Connect with us

Technologies

SpaceX Calls Mission 3 a Success, Despite Losing Starship: How to Rewatch

On its third attempt, SpaceX launched its Starship and cruised into space, but lost the rocket after reentry to Earth.

SpaceX launched its third Starship mission on Thursday, with the space exploration company owned by Elon Musk forging ahead after the first two attempts exploded after takeoff. SpaceX considers those first two missions successful, thanks to the data it was able to collect, and the third mission was the most successful of the bunch.

Shortly before 9:30 a.m. ET Thursday, the company posted three words to the Musk-owned X social media site: «Liftoff of Starship!» A 36-second video showed the rocket engines igniting and then Starship rising amid a cloud of exhaust smoke and up into the sky.

But the rocket did not complete the round trip, as you can see by rewatching the full test flight. «The ship has been lost. No splashdown today,» Dan Huot of SpaceX communications said on the stream. «But we were able to get through some of the early phases of reentry.»

SpaceX quality engineering manager Kate Tice noted on the stream that SpaceX wasn’t intending to recover Starship anyway, and had been planning to crash it into the ocean. 

Starship is arguably the most ambitious effort for Musk, who owns the satellite-based internet company Starlink along with X, EV maker Tesla and the neurotechnology company Neuralink. The Starship missions are critical to SpaceX’s — and Musk’s — goal of getting to and eventually settling the moon and Mars. 

With a new flight trajectory and hopes for new data insights, the space company’s third mission may prove to be its most important yet.

When did the Starship mission launch?

Starship’s third mission launched on March 14. It had been pending favorable weather and a license from the Federal Aviation Administration.

How to rewatch Starship mission 3

SpaceX set up livestreams for watching the third mission. One was the SpaceX account on X, and another was the SpaceX third mission landing page. You could also watch via CNET’s YouTube channel.

You can rewatch the launch on X, on the mission 3 page, or via CNET’s YouTube stream embedded here.

What SpaceX achieved in Starship mission 3

SpaceX’s third Starship mission was designed to test whether the spacecraft can complete certain tasks. After liftoff, the company planned to open Starship’s payload door and transfer its propellant from one part of the vehicle to another. For the first time, SpaceX also attempted to relight its Raptor engine while in space, a test that could be critical for future missions as it eventually tries to propel Starship through space.

Starship took a different flight path this time around, and had planned to land in the Indian Ocean instead of the Pacific Ocean until it lost the vehicle after reentry. In a statement, SpaceX said that the new flight path was designed to maximize «public safety,» but the company didn’t discuss how. 

The new flight path also paved the way for SpaceX to try «in-space engine burns,» a reference to the company attempting to reignite the Raptor engine in space.

«Huge congratulations to the entire team for this incredible day: clean count (glad the shrimpers could get out in the nick of time!), liftoff, hot staging, Super Heavy boost back and coast (and likely a couple engines making mainstage during landing burn!), clean ship ‘insertion’ and coast, payload door cycling and prop transfer demo (to be confirmed!), and ship entry!» SpaceX President Gwynne Shotwell posted on X following the launch, naming the successful components of the test.

The mission was slated to last for 1 hour, 15 minutes. Previous missions, if completed, would have lasted 90 minutes.

What happened in the previous missions?

April mission: Forced detonation

The first Starship mission launched in April 2023. Early on in the mission, the two stages of Starship — the reusable upper stage, called Starship, and its Super Heavy first-stage booster — were supposed to separate. That didn’t happen, and for safety reasons, the SpaceX team was forced to detonate the vehicle just 4 minutes into the mission.

November mission: Explosion due to liquid oxygen

In November 2023, Starship launched on its second mission. That time around, Starship was able to separate its two stages and it reached nominal first-stage engine burn. However, Starship exploded 8 minutes after launch when it tried to vent its liquid oxygen. Oddly, the explosion may not have needed to happen. Earlier this year, Musk said on a real mission carrying payload — meaning the materials a spaceship carries to perform its scientific mission — liquid oxygen wouldn’t be onboard.

«Starship’s second flight test achieved a number of major milestones and provided invaluable data to continue rapidly developing Starship,» the company wrote on its site. «Each of these flight tests continue to be just that: a test. They aren’t occurring in a lab or on a test stand, but are putting flight hardware in a flight environment to maximize learning.»

Corinne Reichert contributed to this report.

Technologies

Gen AI Chatbots Are Starting to Remember You. Should You Let Them?

An AI model’s long memory can offer a better experience — or a worse one. Good thing you can turn it off.

Until recently, generative AI chatbots didn’t have the best memories: You tell it something and, when you come back later, you start again with a blank slate. Not anymore. 

OpenAI started testing a stronger memory in ChatGPT last year and rolled out improvements this month. Grok, the flagship tool of Elon Musk’s xAI, also just got a better memory.

It took significant improvements in math and technology to get here but the real-world benefits seem pretty simple: You can get more consistent and personalized results without having to repeat yourself.

«If it’s able to incorporate every chat I’ve had before, it does not need me to provide all that information the next time,» said Shashank Srivastava, assistant professor of computer science at the University of North Carolina at Chapel Hill.

Those longer memories can help with solving some frustrations with chatbots but they also pose some new challenges. As with when you talk to a person, what you said yesterday might influence your interactions today.

Here’s a look at how the bots came to have better memories and what it means for you.

Improving an AI model’s memory

For starters, it isn’t quite a «memory.» Mostly, these tools work by incorporating past conversations alongside your latest query. «In effect, it’s as simple as if you just took all your past conversations and combined them into one large prompt,» said Aditya Grover, assistant professor of computer science at UCLA.

Those large prompts are now possible because the latest AI models have significantly larger «context windows» than their predecessors. The context window is, essentially, how much text a model can consider at once, measured in tokens. A token might be a word or part of a word (OpenAI offers one token as three-quarters of a word as a rule of thumb). 

Early large language models had context windows of 4,000 or 8,000 tokens — a few thousand words. A few years ago, if you asked ChatGPT something, it could consider roughly as much text as is in this recent CNET cover story on smart thermostats. Google’s Gemini 2.0 Flash now has a context window of a million tokens. That’s a bit longer than Leo Tolstoy’s epic novel War and Peace. Those improvements are driven by some technical advances in how LLMs work, creating faster ways to generate connections between words, Srivastava said. 

Other techniques can also boost a model’s memory and ability to answer a question. One is retrieval-augmented generation, in which the model can run a search or otherwise pull up documents as needed to answer a question, without always keeping all of that information in the context window. Instead of having a massive amount of information available at all times, it just needs to know how to find the right resource, like a researcher perusing a library’s card catalog.

Read more: AI Essentials: 27 Ways to Make Gen AI Work for You, According to Our Experts

Why context matters for a chatbot

The more an LLM knows about you from its past interactions with you, the better suited to your needs its answers will be. That’s the goal of having a chatbot that can remember your old conversations. 

For example, if you ask an LLM with no memory of you what the weather is, it’ll probably follow up first by asking where you are. One that can remember past conversations, however, might know that you often ask it for advice about restaurants or other things in San Francisco, for example, and assume that’s your location. «It’s more user-friendly if the system knows more about you,» Grover said.

A chatbot with a longer memory can provide you with more specific answers. If you ask it to suggest a gift for a family member’s birthday and tell it some details about that family member, it won’t need as much context when you ask again next year. «That would mean smoother conversations because you don’t need to repeat yourself,» Srivatsava said.

A long memory, however, can have its downsides.

You can (and maybe should) tell AI to forget

Having a chatbot recommend a gift poses a conundrum that’s all too common in human memories: You told your aunt you liked airplanes when you were 12 years old, and decades later you still get airplane-themed gifts from her. An LLM that remembers things about you could bias itself too much toward something you told it before.

«There’s definitely that possibility that you can lose your control and that this personalization could haunt you,» Srivastava said. «Instead of getting an unbiased, fresh perspective, its judgment might always be colored by previous interactions.»

LLMs typically allow you to tell them to forget certain things or to exclude some conversations from their memory. 

You may also deal with things you don’t want an AI model to remember. If you have private or sensitive information you’re communicating with an LLM (and you should think twice about doing so at all), you probably want to turn off the memory function for those interactions. 

Read the guidance on the tool you’re using to be sure you know what it’s remembering, how to turn it on and off and how to delete items from its memory. 

Grover said this is an area where gen AI developers should be transparent and offer clear commands in the user interface. «I think they need to be providing more controls that are visible to the user, when to turn it on, when to turn it off,» he said. «Give a sense of urgency for the user base so they don’t get locked into defaults that are hard to find.»

How to turn off gen AI memory features

Here’s how to manage memory features in some common gen AI tools.

ChatGPT

OpenAI has a couple types of memory in its models. One is called «reference saved memories» and it stores details that you specifically ask ChatGPT to save, like your name or dietary preferences. Another, «reference chat history,» remembers information from past conversations (but not everything). 

To turn off either of these features, you can go to Settings and Personalization and toggle the items off. 

You can ask ChatGPT what it remembers about you and ask it to forget something it has remembered. To completely delete this information, you can delete the saved memories in Settings and the chat where you saved that information.

Gemini

Google’s Gemini model can remember things you’ve discussed or summarize past conversations. 

To modify or delete these memories, or to turn off the feature entirely, you can go into your Gemini Apps Activity menu.

Grok

Elon Musk’s xAI announced memory features in Grok this month and they’re turned on by default. 

You can turn them off under Settings and Data Controls. The specific setting is different between Grok.com, where it’s «Personalize Grok with your conversation history,» and on the Android and iOS apps, where it’s «Personalize with memories.» 

Continue Reading

Technologies

It’s OK if You Didn’t Preorder a Switch 2

Commentary: As good as the new console looks, it’s also fine to wait.

FOMO for new tech is hard. And new game consoles are exciting. I get it, and I’ve contributed to that coverage excitement too. The Nintendo Switch 2 finally became available to preorder in the US this week, and as expected, it looks sold out for now. That’ll change over time, but it’s unclear when, or how, and it’s equally unclear what the constant tariff fluctuations might do to future game console pricing. 

That said, having played on the Switch 2 recently at an event, may I help ease your FOMO somewhat by saying you’re probably OK waiting on it?

I felt this way after my full-day Switch 2 experience, and I’ll reiterate it now: As good as the upgrades the Switch 2 has, and as fun as the new Mario Kart and Donkey Kong games seem to be — and the GameCube gaming library also seems like a blast of retro fun — the Switch 2 is very much an iterative upgrade for now. The very best games on the Switch 2, and its most unique exclusives, are likely still to come.

Nintendo has clearly designed the Switch 2, at least for the moment, to exist as a bridge to the current Switch, with many upcoming games intended to work on the original Switch too. Much more than the debut of the first Switch, the Switch 2 is designed to be a system you could wait to upgrade to. In that sense, it’s following the path of the current gen of Xbox Series X and S and PlayStation 5 consoles.

You can build up your Switch library now and be Switch 2-ready when you eventually upgrade

The Switch 2 plays all the Switch games, which wasn’t the case with the Switch and previous Wii U and 3DS hardware. That means you could skip the Switch 2 now if you needed to, play games on the Switch, and then move your library over whenever. Switch 2 versions of games cost more (ranging from $10 to $20 more), but you can just buy the Switch 2 game upgrades later for a similar price — or play the versions you’ve already got minus the enhanced graphics and game extras.

The Switch 2’s current upgrades are good, but not shockingly good

After playing several of the Switch 2 Edition versions of Switch games for a bit, I noticed better frame rates and graphics resolution, but I honestly didn’t find it to be that much different. I’d prefer playing the enhanced Switch 2 editions, but the experience reminded me a bit of the PS5 Pro versus PS5 versions of games when I first played on the console with Sony last year.

If you have a big TV, you’ll likely appreciate the difference. The bigger Switch 2 screen shows off games in higher-res 1080p with HDR, but you could play on the older Switch and be fine. I’m playing on a Switch OLED again, and after the Switch 2 experience, I don’t have massive I-wish-this-were-a-Switch-2-envy. 

I’m sure this will change as games are developed to take better advantage of the amped-up Nvidia-powered Switch 2 GPU, and when more exclusives arrive. It’s similar to how I felt about the Meta Quest 3, which has better graphics than Quest 2 but didn’t feel like an absolute must-get until a year into its release.

You can still play upcoming Nintendo games on OG Switch

While Mario Kart World and Donkey Kong Bananza are Switch 2 exclusives, Metroid Prime 4 Beyond and Pokemon Legends Z-A also play on the Switch. It’s unclear how well these games will play on the Switch versus Switch 2, but you can get a good dose of New Nintendo this year on the older hardware and upgrade the hardware upgrade later. Think of it as a bit of a FOMO buffer.

Looking at Nintendo’s game history, the company often supported its previous consoles for a good couple of years after the new hardware’s release. I’d expect that after 2026 the Switch 2 will start to become the go-to platform for most big game, but I wouldn’t be surprised to see a handful of key Nintendo games still supporting original Switch for another year at least.

There’s no ‘whole new experience’ you’ll miss other than Game Chat, that camera and the mouse 

The original Switch was an eye-opener because it was a portable, full game console that could dock with your TV and turn into a shareable console with modular controllers. It was different from anything Nintendo had made before. The Switch 2 is mostly the same proposition, just nicer.

You won’t feel the same regret for missing out on a whole new way to play this time, since it’s a continuation of the same idea. There are two new features you might envy: audio or video Game Chat among friends and the new Joy-Cons working like mice in some supported games. But Game Chat works only with other Switch 2 owners and needs a Switch Online subscription. The mouse functions are fun at times, but could also end up as just a gimmick. For now, the Switch 2 hasn’t pulled that many wild new functions out of its hat, but that could change, knowing Nintendo. There are also some fun camera-connected party game modes for Mario Party Jamboree if you happen to connect a camera, but no other games even have new camera-based features yet.

It’s fine to wait, but tariffs are still a question mark

I’m saying this well before I’ve had a chance to review the Switch 2, and for sure, it looks like the best Nintendo console in a long while and worth upgrading to. But take some comfort that missing out on getting one early this time isn’t quite as big a deal as it was in 2017, even if you’re feeling the pull of regret. 

The only wild card remains the question of the effect tariffs will have on future console pricing. Will it fluctuate? I hope not, but the prices of Nintendo’s Switch 2 accessories have already gone up as a result of Trump’s chaotic tariff policies, and it’s unclear if that might happen again. The state of pricing and consumer electronics is still in an unknown zone, but in the meantime, you can still have a lot of fun on the Switch you already have, now and even in the near future.

Continue Reading

Technologies

Apple to Shift All US iPhone Assembly to India Amid Tariff Turmoil, Report Says

The manufacturing move aims to address massive US tariffs against China that could spur higher prices on the company’s biggest-selling product.

Apple could be sourcing its entire line of iPhones for the US market — about 60 million devices a year — from assembly facilities in India by the end of 2026, according to a report from the Financial Times

The planned move comes against the backdrop of the Trump administration imposing tariffs against China of up to 145%, although some products such as mobile phones and computers have been exempted for the time being. Apple has long centered its iPhone production in China, making it vulnerable to any trade war between the two countries and spurring speculation that tariffs could mean price increases for the company’s biggest-selling product. 

By moving third-party assembly of US iPhones to India, Apple could avoid the most significant cost pressure of a trade war, though India itself faces new tariffs as well.

The company had already begun shipping iPhones made in India, adding to its product reserves, before new tariffs became active.

A representative for Apple did not immediately respond to a request for comment.   

Continue Reading

Trending

Copyright © Verum World Media