 |
|
|
|||||
SpaceX’s Starlink orbital internet satellites are falling out of low earth orbit at an increasingly alarming rate, with one to two satellites now reentering Earths atmosphere every single day. According to Harvard-Smithsonian Center astrophysicist Jonathan McDowell, that number will only go up as more satellites end their useful lifetime and the low earth orbit (LEO) constellation numbers skyrocket. This is as much a design problem as anything. While the numbers vary, right now there are around 10,200 active satellites in low earth orbit. Of those, about 8,475 are Starlinks. In other words, about 80% of all those satellites belong to Elon Musk’s company. By 2030, the European Space Agency expects the number of satellites in LEO will increase to about 100,000. This is mostly thanks to SpaceXwhich plans, pending regulatory approval, to expand its fleet to a staggering 42,000 satellitesbut also to Starlink-clone networks like Jeff Bezos’s 3,200-satellite Kuiper, and the Chinese GuoWang and Qianfan, which plan to launch a total of 18,000 units combined. Designed to fall Each Starlink satellite has a five-year lifespan. They zip across the sky in low earth orbit. There, objects still feel about 95% of the gravity we experience on the ground. What keeps them from plummeting is their sideways velocity of 17,000-plus mph. These vehicles are essentially falling around the Earth, inches at a time. But even at that altitude, the thin atmosphere creates drag, with air particles hitting and slowing the satellites down. To compensate, they fire up their onboard krypton and argon thrusters, which lift them up to maintain their orbital path. When the fuel runs out, the satellite can no longer boost, its orbit decays, and it comes crashing down. Before that time comes, SpaceX de-orbits the satellites on a controlled crash, allowing them to aim at an empty stretch of ocean as opposed to making a random entry. Why this is a problem As McDowell explainsand SpaceX itself admitssome satellites will not disintegrate upon reentry, though they are designed to do so. They [Starlink satellites] are designed to completely burn up, McDowell said in a recent interview with EarthSky. Now we’re not sure we really believe that they really burn up, but at least for the most part they melt. There have been many other incidents of space objects falling to Earth, including big chunks of space stations like the American Skylab and the Soviet Salyut 7; parts of rockets like a European Ariane 5 nose; satellites like the Russian Kosmos 2251 (which collided with an Iridium communications satellite); and even the trunk of SpaceX’s very own Crew-9 Dragon spaceship. But thanks to the extraordinary number of units deployed, the Starlink constellation represents an outsize concern to everyone on Earthand also to other satellites in low earth orbit. If one of Musk’s satellites crashes against another satellite, it could start a chain reaction called the Kessler Syndrome, which you can see in action in Alfonso Cuaróns film Gravity. This is the nightmare of runaway debris collisions devastating all low earth orbit satellites. A single crash could create cascading debris fields, wiping out the infrastructure of global GPS, communication, financial systems, and weather monitoring. Worst-case scenario, it could plunge civilization into chaos. Right now, SpaceX is essentially launching bullets into an orbital game of Russian roulette. With thousands of Starlink satellites circling the globe, McDowell says that the current de-orbit rate is just the beginning. As the first generation of Starlinks reach their five-year expiration date, we are seeing four or five satellites per day being intentionally plunged back to Earth. This number is set to multiply as more and more Starlinks get to their end of life. As the constellation grows into the tens of thousands, we risk turning our upper atmosphere into a perpetual fireworks show of burning toxic metal that sometimes crashes into Earth. Designed for full demise Back in July 2024, Musks company assured regulators and the public that its satellites were designed for “full demise,” claiming they would vaporize into harmless dust. That turned out to be fantasy: Eight months later, New Scientist revealed that a 2.5-kilogram chunk of aluminum from a Starlink satellite slammed into a farm in Saskatchewan, Canada. SpaceX was forced to admit that this piecea modem enclosure lidwas supposed to have vanished completely but didn’t. Musks safety guarantees were proven wrong by a 5-pound piece of metal lying in a farmer’s field. SpaceX claims this has happened only once with a satellite that was part of a failed launch. However, in January 2025 a new fireball crossed Chicago’s sky. As he posted on X at the time, McDowell believes this was Starlink-5693. In response to McDowells post, Michael Nicolls, VP of Starlink engineering at SpaceX,said it was an uncontrolled reentry caused by a faulty component. The worrying bit of his explanation: There is still work to do to guarantee this, especially for satellites with degraded attitude control. But as you noted, the sats nearly completely demise upon reentry.” [my emphasis] Too much junk SpaceX now claims it uses a “belt-and-suspenders approach” to safety, the nerd way to refer to using multiple redundant systems to prevent a single point of failure. It says the risk of human harm is “less than 1 in 100 million.” The company has said that for its Starlink V2 Mini satellite, about 5% of a satellite’s mass could potentially survive reentry, but insists its mostly harmless silicon fragments with the impact energy of a falling apple. Musk claims his latest Starlink V2 satellites are designed with better altitude and attitude controls to target reentry corridors with high accuracyroughly within 10% of an orbit ground track, which translates to about 10 minutes of flight time. SpaceX says it conducts plasma chamber tests simulating atmospheric conditions to better understand how components break up during reentry, seeking to improve prediction of debris survival. But no matter the improvements, every new satellite launched adds to an increasingly fragile orbital environment. SpaceX hasnt replied to Fast Companys request for comment. But theres more to consider beyond potential bodily harm. As these satellites burn up, they pollute the stratosphere with metal particles, creating what scientists call “anthropogenic meteor showers.” Researchers are now raising alarms that these metals, particularly aluminum, could linger for years and catalyze the destruction of the ozone layer. Atmospheric chemist Daniel Murphy told Science magazine in November 2024, Almost no one is thinking about the environmental impact on the stratosphere. Laser mass spectrometry studies detected elevated levels of lithium, aluminum, copper, and lead in the stratosphere, exceeding natural meteor input. These metals come from satellite reentries and may nearly double natural metal aerosol concentrations, threatening ozone protection. Currently, about 2,000 satellite reentries per year emit 17 metric tons of aluminum oxide nanoparticles into the stratosphere. The figure is rapidly rising as mega-constellations multiply. Astronomer Samantha Lawler told Science, We cant keep using the ground and the atmosphere as a dumpster. Heres how the European Space Agency CEO Josef Aschbacher warned about the existence of Musks satellites to the Financial Times back in 2021: You have one person owning half of the active satellites in the world. Thats quite amazing. De facto, he is making the rules. The rest of the world including Europe . . . is just not responding quick enough. We are watching as a billionaires unchecked ambition reshapes the orbital commons without real oversight. Space isnt meant to be Musks backyard. While humanity pays the price, Musk just shrugs and keeps aiming at planetary domination.
Category:
E-Commerce
Its hard to believe, but were only a few weeks away from Halloween, and after that comes Novemberand the unofficial start of the holiday season. If you travel during this time, youll know that finding cheap flights can be difficult. To get the best prices, people traditionally turn to comparison sites like Kayak and SkyScanner. However, as artificial intelligence seems to be taking over everything, and the tech industry wont stop shouting about its benefits, I decided to try three conversational AI tools to see if they could help me find the cheapest flight deal for the Thanksgiving period. Heres how that went. Flights listed on ChatGPT The first tool I turned to was AI chatbot king, ChatGPT. I gave it the following prompt for a hypothetical holiday trip, which is the exact same prompt I issued the other AI chatbots I tried for this article: I want you to find me the cheapest tickets for a round-trip flight for this Thanksgiving period. My departure city is New York City, and my destination city is Dallas. ChatGPT asked me a series of questions about trip specifics, including exact dates and preferred airports, or gave me the option for defaults it selected (all NYC DFW/DAL airports from Wednesday, November 26, 2025 to Sunday, November 30, 2025). I chose the latter. OpenAIs chatbot then spit out a bunch of information, including which airlines and airports had the cheapest options, and asked me if I wanted to see further results on specific itineraries. It even offered to show me total landed cost options (which means it would let me know how much the flight would cost if I checked bags, too). It also offered to set up price alerts for me. But it didnt stop there. As ChatGPT will carry on a conversation for as long as you want (and as I didnt want to ask and answer questions for 30 minutes), the chatbot also gave me three ABC options: A Show me the three absolute cheapest round-trip itineraries (all carriers, show baggage fees). B Show me the three best nonstop options (if any) ranked by total cost & convenience. C Compare a cheap Spirit/Frontier itinerary versus a reliable nonstop (AA/Delta) including checked-bag costs. [ChatGPT previously told me low cost carriers often show the lowest base fares]. I chose option A. Ultimately, ChatGPT returned three itinerary options with the absolute cheapest with one checked bag being between $190 and $220. It also gave me a direct link to the carriers website so I could book that option. Flights listed on Google Flight Deals Now that I had ChatGPTs answer, I next gave the same prompt to Googles new Flight Deals, its AI-powered Google Flights search tool. Google launched Flight Deals last month, billing it as an AI-powered search tool within Google Flights that is designed for flexible travelers whose number one goal is saving money on their next trip. Flight Deals lets you prompt the service like you would as though youre talking to a friendin natural languageand it will return flight itineraries that best fit your needs. I entered the same prompt I used with ChatGPT. Frustratingly, Flight Deals then asked me to confirm where I was flying from. I replied with NYC and then had to also select New York from the drop-down menu. However, I only received one result: a $249 nonstop United flight from Monday, November 24, to Friday, November 28. Flight Deals said it checked departures between Nov 24 and Nov 27, and returns between November 28 and December 1, which generally matches the Thanksgiving period I indicated in my prompt. A disclaimer for the results stated that The results shown are flights that are either significantly cheaper than usual for a route, time of year, trip length, and seating class, or are among the lowest-priced options for destinations that match your search. Unlike ChatGPT, Google Flight Deals did not allow me to ask follow-up questions or provide any tips on finding cheap flights. It also didnt tell me whether the $249 flight included checked baggage. Clicking on the sole result took me to Google Flights traditional interface, which showed additional flight results. Flights listed on iMean.ai Finally, I gave the same prompt I used for the others to iMean.ai, one of a growing number of dedicated conversational AI travel assistants. Even though iMean.ai’s interface looks like a more colorful version of ChatGPT, the sites AI agent didnt waste time asking me questions in an attempt to refine my prompt like OpenAIs chatbot did. Instead, it informed me that it searched through 302 flight options and determined that the best itinerary matching my query was one that split the departure and return flights between two airlines. The outbound flight leaves New York City on Wednesday, November 26, and the return flight leaves Dallas on Sunday, November 30. The total cost: $334. iMean.ai’s agent, like ChatGPT, provided me with the option to continue chatting with it to ask more questions or refine my needs. And unlike ChatGPT, iMean.ai displayed the results in a useful split-screen interface that helpfully laid out details, such as flight times, for each leg of the trip. Clicking on the accompanying View button took me to Kayak, where I could buy the selected tickets. Should you use AI chatbots to find cheap flights? In the end, consulting with three different AI tools resulted in the agents returning three different flight options for my hypothetical Thanksgiving trip, all on different dates, different airlines, and at different price points (ChatGPT: $190 to $220, Google Flight Deals: $249, iMean.ai: $334). Based on price alone, ChatGPT found me the cheapest tickets for a flight from New York City to Dallas during the Thanksgiving period. But though I was happy with the price, the experience of using AI chatbots to help me find cheap flight deals left me with a nagging feeling: uncertainty. ChatGPT presented me with too many questions and options. I felt that if I kept engaging with it, I would be sucked into a never-ending succession of possibilities that would make it nearly impossible to choose. When was the right time to stop prompting and make a choice? I didnt know. Conversely, Google Flight Deals provided me with a single option. This left me unsure as to whether it was actually the best choice (according to ChatGPT, no). iMean.ai was a mix of the two. It was conversational like ChatGPT, but provided fewer options, like Google Flights. However, it also presented me with an option that was more expensive than the ones both ChatGPT and Google Flight Deals presented, leaving me questioning its results. Ultimately, my experience with the three chatbots left me wanting to return to the traditional flight comparison websites, like Skyscanner and Kayak, that I am used to. If youre thinking of using chatbots to find deals on flightsfor this upcoming holiday season, it cant hurt to check out various AI agents to see what information they return, but Id still check the results of any AI recommendations against the results of traditional flight checking tools.
Category:
E-Commerce
Nearly every company I work with is focused on using AI to drive productivity and efficiency. They are starting to see real gains, and thats leading to excitement about AIs future potential. However, AI used to drive efficiency is only the starting line, and theres real risk if we stop there. In my work with Fortune 500 leaders across the C-suite, from chief HR officers (CHROs) to CTOs and CMOs, Ive seen that the very best organizations recognize a bigger opportunity: using AI to help managers build connection and trust with their teams. The companies that are able to leverage AI both to drive efficiency gains and to build highly motivated teams will be the ones that come out ahead. If youre only using AI for productivity, youre at risk AI is transforming work, and nearly every company I talk with is applying it to boost productivity from automating tasks to streamlining workflows and scaling output like coding and design. However, if we think about AI only as a tool for efficiency and cost cutting, were missing the bigger picture. Worse, we risk widening the trust gap that already exists in so many workplaces. The data is striking. A recent Upwork survey of 2,500 global workers, including 1,250 C-suite executives, found that 67% of top AI performers said they trust AI more than their colleagues, and 64% said they have a better relationship with AI than with their teammates. There is real risk present in those statements. At the end of the day, even with increased AI use, humans still have to work together to get things done. If we dont trust each other, efficiency gains from AI may be lost as organizations get mired in conflict, gossip, and fearthe hallmarks of low-trust company cultures. So, the question for every executive becomes: How do we equip the next generation of managers to use AI in service of connection? Here are three powerful ways. 1. AI as Your Leadership Memory The best managers Ive seen dont just lead, they remember. They pick up on the way individuals prefer feedback, they create a spark by recognizing people in a unique way, and they remember those personal details that tell someone they’re not just a cog in the wheel. Those moments of recognition build trust and loyalty. But today’s leaders are stretched. Teams are bigger, hierarchies flatter, and we simply cant hold all of that in our heads. As a leader myself, I often struggle to remember every detail about how each person on my team prefers to work or communicate. And every manager Ive coached has felt the same, because its hard to stay personal when youre juggling so much. Thats where AI steps in to reinforce what you already do best. Imagine before a one-on-one, your AI leadership memory gently reminds you that your direct report prefers written feedback over verbal, or that last time you spoke, they mentioned their childs soccer tournament. It can nudge you with thoughtful opening lines, and maybe even help you frame a difficult message so it lands in the best possible way. Thats not replacing the personal touch, its enabling it to scale. Leaders can use tools like Rising Team that automatically pull in insights that colleagues have shared, or manually upload materialslike personality assessmentsthat their team members have chosen to share. That way the AI can surface those details when they are helpful, without needing to use any private information. AI is helping you remember what matters, in the moment that it matters. 2. AI as Your Coaching Partner Some of the most meaningful moments in peoples careers come from the hardest conversations. Great leaders can deliver constructive feedback that helps people grow, diffuse conflict in a way that builds trust, and help teams be resilient amidst major challenges. Many managers just freeze, or wing it in conversations like these, not because they dont care, but because they dont know how to approach them and dont have a way to practice. And today, HR business partners can’t be there for every one of those moments. What if you could practice, and have your teams real dynamics baked in? Ive seen this dramatically shift things. Role-play with AI tailored to your specific engineer who needs time to process, or your marketer who craves blunt feedbackthats when AI coaching becomes real, actionable readiness. By practicing with a tool that knows your people, managers show up with clarity, empathy, and trust. 3. AI as Your Team Experience Builder Connection doesnt just happen in one-on-ones. Teams build trust and alignment in shared experiences, whether its learning a critical skill, tackling a big strategy shift, or building insights about how to work together as a team. But creating and facilitating those sessions takes time and expertise, and most managers dont have the time or the support to do it well. This is where AI can help. Think of it like working with a facilitation proone who knows theory and your team context. It can layer in warm-ups, activity ideas, reflective questions, and even capture what people say, track whos engaged, and surface next steps. With help from AI, managers are now capable of bringing their teams together to build trust and connection in synchronous experiences that were too hard or expensive to do before. As an executive, imagine rolling out a new company initiative or framework. Instead of relying on slides and top-down presentations, AI can now help your managers lead team sessions with reflective prompts, collaborative exercises, and clear action plans. This ensures that experiences across the organization are consistent, measurable, and engaging. Beyond Productivity: Building a System for Connected Leadership AI is often portrayed as a catalyst for productivity. And yes, its great for that, but I believe the real frontier is AI as a force to drive connectiona leadership system for modern teams. Because the future isn’t about choosing AI or humanity. Its about how we use AI to amplify our humanity, and build teams that are not just productive, but also high trust, resilient, and great at delivering results together.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||