News
SpaceX Starlink launch suffers last-second scrub, ULA up next [update: double scrub]
Update: ULA has scrubbed today’s NROL-44 launch attempt after the weather at the launch site substantially worsened. The Delta IV Heavy rocket’s next shot at launch is now scheduled no earlier than 11:58 pm EDT (03:58 UTC), Tuesday, September 29th, just two hours after a SpaceX Falcon 9 is scheduled to launch the US military’s fourth upgraded GPS III satellite.
SpaceX’s eleventh Starlink launch of the year was scrubbed ~30 seconds before liftoff by bad weather, likely delaying the mission a few days and leaving ULA’s latest Delta IV Heavy launch attempt next in line.
Scheduled to lift off at 10:22 am EDT on Monday, September 28th, SpaceX’s 12th operational Starlink launch (V1 L12) nearly made it to liftoff before the company called the mission off, prioritizing mission success above all else. Given that SpaceX’s Starlink program puts the company in the unique position of being its own launch customer, the decision to let a relatively mild weather violation delay a Starlink mission by at least a few days is unintuitively encouraging.
It’s no secret that SpaceX has become the most successful private launch company in history and a commercial force to be reckoned with, handily overtaking United Launch Alliance (ULA) and Arianespace to acquire a vast majority of the commercial launch market share. Falcon 9 is on track to become the fastest commercial rocket in history to cross the 100-launch milestone and SpaceX is already well on its way to regularly out-launching entire countries with 20+ missions per year. The single biggest risk facing the company is arguably complacency and an infamous tendency known as “launch fever.”

At the cutting edge of spaceflight, constant, exhaustive vigilance is ultimately the only thing standing between a reliable rocket or spacecraft and catastrophic failure. Perhaps the single biggest threat to that vigilance is the somewhat understandable desire to avoid launch delays – a fact of life for rocketry that nevertheless costs time, money, and (to some) reputation. The term “launch” or “go fever” was originally colloquialized to describe the irresponsible managerial pressure to launch largely responsible for both of NASA’s catastrophic Space Shuttle failures.
Some (if not most) parts of SpaceX almost assuredly would rather avoid launch delays. The fact that the company continues to accept Starlink launch delays and respect Falcon 9’s limits strongly implies that SpaceX has found ways to prevent launch fever while still pushing the envelope of launch cadence and rocket reuse. Starlink-12, for example, was originally meant to launch on September 17th but was delayed ~10 days by strong ocean currents before being scrubbed seconds before launch on September 28th. Combined with the fact that SpaceX is technically free to accept more risk on its own Starlink launches, compounded delays will inevitably test the limits of any organization’s resolve.

While the argument that SpaceX is technically the only direct stakeholder in Starlink missions is a bad-faith argument that could easily be made to push for increased risk tolerance, it’s only true in a vacuum. A Falcon 9 failure during a Starlink launch would still have major consequences for all of SpaceX’s customers, particularly delaying critical NASA astronaut and US military launches until a lengthy accident investigation is completed. SpaceX executives and managers involved in launch go/no-go decisions clearly understand this and act accordingly.
Starlink-12 will likely be recycled for another launch attempt sometime after ULA’s next Delta IV Heavy launch attempt and probably after SpaceX’s own GPS III SV04 mission for the US military, scheduled no earlier than (NET) 12:02 am EDT (04:02 UTC) and 9:55 pm EDT (01:55 UTC), September 29th, respectively. Catch ULA’s latest NROL-44 launch attempt at the company’s official webcast below.
Check out Teslarati’s Marketplace! We offer Tesla accessories, including for the Tesla Cybertruck and Tesla Model 3.
Elon Musk
Tesla AI boss reveals how big Optimus is going to get
Tesla’s Optimus chief corrected himself on X, confirming a staggering 10 million robot production target.
Tesla’s Optimus program has a new number attached to it, after Ashok Elluswamy, the executive who has run the humanoid robot program since June 2025, posted a three word correction on X Thursday, “Correction, 10 million robots.”
The line clarifies the long term annual capacity Tesla is building toward its planned second Optimus production line at Gigafactory Texas, a figure Musk has cited repeatedly since last year’s shareholder meeting.
The scale is worth noting, because ten million robots a year would mean Tesla building more units annually than most countries sell in new cars. Tesla has framed this as a second line, not the first. The buildout is happening in two phases: a roughly one million unit per year line inside Tesla’s Fremont factory, installed on the floor space vacated when Model S and Model X production ended earlier this year, and a much larger dedicated facility under construction at Giga Texas that broke ground on its first steel structure in May. That Texas facility is the one Elluswamy’s correction refers to, and is expected to reach volume production sometime in 2027.
Correction, 10 million robots https://t.co/0z4nyQNTzp
— Ashok Elluswamy (@aelluswamy) July 30, 2026
Tesla Optimus project fires up as Musk sees production line progress
Elluswamy took over Optimus from Milan Kovac last summer and has spent the months since talking up the program’s trajectory. Elon Musk has also floated the ten million figure at Tesla’s 2025 shareholder meeting.
Ending Model S and Model X production to make room for the first Optimus line was one of the more consequential manufacturing decisions in the company’s recent history, retiring two flagship vehicles in favor of a robot that has yet to enter mass production. Musk has previously estimated per unit production costs at $20,000 to $25,000 once Tesla reaches a million units a year, though he hasn’t said what that cost looks like at ten times the volume.
News
Autonomous vehicle red tape gets slashed by Trump Administration
The Trump Administration today made several key moves to help with the deployment of autonomous vehicles by cutting overreaching red tape that has stifled growth and innovation for years.
The moves, which were put forth by the National Highway Traffic Safety Administration (NHTSA), aim to grant temporary exemptions to at least one company currently, although that could expand in the coming months. Additionally, it will work with organizations to develop standards and a sound but efficient regulatory landscape.
Zoox is the only company mentioned explicitly by the Trump Administration in its press release announcing the new terms today. They will receive a temporary two-year exemption that will allow the commercial deployment of up to 2,500 vehicles annually for two years.
There is a potential exemption for Robomart, Inc., which “requests a temporary exemption from certain FMVSS No. 500 requirements for a low-speed vehicle operated by an ADS without a human driver onboard. NHTSA will publish a separate notice seeking public comment on its merits once the initial evaluation is complete,” the agency said.
Here are the five new terms that Secretary Sean Duffy has implemented through the NHTSA today:
- Allow Zoox to commercially deploy its robotaxis through a temporary exemption.
This temporary exemption will allow the commercial deployment of up to 2,500 vehicles annually for two years, subject to an enhanced, adaptable oversight structure that can evolve as Zoox’s technology advances. - Accelerate development of first-ever AV performance standards through a partnership with SAE Industry Technologies Consortia (ITC).
This partnership will fund a three-year, $5 million “A2SCEND” consortium, bringing together experts to gather data and accelerate creation of the first-ever AV performance standards. This project will inform a single national standard for AV safety to eliminate the patchwork regulatory landscape that has stifled innovation for years. - Publish an interim final rule that allows vehicles manufactured prior to an exemption to be eligible for a commercial deployment exemption.
This rule will modernize the application process and improve access to exemptions for innovators, including AV developers, by granting the NHTSA Administrator the discretion to apply temporary exemptions to vehicles manufactured prior to the effective date of an exemption grant. - Streamline the application process for Part 555 exemptions by updating guidance and soliciting feedback from the public.
By updating the Part 555 exemption process—which allows automakers to temporarily sell a limited number of non-compliant vehicles, primarily to test new technologies—NHTSA is aiming to create a more flexible oversight structure for exemptions and summarize recent AV framework activities, including expanded exemption pathways, streamlined crash reporting, and ongoing efforts to modernize Federal Motor Vehicle Safety Standards (FMVSS). - Establish a new Federal Docket for public feedback on NHTSA’s updated safe AV development and deployment guidance.
NHTSA is updating its technical guidance for AVs for the first time since 2017—focusing on key safety areas like emergency responder interactions, safety management systems, remote assistance, and post-crash behavior to help the industry scale up driverless deployments safely.
Additionally, the NHTSA said it has modernized some safety standards by proposing updates to:
- FMVSS 102 – Transmission shifting
- FMVSS 103/104 – Windshield defrosting and wiping
- FMVSS 110 – Tire placards
- FMVSS 135 – Braking systems
- FMVSS 101 – Controls and displays
- FMVSS 108 – Vehicle lighting
- FMVSS 111 – Mirrors and rearview display
- FMVSS 126 – Electronic stability control systems
- FMVSS 201/208 – Sun visors and warning labels
These changes aim to make the regulatory process for autonomous vehicles more streamlined and efficient, which could help the U.S. gain dominance over autonomous vehicle systems moving forward.
Elon Musk
Elon Musk has a crazy prediction about AI in two years
Elon Musk is, in many respects, one of the biggest and most influential figures in modern-day artificial intelligence.
Given that Tesla, SpaceX, and xAI are all looked at in their respective fields as leaders to an extent, each of them has a heavy influence on the future of AI, even though two of them are not thought of, at face value, as AI companies.
Musk has grand expectations for what is to come with AI, not only as a form of assistance to make human lives easier, but to make humans multiplanetary and solve some of the biggest issues that face us today. But even he is astounded by AI’s pace of progress.
He believes that in two years, AI will be so mind-blowing it might be unrecognizable.
Given that AI from 2 years ago feels so old that it should be in a museum, then obviously AI 2 years from now will be mind-blowing https://t.co/TcsKZ8o8OE
— Elon Musk (@elonmusk) July 30, 2026
This progress can be seen in a variety of ways, but perhaps the most popular way people have shown AI’s progress, especially on social media, is through an incredibly arbitrary way of watching Will Smith eat spaghetti:
The progression in AI of Will Smith eating spaghetti (2023 – 2026) pic.twitter.com/VDv82mB5gs
— internet hall of fame (@InternetH0F) February 10, 2026
This is a great way to show people how AI is improving, especially from a perspective that examines how it can manufacture images and video from prompts. AI is an incredibly complex concept, however, and it goes much deeper than Will Smith eating Italian food.
Musk’s most widely adopted method of AI is likely Tesla Full Self-Driving, which impacts millions of people as they utilize it to increase safety with their travel. Musk has routinely pushed incredibly aggressive timelines for self-driving, especially unsupervised.
Perhaps this perspective is why he feels that things will be solved in a timeframe that is much more aggressive than most of us would think. Regardless, the progress of AI is moving fast, and it seems that Musk’s expectations for it could be high.
But if it can actually achieve full-length motion pictures and even more realistic production value, it will be hard to distinguish between reality and AI very soon.

