, machine learning, and application orchestration.

The rise of software-defined vehicles introduces a new abstraction layer where software components are decoupled from hardware. This enables continuous updates, remote configuration, and the deployment of new services throughout the vehicle lifecycle.

Key technologies and standards

Automotive IoT depends on a combination of communication technologies, software frameworks, and hardware components. Key technologies include:

• Cellular connectivity: LTE, LTE-M, NB-IoT, and 5G provide wide-area communication for connected vehicles.
• Vehicle-to-Everything (V2X): Enables communication between vehicles, infrastructure, pedestrians, and networks.
• CAN, LIN, and Ethernet: In-vehicle communication protocols connecting sensors and control units.
• Telematics platforms: Systems that collect, process, and analyze vehicle data for fleet and operational insights.
• Edge computing: Local data processing to reduce latency and bandwidth usage.
• Over-the-Air (OTA) updates: Mechanisms to remotely update software and firmware.
• Cloud IoT platforms: Infrastructure for data storage, analytics, and application management.

Standards and industry initiatives also play a key role in ensuring interoperability and scalability across Automotive IoT deployments. These include automotive-grade Linux, AUTOSAR frameworks, and emerging standards for V2X communication.

Main IoT use cases

Automotive IoT enables a wide range of applications across multiple industries, extending beyond traditional passenger vehicles.

• Fleet management: Real-time tracking, route optimization, fuel efficiency monitoring, and driver behavior analysis.
• Predictive maintenance: Continuous monitoring of vehicle components to detect failures before they occur.
• Usage-based insurance: Insurance models based on driving behavior and vehicle usage data.
• Smart logistics: Integration with supply chain systems to track goods and optimize delivery operations.
• Connected public transport: Monitoring and optimization of buses, trains, and shared mobility services.
• Autonomous and assisted driving: Data exchange supporting advanced driver assistance systems (ADAS).
• Energy and EV management: Monitoring battery performance, charging infrastructure, and energy consumption.

These use cases demonstrate how Automotive IoT extends into industrial IoT, smart city infrastructure, and energy systems, creating interconnected mobility ecosystems.

Benefits and limitations

Automotive IoT delivers significant operational and strategic benefits. It improves visibility into vehicle performance, enhances safety through real-time monitoring, and enables new business models based on data-driven services.

• Operational efficiency: Optimized routing, reduced downtime, and better resource utilization.
• Enhanced safety: Real-time alerts, remote diagnostics, and driver assistance features.
• New revenue streams: Subscription services, data monetization, and mobility-as-a-service models.
• Lifecycle management: Continuous software updates and remote maintenance capabilities.

However, several constraints and challenges remain:

• Connectivity limitations: Coverage gaps and variable network performance can impact reliability.
• Latency requirements: Critical applications require ultra-low latency that not all networks can guarantee.
• Security risks: Connected vehicles expand the attack surface for cyber threats.
• System complexity: Integration of hardware, software, and networks increases development and maintenance complexity.
• Data governance: Managing ownership, privacy, and compliance across jurisdictions is challenging.

Balancing these benefits and limitations is a key consideration for organizations deploying Automotive IoT solutions at scale.

Market landscape and ecosystem

The Automotive IoT ecosystem is composed of multiple stakeholders, each contributing to different layers of the value chain.

• Automotive OEMs: Integrate connectivity and software capabilities into vehicles.
• Tier 1 suppliers: Provide hardware components such as sensors, ECUs, and telematics units.
• Connectivity providers: Mobile network operators and MVNOs enabling global connectivity.
• Cloud and platform vendors: Offer infrastructure for data processing and application development.
• Software providers: Develop operating systems, middleware, and application frameworks.
• System integrators: Combine technologies into end-to-end solutions for enterprises.

The shift toward software-defined mobility is also changing competitive dynamics. Traditional automotive players are increasingly collaborating with technology companies, while new entrants focus on software platforms and data services.

Future outlook

The evolution of Automotive IoT is closely tied to advances in connectivity, computing, and software architectures. The rollout of 5G and future 6G networks is expected to enable higher data throughput and lower latency, supporting more advanced use cases such as cooperative driving and real-time vehicle coordination.

Software-defined vehicles will continue to gain traction, enabling continuous feature deployment and reducing dependency on hardware upgrades. This shift will likely accelerate the adoption of subscription-based services and new revenue models.

Edge computing will also play a larger role, particularly for applications requiring immediate decision-making. At the same time, increasing regulatory scrutiny around data privacy and cybersecurity will shape how Automotive IoT systems are designed and operated.

Overall, Automotive IoT is moving toward a more integrated and platform-driven model, where vehicles are part of a broader digital ecosystem spanning transportation, energy, and urban infrastructure.

Frequently Asked Questions

What is Automotive IoT?

Automotive IoT refers to the use of connected sensors, communication technologies, and software platforms in vehicles to enable data exchange with external systems such as cloud services and infrastructure.

What is a connected vehicle?

A connected vehicle is a vehicle equipped with internet connectivity and communication systems that allow it to send and receive data in real time.

What is telematics in Automotive IoT?

Telematics is the technology that combines telecommunications and data analytics to monitor vehicle location, performance, and usage.

What is a software-defined vehicle?

A software-defined vehicle is a vehicle where functionality is primarily controlled and updated through software rather than fixed hardware configurations.

What connectivity technologies are used in Automotive IoT?

Automotive IoT commonly uses cellular technologies such as LTE and 5G, as well as short-range communication technologies like Wi-Fi and Bluetooth.

What are the main challenges of Automotive IoT?

Key challenges include connectivity reliability, cybersecurity risks, system complexity, and data privacy management.

Related IoT topics

• Edge Computing for IoT
• 5G and Cellular IoT Connectivity
• Industrial IoT (IIoT)
• IoT Security and Device Authentication
• Digital Twins in IoT
• Smart Cities and Urban IoT Infrastructure

The post Automotive IoT: Connected Vehicles, Telematics and Software-Defined Mobility appeared first on IoT Business News.

• LinkedIn CEO Ryan Roslansky said “soft skills” are more important as AI takes over routine work.
• He highlighted four communication-specific skills to hone.
• Other Silicon Valley leaders have made more dramatic predictions about AI’s impact on jobs.

Personality hires, rejoice. LinkedIn’s CEO said soft skills are getting a hard rebrand.

In an interview with the “Tools and Weapons” podcast, LinkedIn CEO Ryan Roslansky said AI is automating routine tasks. He argued that the shift is elevating four human-centered skills — curiosity, courage, communication, and compassion.

“These turn out to be some really, really important skills to do your job well,” the CEO said. “The focus and emphasis on those, along with the AI, is what I think gives us the opportunity to dream big and paint a much more positive picture that exists with humans and technology together moving forward.”

Roslansky has a broad view of the job market based on LinkedIn data, though he did not cite specific figures in the interview. His view contrasts with some other Silicon Valley voices (and a growing number of Americans).

This year, OpenAI investor Vinod Khosla predicted that today’s five-year-olds won’t need a job. Boris Cherny, the creator of Anthropic’s Claude Code, said the job title “software engineer” will fade away this year. Nvidia CEO Jensen Huang recently said he wants high-paid engineers to spend half their salary on AI tokens.

Roslansky’s vision is less jarring. Instead, he argues that AI is reshaping how people think about their jobs, encouraging workers to view their roles as a “collection of tasks” rather than a fixed title.

He breaks those tasks into three buckets: ones AI can fully automate, ones it can augment, and ones that remain deeply human — like resolving conflict, persuading a team, or setting strategy.

“These skills, they’re important, but they’ve historically been talked about as soft skills,” Roslansky said. “In a professional world where people are actually much better at these skills and have really honed their craft on it, I think that it makes things a lot better.”

As AI handles more automated responsibilities, Roslansky said that agents can free up time for coworker conversations, putting a greater premium on communication, judgment, and emotional intelligence.

He said his thought process has given him a hopeful view of AI’s future. Still, he said he doesn’t have a crystal ball and could ultimately be wrong.

“Sometimes when you’re mired in the technology, and especially with AI, and you kind of draw out where this could potentially go, it leads you to some dark places,” he said. “I believe that humans play such an integral role in shaping where that technology should go.”

• James McMillian is the president of Tone House, a strength training facility in New York City.
• McMillian shared a few tips for building muscle fast, especially if you don’t have a lot of time.
• Compound exercises work multiple muscle groups simultaneously, while explosive movements develop power and speed.

Since childhood, James McMillian has learned the value of slowly getting stronger.

McMillian said he was diagnosed with asthma when he was born, and doctors advised his parents against letting him play sports like youth tackle football. “It made me tell myself that I need to train harder,” McMillian told Business Insider.

Before and after practice, McMillian recalled running up the stadium steps six times. He’d also run to practice or run home with his sports equipment. Even though his asthma wasn’t induced by sports, he noticed that the extra training kept him from being as winded.

McMillian is now the president and a trainer at Tone House, a strength training and conditioning-focused training facility deemed by Business Insider as “the hardest workout in New York City.” He applies that same mentality from his upbringing to Tone House workouts that he coaches.

“You’re going to be exposed for your weaknesses, but then you should always turn your weaknesses into your strengths,” he said.

McMillian believes progressively overloading — slowly adding weight, reps, or intensity — is the best way to build strength.

“I always tell people, ‘If you don’t use it, you’ll lose it,” he said, particularly since we lose muscle mass as we age. “You’re going to want to pick up your kid. You’re going to want to hold those groceries. You’re going to want to hold your Louis Vuitton bag a little bit longer at an event. You have to be strong to do that.”

McMillian shared a few of his favorite quick movements to build muscle, whether you’re a complete beginner, pressed for time at the gym, or wanting to speed up your progress.

Raise a glass to goblet squats

McMillian’s favorite movement for beginners is the goblet squat, or squatting while holding a heavy kettlebell or dumbbell.

“When you’re going into that squat, as you allow those knees to go out, you’re working that lower body, your glutes, your hamstrings,” he said. At the same time, you’re also engaging your lats, triceps, forearms, and upper back from holding the weight.

To make the goblet squat more challenging, he said, you can elevate your heels to focus on building quad strength. McMillian also said it’s easy to start with a lighter weight, say 10 pounds, and gradually work your way up.

Save time with compound exercises

Some of the best time-savers are compound exercises, which target multiple parts of the body at once and “allow you to shorten your time at the gym,” McMillian said.

A favorite of McMillian’s is transitioning from a squat to an overhead press, thus targeting both the upper and lower body.

He also likes to string together multiple consecutive movements. He opts for two rounds of 20 shoulder presses, 20 hammer curls, and 20 bent-over rows, all right after each other to activate different parts of his upper body.

“It’s intense, it burns, but it’s something that people can do with a single dumbbell,” he said, thus also saving time on swapping out equipment.

It’s also customizable, he said: You can use lighter weights and do fewer reps, but increase the number of rounds, for example.

Play with tempo

Explosive exercises — deliberately slowing down your tempo and “exploding” at the end of the movement — are also great for improving your strength, McMillian said.

He’s a fan of slowing down his back squats: lowering slowly for five seconds, pausing for another five at the bottom, and then quickly pushing up.

“When you do things a little bit slower, controlled, it builds strength,” he said, noting there’s more than one way to challenge yourself.

• Having children was never high on my priority list.
• I went through some difficult times in my 30s. I’m glad I didn’t have the added stress of kids.
• I’ve chosen to focus on my ambitions, mental health, and other aspects of my fulfilling life.

I always felt like I disappointed my mother. Her lifelong wish was to become a grandmother, but having children was never high on my priority list.

When she died unexpectedly, I was 35 and struggling financially and mentally.

Her death threw my life into even more chaos — and not being a mother helped me survive it.

As I approach 40, I get asked, “Why don’t you just have a baby?” more and more often.

Although well-meaning, the question reduces a complex, private decision to small talk — as though my ambitions, finances, mental health, and grief matter less than my reproductive function.

What I witnessed growing up shaped my decision not to have children

Growing up, I watched my mother work six days a week running her esthetician business while doing all of the cooking, cleaning, and childcare at home. She was so busy being a working wife and mother to two kids that she put off many of her dreams until retirement.

She loved her work and continued tinting eyelashes and applying gel nails for her clients until the pandemic forced her to retire at 68. The prospect of enjoying her retirement kept her going through the UK’s lockdowns — but she died just a few months after restrictions were lifted.

Watching her sacrifice so much time — only to run out of it — shaped how I think about my own life.

Career and financial instability made survival my priority

At 32, I left a stable career in humanitarian aid to pursue my creative ambitions. It didn’t go well at first, but by my mid-30s, I felt like I was finally getting somewhere with my fledgling freelance writing business.

My partner was also out of work at the time, which only intensified the pressure. When we went shopping at our local market, we bought discounted bags of vegetables, and I crossed my fingers that my card payment would go through.

Every day, I struggled with the shame and sometimes-crippling anxiety and depression brought on by prolonged financial hardship.

At a time when many of our friends were having babies, we were simply trying to stay afloat. And since my partner (who’s 17 years my senior) has never wanted children, they weren’t in our plans.

Grief put my life on hold for five years

Then, in October 2021, my world fell apart. My mother and I were close, and I was unprepared for the heartbreak of losing her so suddenly — especially as the pandemic had kept us apart for so long.

Grief consumed me, making it difficult to work, and I mostly lived off the little savings I had scraped together.

Within a year, I landed a remote role at a marketing agency that gave me the stability I needed to start rebuilding my life. When I was made redundant at the end of 2024, not having a child to support gave me time to think about what to do next without additional pressure.

Now, nearly five years after my mother died, I finally have the strength to pursue the dreams I started chasing years ago.

I’m building a life that feels right for me

Last year, my nephew was born — the grandchild my mother never got to meet.

I can’t deny that watching my brother become a father made me wonder what parenthood might be like. But while I adore my nephew, loving him doesn’t make me yearn for the upheaval motherhood would bring. And without my mom to share in that chapter of life, I simply don’t feel the pull.

As I enter my 40s, I don’t feel like I’m “missing out”. Being child-free helped me survive the darkest time of my life — and gave me the chance to rebuild on my own terms.