News
SpaceX wants to attempt Starship booster catch during first orbital launch
An updated document submitted by SpaceX to the US Federal Communications Commission (FCC) has revealed details about the company’s plan for the first Starship booster ‘catch’ attempt.
The document follows a different batch submitted by SpaceX in June 2021, when the company detailed its plans for Starship’s orbital launch debut as background while requesting permission from the FCC to use Starlink dishes for in-flight telemetry. A month earlier, a different request focused on more standard telemetry antennas had already revealed that even if the mission went perfectly, Starship would not fully reach orbit on its first attempted spaceflight. It also confirmed that SpaceX had no intention of recovering the upper stage or Super Heavy booster assigned to Starship’s launch debut – a sort of implicit acknowledgment that success was (then) not expected on the first try.
Twelve months later, SpaceX has submitted an updated overview of Starship’s orbital launch debut in a new request for permission to use multiple Starlink dishes on both stages. While most of the document is the same, a few particular details have changed about Super Heavy’s role in the mission.
This time around, SpaceX says that the Super Heavy booster will “will separate[,] perform a partial return[,] and land in the Gulf of Mexico or return to Starbase and be caught by the launch tower.” Prior to this document, SpaceX’s best-case plans for the first Super Heavy booster to launch never strayed from a controlled splashdown in the Gulf of Mexico – potentially demonstrating that it would be safe to attempt booster recovery on the next launch but all but guaranteeing that the first booster would be lost at sea.
A year later, SpaceX appears to be a bit more confident and wants to leave itself the option to attempt to recover the first Super Heavy booster that launches. However, the company has dramatically complicated the process of testing early Super Heavy and Starship recovery (and thus reuse) by fully removing traditional and predictable landing legs and designing its latest prototypes such that the only way they can be recovered in one piece is with a giant mechanized ‘launch tower’ nicknamed Mechazilla.
The launch tower and its three mobile arms will play a crucial role in all aspects of orbital Starship launches. The first arm swings out to brace Super Heavy for Starship installation and connect the upper stage to power, propellant supplies, and other launch pad utilities. A more exotic pair of arms nicknamed ‘chopsticks’ has a more complex job. On top of using the chopsticks to lift, stack, and demate Starships and Super Heavy boosters and almost any weather and wind conditions, SpaceX wants to use the arms as an incredibly complex and precarious rocket recovery system.
For a booster or Starship “catch,” the rocket will approach the tower, enter the gap between the splayed arms, hover in place while the arms close around it, and eventually come to rest on hardpoints that appear to offer about as much surface area as a coffee table. Based on a simulation of the process shown by Elon Musk, calling it a “catch” is a misnomer, as the arms will mainly move in one dimension (open/close) and can’t actually ‘grab’ the rocket in any real sense. As built and shown, they are closer to a tiny fixed landing platform capable of minor last-second positional adjustments.
Eventually, the chopsticks could shave a small amount of time off of post-recovery processing, removing the need for a crane (or the same arms) to attach to a landed booster or ship. They could also shave off the dry mass required for landing legs, though all interplanetary ships will still need legs. However, they will also inherently make proving their own efficacy a nightmare. By all appearances, the current recovery mechanisms on the arms and the landing hardpoints on ships and boosters mean that a ‘catch’ could fail if either stage is more than a foot or two from a perfect bullseye or rotated a few degrees in the wrong direction. With the method SpaceX has devised, even the tiniest error could easily end with a massive, pressurized, partially-fueled rocket destroying the chopsticks and plummeting a few hundred feet to the ground, guaranteeing an explosion that could damage surrounding infrastructure or start fires that might.
In the event of larger anomalies during a landing attempt, Starship or Super Heavy could accidentally impact the launch tower, damaging or even outright destroying the skyscraper-sized structure. Ultimately, the immense risk posed by any catch attempt means that unless SpaceX has miraculously gotten the design of everything involved nearly perfect on its first try, the company will have to be extraordinarily cautious and expend a large number of ships and boosters to avoid rendering its only Starship launch tower unusable.
At least to some extent, SpaceX likely knows this and Super Heavy would likely need to be in excellent health and perform perfectly during the ascent and boostback portions of its launch debut to be cleared for a catch attempt. Ultimately, Starship’s first orbital launch could end up being even more of a spectacle than it’s already guaranteed to be.
News
Tesla intertwines FSD with in-house Insurance for attractive incentive
Every mile logged under FSD now carries a documented financial value—lower risk, lower cost—based on Tesla’s internal driving data rather than external crash statistics alone.
Tesla intertwined its Full Self-Driving (Supervised) suite with its in-house Insurance initiative in an effort to offer an attractive incentive to drivers.
Tesla announced that its new Safety Score 3.0 will automatically have a perfect score of 100 with every mile driven with Full Self-Driving (Supervised) enabled.
The change is designed to boost customers’ average safety scores and deliver noticeably lower monthly premiums.
The move marks the clearest link yet between Tesla’s autonomous driving technology and its proprietary insurance product. Tesla Insurance already relies on real-time vehicle data—such as acceleration, braking, following distance, and speed—to calculate a Safety Score between 0 and 100. Higher scores have long translated into cheaper rates.
Under the previous system, however, even brief manual interventions could drag down the average, frustrating owners who rely heavily on FSD. Version 3.0 eliminates that penalty for supervised autonomous miles, effectively treating FSD-driven segments as the safest possible driving behavior.
The incentive is immediate and financial. Drivers who keep FSD engaged for the majority of their trips will see their overall score rise, potentially shaving hundreds of dollars off annual premiums.
Tesla framed the update as a direct response to customer feedback, many of whom had complained that the old scoring model punished the very behavior it was meant to encourage.
For now, the program applies only to new policies in six states: Indiana, Tennessee, Texas, Arizona, Virginia, and Illinois.
Existing policyholders are not yet included, a point that drew swift questions from the Tesla community. Many owners in other states, including California and Georgia, expressed hope that the benefit would expand nationwide soon.
The announcement arrives as Tesla continues to roll out FSD Supervised updates and push for regulatory approval of more advanced autonomy. By tying insurance savings directly to FSD usage, the company is putting its own actuarial weight behind the technology’s safety claims.
Every mile logged under FSD now carries a documented financial value—lower risk, lower cost—based on Tesla’s internal driving data rather than external crash statistics alone.
Tesla has not disclosed exact premium reductions or the full rollout timeline beyond the six launch states.
Still, the message is clear: the more drivers trust FSD Supervised, the more Tesla Insurance will reward them. In an era when legacy insurers remain cautious about autonomous tech, Tesla is betting that its own data will prove the safest miles are the ones driven hands-free.
Elon Musk
Tesla finalizes AI5 chip design, Elon Musk makes bold claim on capability
The Tesla CEO’s words mark a strategic shift. Tesla has long emphasized software-hardware co-design, squeezing maximum performance from every transistor. Musk previously described AI5 as optimized for edge inference in both Robotaxi and Optimus.
Tesla has finalized its chip design for AI5, as Elon Musk confirmed today that the new chip has reached the tape-out stage, the final step before mass production.
But in a brief reply on X, Musk clarified Tesla’s AI hardware roadmap, essentially confirming that the new chip will not be utilized for being “enough to achieve much better than human safety for FSD.”
He said that AI4 is enough to do that.
Instead, the AI5 chip will be focused on Tesla’s big-time projects for the future: Optimus and supercomputer clusters.
Musk thanked TSMC and Samsung for production support, noting that AI5 could become “one of the most produced AI chips ever.” Yet, the key pivot came in his direct answer: vehicles no longer need the bleeding-edge silicon.
And thank you to @TaiwanSemi_TSC and @Samsung for your support in bringing this chip to production! It will be one of most produced AI chips ever.
— Elon Musk (@elonmusk) April 15, 2026
Existing AI4 hardware, which is already deployed in hundreds of thousands of HW4-equipped Teslas, delivers safety metrics superior to human drivers for Full Self-Driving. AI5 will instead accelerate Optimus robot development and massive Dojo-style training clusters.
The Tesla CEO’s words mark a strategic shift. Tesla has long emphasized software-hardware co-design, squeezing maximum performance from every transistor. Musk previously described AI5 as optimized for edge inference in both Robotaxi and Optimus.
Now, with AI4 proving sufficient, the company avoids costly retrofits across its fleet while redirecting next-generation compute toward higher-value applications: dexterous robots and exponential training scale.
But is it reasonable to assume AI4 enables unsupervised self-driving? Yes, but with important caveats.
On the hardware side, the claim is credible. Tesla’s FSD stack runs end-to-end neural networks trained on billions of miles of real-world data. Internal safety data reportedly shows AI4-equipped vehicles already outperforming average human drivers by a significant margin in controlled metrics (collision avoidance, reaction time, edge-case handling).
Dual-redundant AI4 chips provide ample headroom for the driving task, leaving bandwidth for future model improvements without new silicon. Musk’s assertion aligns with Tesla’s pattern of over-provisioning compute early, then optimizing ruthlessly, exactly as HW3 once sufficed before HW4 scaled further.
Optimus and our supercomputer clusters.
AI4 is enough to achieve much better than human safety for FSD.
— Elon Musk (@elonmusk) April 15, 2026
Unsupervised autonomy, meaning Level 4 or higher, is not solely a compute problem. Regulatory approval remains the primary gate.
Even if AI4 achieves “much better than human” safety statistically, agencies like the NHTSA demand exhaustive validation, liability frameworks, and public trust.
Tesla’s supervised FSD has shown rapid gains in recent versions, yet real-world edge cases, like construction zones, emergency vehicles, and adverse weather, still require driver intervention in many jurisdictions. Competitors like Waymo operate limited unsupervised fleets, but only in geofenced areas with extensive mapping. Tesla’s vision-only, fleet-scale approach is more ambitious—and harder to certify globally.
In short, Musk’s post is both pragmatic and bullish. AI4 is likely capable of unsupervised FSD from a technical standpoint. Whether regulators and consumers agree, and how quickly, will determine if Tesla’s bet pays off.
The company’s capital-efficient path keeps existing cars relevant while pouring future compute into robots. If the safety data holds, unsupervised autonomy could arrive sooner than many expect.
Elon Musk
Elon Musk signals expansion of Tesla’s unique side business
Long envisioning the Tesla Diner as more than a charging stop, Musk has clearly adopted the idea that the Supercharger and Restaurant combo is a good thing for the company to have. It’s a blend of classic American drive-in culture with futuristic Tesla flair, complete with a 1950s-inspired design, movie screens, and on-site dining.
Elon Musk has signaled an expansion of Tesla’s unique side business, something that really has nothing to do with cars or spaceships, but fans of the company have truly adopted it as just another one of its awesome ventures.
Musk confirmed on Wednesday that Tesla would build a new Diner location in Palo Alto, Northern California. After hinting last October that it “probably makes sense to open one near our Giga Texas HQ in Austin and engineering HQ in Palo Alto,” it seems one of those locations is being set into motion.
Sure
— Elon Musk (@elonmusk) April 15, 2026
Long envisioning the Tesla Diner as more than a charging stop, Musk has clearly adopted the idea that the Supercharger and Restaurant combo is a good thing for the company to have. It’s a blend of classic American drive-in culture with futuristic Tesla flair, complete with a 1950s-inspired design, movie screens, and on-site dining.
He first floated broader expansion plans shortly after the LA opening in July 2025, noting that if the prototype succeeded, Tesla would roll out similar venues in major cities worldwide and along long-distance Supercharger routes.
Earlier hints included a confirmed second site at Starbase in Texas, tied to SpaceX operations, underscoring the Diner’s role in enhancing Tesla’s ecosystem behind vehicles.
The Los Angeles location on Santa Monica Boulevard in West Hollywood has served as a high-profile test case. Opened in July 2025 at 7001 Santa Monica Blvd., it features the world’s largest urban Supercharging station with 80 V4 stalls open to all NACS-compatible EVs, over 250 dining seats, rooftop views, and 24/7 service.
The retro-futuristic building replaced a former Shakey’s and quickly became a destination. Tesla reported selling 50,000 burgers in the first 72 days—an average of over 700 daily—drawing crowds with Cybertruck-shaped packaging, breakfast extensions until 2 p.m., and movie screenings.
Palo Alto stands out as a logical next step for several reasons. As Tesla’s longstanding engineering headquarters in the heart of Silicon Valley, the city is home to thousands of Tesla employees, engineers, and executives who could benefit from a convenient, branded gathering spot.
The area boasts high EV adoption rates, dense tech talent, and heavy traffic along key corridors, making a large Supercharger-diner an ideal fit for both daily commuters and long-haul travelers.
Proximity to Stanford University and the innovation ecosystem would amplify its appeal, potentially serving as a showcase for Tesla’s vision of integrated mobility and lifestyle experiences. It could be a great way for Tesla to recruit new talent from one of the country’s best universities.
If Tesla and Musk decide to move forward with a Palo Alto diner, it would build directly on the LA prototype’s momentum while addressing Musk’s earlier calls for expansion near core Tesla hubs.
Whether it materializes as a full confirmation or evolves from these hints remains to be seen, but the pattern is clear: Tesla is testing ways to make charging stops memorable. For EV drivers and enthusiasts alike, a Silicon Valley outpost could blend cutting-edge tech with nostalgic comfort, further embedding Tesla into everyday culture. As Musk’s comments suggest, the future of the Diner looks promising.
Tesla intertwines FSD with in-house Insurance for attractive incentive
Tesla finalizes AI5 chip design, Elon Musk makes bold claim on capability
