News
SpaceX simulates lifting Starship with launch tower ‘arms’
SpaceX has taken Starbase’s rocket-catching launch tower ‘arms’ to new heights during the latest series of proof tests.
That process began in earnest on January 4th, when SpaceX lifted, opened, and swung the tower’s building-sized arms for the first time. Four days later, SpaceX performed a variation on the first round of tests, again slowly lifting the assembly up the side of the launch tower and opening and closing the arms. The most notable difference was the addition of several tandem swing tests, which hinted at more applied tests that were soon to come. SpaceX also performed some basic tests with a third Starship fueling arm higher up on the tower, very slowly swinging it towards where Starship would be standing.
On Sunday afternoon, a third major round of testing kicked off. This set of tests was considerably more focused than the prior two, suggesting that it was more of a simulation of the main purpose of the arms.
Instead of lifting a few dozen feet and performing basic actuation and coordination tests, SpaceX simply lifted the arm assembly up along the tower’s exterior and didn’t stop. There were a few temporary pauses but the arms ultimately reached the approximate height they’d need to reach to stack a Starship on top of a Super Heavy booster. In fact, despite being (in)famous for being partially designed to catch boosters and ships out of mid-air, the main purpose of the arms – and likely the only reason they exist at all – is to safely, accurately, and precisely lift, install, and stack Starships and Super Heavy boosters.
SpaceX could obviously use a giant crawler or tower crane to accomplish a similar feat but cranes – especially large and tall ones – are extremely sensitive to wind conditions and effectively become very unsafe to operate in anything more than a brisk breeze. To put it lightly, even the average weather on the South Texas Gulf Coast is anything but conducive to the routine and reliable operation of giant cranes, which is exactly what SpaceX would need to avoid near-future Starship launch and recovery operations being constantly delayed by weather.
On January 9th, SpaceX appeared to test exactly that function. Before the day’s testing began, workers installed a large steel bar believed to be a weight simulator between the arms. Just like a booster would, the simulator sat – one end resting on both arms – on two small steel appendages identical to those present on all recent Super Heavy prototypes. On top of serving as a hardpoint for cranes, the downward-facing end of the L-shaped structures are capped with a small steel tip designed to take the whole weight of a Super Heavy. Those two minuscule steel caps – each no more than a foot wide – are what SpaceX (or at least CEO Elon Musk) wants Super Heavy to ‘land’ on to be “caught” by the launch tower’s arms.
More importantly, those caps – covering heavy-duty bearings – are also what the arms will ‘grab’ and manipulate to carefully position Super Heavy boosters for launch mount installation. To do so, each arm has a pair of parallel screw rods that can move together to shift the booster closer to or further away from the launch tower or move in opposite directions to slightly rotate it.
Once the arms reached the top of the tower, SpaceX performed several swing tests, mirroring the kind of movements they would use to carefully lift Starship, swing it over top of Super Heavy, and mate the two stages. Ultimately, the tower seemed to complete the simulation without any showstopping issues. Up next, it’s possible that SpaceX will add weights to the simulator bar to fully simulate the 100-200 ton masses of Starship and Super Heavy. Eventually, SpaceX may also use Starship S20 and Super Heavy B4 to fully qualify the arms by actually lifting, stacking, and removing both stages.
Elon Musk
Tesla’s Elon Musk: 10 billion miles needed for safe Unsupervised FSD
As per the CEO, roughly 10 billion miles of training data are required due to reality’s “super long tail of complexity.”
Tesla CEO Elon Musk has provided an updated estimate for the training data needed to achieve truly safe unsupervised Full Self-Driving (FSD).
As per the CEO, roughly 10 billion miles of training data are required due to reality’s “super long tail of complexity.”
10 billion miles of training data
Musk comment came as a reply to Apple and Rivian alum Paul Beisel, who posted an analysis on X about the gap between tech demonstrations and real-world products. In his post, Beisel highlighted Tesla’s data-driven lead in autonomy, and he also argued that it would not be easy for rivals to become a legitimate competitor to FSD quickly.
“The notion that someone can ‘catch up’ to this problem primarily through simulation and limited on-road exposure strikes me as deeply naive. This is not a demo problem. It is a scale, data, and iteration problem— and Tesla is already far, far down that road while others are just getting started,” Beisel wrote.
Musk responded to Beisel’s post, stating that “Roughly 10 billion miles of training data is needed to achieve safe unsupervised self-driving. Reality has a super long tail of complexity.” This is quite interesting considering that in his Master Plan Part Deux, Elon Musk estimated that worldwide regulatory approval for autonomous driving would require around 6 billion miles.
FSD’s total training miles
As 2025 came to a close, Tesla community members observed that FSD was already nearing 7 billion miles driven, with over 2.5 billion miles being from inner city roads. The 7-billion-mile mark was passed just a few days later. This suggests that Tesla is likely the company today with the most training data for its autonomous driving program.
The difficulties of achieving autonomy were referenced by Elon Musk recently, when he commented on Nvidia’s Alpamayo program. As per Musk, “they will find that it’s easy to get to 99% and then super hard to solve the long tail of the distribution.” These sentiments were echoed by Tesla VP for AI software Ashok Elluswamy, who also noted on X that “the long tail is sooo long, that most people can’t grasp it.”
News
Tesla earns top honors at MotorTrend’s SDV Innovator Awards
MotorTrend’s SDV Awards were presented during CES 2026 in Las Vegas.
Tesla emerged as one of the most recognized automakers at MotorTrend’s 2026 Software-Defined Vehicle (SDV) Innovator Awards.
As could be seen in a press release from the publication, two key Tesla employees were honored for their work on AI, autonomy, and vehicle software. MotorTrend’s SDV Awards were presented during CES 2026 in Las Vegas.
Tesla leaders and engineers recognized
The fourth annual SDV Innovator Awards celebrate pioneers and experts who are pushing the automotive industry deeper into software-driven development. Among the most notable honorees for this year was Ashok Elluswamy, Tesla’s Vice President of AI Software, who received a Pioneer Award for his role in advancing artificial intelligence and autonomy across the company’s vehicle lineup.
Tesla also secured recognition in the Expert category, with Lawson Fulton, a staff Autopilot machine learning engineer, honored for his contributions to Tesla’s driver-assistance and autonomous systems.
Tesla’s software-first strategy
While automakers like General Motors, Ford, and Rivian also received recognition, Tesla’s multiple awards stood out given the company’s outsized role in popularizing software-defined vehicles over the past decade. From frequent OTA updates to its data-driven approach to autonomy, Tesla has consistently treated vehicles as evolving software platforms rather than static products.
This has made Tesla’s vehicles very unique in their respective sectors, as they are arguably the only cars that objectively get better over time. This is especially true for vehicles that are loaded with the company’s Full Self-Driving system, which are getting progressively more intelligent and autonomous over time. The majority of Tesla’s updates to its vehicles are free as well, which is very much appreciated by customers worldwide.
Elon Musk
Judge clears path for Elon Musk’s OpenAI lawsuit to go before a jury
The decision maintains Musk’s claims that OpenAI’s shift toward a for-profit structure violated early assurances made to him as a co-founder.
A U.S. judge has ruled that Elon Musk’s lawsuit accusing OpenAI of abandoning its founding nonprofit mission can proceed to a jury trial.
The decision maintains Musk’s claims that OpenAI’s shift toward a for-profit structure violated early assurances made to him as a co-founder. These claims are directly opposed by OpenAI.
Judge says disputed facts warrant a trial
At a hearing in Oakland, U.S. District Judge Yvonne Gonzalez Rogers stated that there was “plenty of evidence” suggesting that OpenAI leaders had promised that the organization’s original nonprofit structure would be maintained. She ruled that those disputed facts should be evaluated by a jury at a trial in March rather than decided by the court at this stage, as noted in a Reuters report.
Musk helped co-found OpenAI in 2015 but left the organization in 2018. In his lawsuit, he argued that he contributed roughly $38 million, or about 60% of OpenAI’s early funding, based on assurances that the company would remain a nonprofit dedicated to the public benefit. He is seeking unspecified monetary damages tied to what he describes as “ill-gotten gains.”
OpenAI, however, has repeatedly rejected Musk’s allegations. The company has stated that Musk’s claims were baseless and part of a pattern of harassment.
Rivalries and Microsoft ties
The case unfolds against the backdrop of intensifying competition in generative artificial intelligence. Musk now runs xAI, whose Grok chatbot competes directly with OpenAI’s flagship ChatGPT. OpenAI has argued that Musk is a frustrated commercial rival who is simply attempting to slow down a market leader.
The lawsuit also names Microsoft as a defendant, citing its multibillion-dollar partnerships with OpenAI. Microsoft has urged the court to dismiss the claims against it, arguing there is no evidence it aided or abetted any alleged misconduct. Lawyers for OpenAI have also pushed for the case to be thrown out, claiming that Musk failed to show sufficient factual basis for claims such as fraud and breach of contract.
Judge Gonzalez Rogers, however, declined to end the case at this stage, noting that a jury would also need to consider whether Musk filed the lawsuit within the applicable statute of limitations. Still, the dispute between Elon Musk and OpenAI is now headed for a high-profile jury trial in the coming months.
Tesla’s Elon Musk: 10 billion miles needed for safe Unsupervised FSD
Tesla earns top honors at MotorTrend’s SDV Innovator Awards
