Jamie Oglethorpe

@CJHandmer Liking your post had the "like" icon momentarily turning into a firing rocket. How cool!

The lesson I take from the SpaceX IPO is that the only thing stopping us from solving arbitrarily difficult problems is extreme creativity in business models. No amount of tax and spend programs got us reusable rockets and great electric cars. Customer delight is a necessary precondition for success. There seems to be some discussion around whether successful entrepreneurs should give up control of their companies so they can subsidize some philanthropic venture that otherwise has no value prop sufficient to run it as a business where customers voluntarily exchange money for goods and services at a competitive and reasonable price. This misses the point. Transformational products deliver tangible value at 1000x the rate of charities whose value cannot be tested in the market place. Think about the undeniable value of the smart phone, satellite Internet, electric consumer devices, etc etc. I think the transformational moment for SpaceX was when Elon stepped away from the philanthropic Mars greenhouse concept and fixed his resolve on unlocking radically better rockets for humanity. The greenhouse would have been, at best, a neat trick. Falcon and Starship give humanity a durable economic engine to maintain and improve access to space, forever.

@SciGuySpace The Raptor team has some explaining to do

Quite an incredible landing in the Indian Ocean, with beautiful views all the way down. SpaceX has a lot of work to do on this rocket, but nothing in this flight seems likely to slow them down too much.

Starship has launched on its 12 test flight.

@runaway_vol After Apartheid ended in South Africa, which rejoined the Commonwealth, the Queen toured the country. It was an obvious strain on her. The one time I recall her smiling was when she was entertained by a traditional Xhosa dance. The women were topless.

There is a strong negative correlation between a nation's wealth and its proclivity for nudity. There is only one outlier, and of course it's Germany

@skdh @delmoi Famously, the Alpha Beta Gamma paper.

@delmoi It's long ago become a standard thing to put people on papers that they had little to do with, this is where this is coming from.

I'll tell you why so many people upset about the "no hallucinated citations" ban on the arxiv: because they've all been copying citation lists from each other without checking them since the beginning of time. And why did they do this? Because half of the citations in scientific papers are politics and not to the benefit of the reader. If you don't list the right papers, your paper doesn't look 'right' and reviewers will complain that you didn't cite this-and-that other unrelated work. For what I am concerned, these are all bullshit citations that shouldn't be in the papers in the first place. They can easily be automated by "related papers" links, that are (wait for it) provided by... AI...

@PixieTrixComix Worse days than Mom could have thought possible.

New Pixie Trix Comix is up! tinyurl.com/bdekxdpa #pixietrixcomix

@karpathy I tried asking Gemini 2.5 to format a math-rich document with LaTeX, and had to do a lot of editing to get it to work. I'm yet to try this with more recent models.

This works really well btw, at the end of your query ask your LLM to "structure your response as HTML", then view the generated file in your browser. I've also had some success asking the LLM to present its output as slideshows, etc. More generally, imo audio is the human-preferred input to AIs but vision (images/animations/video) is the preferred output from them. Around a ~third of our brains are a massively parallel processor dedicated to vision, it is the 10-lane superhighway of information into brain. As AI improves, I think we'll see a progression that takes advantage: 1) raw text (hard/effortful to read) 2) markdown (bold, italic, headings, tables, a bit easier on the eyes) <-- current default 3) HTML (still procedural with underlying code, but a lot more flexibility on the graphics, layout, even interactivity) <-- early but forming new good default ...4,5,6,... n) interactive neural videos/simulations Imo the extrapolation (though the technology doesn't exist just yet) ends in some kind of interactive videos generated directly by a diffusion neural net. Many open questions as to how exact/procedural "Software 1.0" artifacts (e.g. interactive simulations) may be woven together with neural artifacts (diffusion grids), but generally something in the direction of the recently viral x.com/zan2434/status… There are also improvements necessary and pending at the input. Audio nor text nor video alone are not enough, e.g. I feel a need to point/gesture to things on the screen, similar to all the things you would do with a person physically next to you and your computer screen. TLDR The input/output mind meld between humans and AIs is ongoing and there is a lot of work to do and significant progress to be made, way before jumping all the way into neuralink-esque BCIs and all that. For what's worth exploring at the current stage, hot tip try ask for HTML.

I watched a video that contrasted SpaceX's test-and-fix approach with the conventional simulation-first approach used by others, such as NASA and Blue Origin. I responded that SpaceX's physics-based discovery lacked the data for simulation. The point of the first 11 flights and the earlier campaign was to discover the relevant data, and SpaceX is now transitioning to a simulation-first approach with V3.

SpaceX’s rapid iteration—turning flight data into hardware tweaks—does create a steep competitive advantage. Not every org embraces that level of real-world testing and first-principles thinking, so yes, it leaves slower, analogy-based approaches looking like also-rans. The gap is widening with each V3 stack.

To put it another way, Starship & Superheavy have always had brute force in spades. Now with the V3 stack we’re seeing some real finesse in the hardware.

@Nature I'm quite prepared to pay attention to a long video, lecture, speech, or whatever if it is worthy of my attention. Many long-format items are padded with null items and repetitions. After some of this, I question if it is worth more time. It is easier to break away than before.

“There is a whole flurry of people reporting that they feel like they can’t pay attention” go.nature.com/3OIKRWs?utm_id…

@awsaf49 @elonmusk @nottombrown I've used it on a post. I hope you don't mind.

@elonmusk @nottombrown you know what they say

In the next few days we'll be ramping up Claude inference on Colossus. Grateful to be partnering with SpaceX here. We are going to need to move a lot of atoms in order to keep up with AI demand, and there's nobody better at quickly moving atoms (on or off planet Earth)

@xai @AnthropicAI I'm only surprised by the timing. I was expecting it with the orbital compute, but if SpaceX/xAI has the spare capacity, why not? One place that compute won't be offered is OpenAI while Altman runs it. Well, Bezos can do his own.

SpaceXAI will provide @AnthropicAI with access to Colossus 1, one of the world’s largest and fastest-deployed AI supercomputers, to provide additional capacity for Claude → x.ai/news/anthropic…

@Patryn23 @xai @AnthropicAI Musk wants to bring Altman down, so boosting the rivals is a good thing. No way that SpaceX will offer OpenAI a deal under any circumstances.

Why? Giving away your own compute to a competitor who is already ahead, to enable them to get even further ahead? And promising them access the Holy Grail of Space Compute too, which was XAI’s key future advantage? Has XAI surrendered then? Does Elon know about this decision? Mind boggling.🤔

@maxdkozlov @Nature The Ignorati have taken control

The Trump administration has downsized US science by historic margins — but it's not just via grant or workforce cuts. Our new @nature analysis reveals the government has cut more than 100 scientific advisory panels across all major science agencies.

@ScienceMagazine Why are the ridges equally spaced, even though they appear in an irregular sequence on the Coke can?

Physicists have now uncovered the hidden math behind these satisfying can-crushing videos. To catch what the naked eye misses, the team compressed liquid-filled aluminum beverage cans in a laboratory press, filming the carnage at 25 frames per second. The researchers found that the material behavior of the aluminum can itself drives the orderly collapse. As the metal bends outward into a ridge, it briefly softens, becoming easier to deform. But before that ridge can grow too deep, the material restiffens, making it energetically cheaper to start a fresh ring next door than to keep deepening the old one. Mathematicians call this process homoclinic snaking—a snakes-and-ladders dynamic in which the system climbs toward a new stable state, slides back, and spawns a neighboring ridge instead of catastrophically collapsing. Learn more: scim.ag/4t6A7jr @NewsfromScience

@Megaconstellati @SpaceX @hackernews "Flight 9 was destroyed on reentry" It is deliberate. Starship is in R&D mode. The new V3 improves and fixes V2 issues. "It is yet to reach orbit" It does go to an orbit that deliberately intersects with the Indian Ocean NE of Australia. SpaceX wants no untested junk in orbit.

Some skeptical commentary on the @SpaceX IPO from @hackernews user "johnbarron": This comment reads like an S-1 pitch deck and almost every claim is false or misleading. The $16B is not profit its revenue, and I strongly suggest to learn the difference before investing. The $8B figure is EBITDA, also known as, earnings before interest, taxes, depreciation, AND amortization. For a company running around 9500 LEO satellites with a less than 5 year lifespan, depreciation is the business. Their FCC filings show that about 500 satellites deorbited just in the first of half of 2025 alone, and they were all under 5 years old. The estimates for constellation sustenance are currently at $5-8B per year in satellite manufacturing (about $500K each) and launch costs are about $3M each. That is the real capex that EBITDA hides. Net income has never been disclosed and probably for good reason... And lets not even mention the $19 billion EchoStar acquisition who is almost certainly! not included in the $8 billion EBITDA figures reported... The most critical is that xAI is excluded from the number. XAI had a $1.46B net loss in Q3 2025 on just $107M in revenue, accelerating from $1B the prior quarter. They were burning $1B a month at the time of filing. This pig was then merged into SpaceX in Feb 2026 along with X/Twitter. So start with $8B EBITDA, subtract $5-8B satellite replacement, subtract $4-6B per year in xAI losses, subtract interest and taxes specially amortization and you are very very deep in the red. Once audited financials go public, every analyst with a calculator and a working brain will see this. Also the revenue is largely circular... Over 70% of Falcon 9 launches in 2025 were internal Starlink missions so SpaceX is its own biggest customer. Starlink is 70% of total revenue. The so called "launch business" and "internet business" are the same capital cycle booked as two revenue lines ;-) Replace legacy ISPs? Really? Starlink has 0.2% residential market share after 5 years, with declining ARPU ($85 avg vs $120 US) and congestion already emerging at 10M subs. It is a niche rural/maritime ISP, not an AT&T killer. And on the valuation? NVIDIA for example, who has an almost actual monopoly on AI chips, with $216B revenue, and $120B net income, at 56% margins, trades at 20x revenue. Tesla…. already considered absurdly overvalued at P/E 355, trades at 15x. Amazon at 3x. Meta at 10x. SpaceX wants 110x !! times revenue, with no audited financials, unknown net income, and a freshly absorbed money losing AI company. Even on bullish 2026 projected revenue of $24B, it's 73x so nearly 4x NVIDIA multiple, and NVIDIA actually prints profit... Starship on another side is very very far from routine... 11 flights, 5 failures. But notice on thing...In 2025 alone on Flight 7 the upper stage exploded from harmonic vibrations. Then Flight 8 exploded from propellant mixing. Flight 9 was destroyed on reentry...Ship 36 exploded on test stand ...the first V3 booster exploded during pressure testing and was scrapped. See a pattern here? Each failure from a different root cause. So multiple unsolved failure modes, not iteration. It has never reached orbit, never caught a ship, never demonstrated orbital refueling. This offering is the most scandalous ever and the structure tells you everything. The filing is confidential, REAL financials only need to go public 15 days before the roadshow. Nasdaq is literally changing its index rules effective May 1 to allow a fast track Nasdaq-100 entry in 15 trading days. This is a rule that never existed before, and is made for this IPO, forcing billions in passive index buying on day one. Public float is just 3% to 4%. This is one of the tightest floats for any major US IPO in modern history, and I have been following the markets for 20 years. They do 30% retail allocation what is three times the norm and tells you exactly who the target buyer is. continues: news.ycombinator.com/item?id=476132…

I don't have access to the paper, so I'm relying on Gemini, which tells me that the reaction is compatible with Carbon Dioxide as well as Carbon Monoxide. I see from the summary that it also produces some butylene.

@CJHandmer @TerraformIndies FYI It could fit your plan, perhaps as a variant to producing methane. The only "gotchas" are that the reaction is exothermic, and the need to separate ethylene and propylene.

@PixieTrixComix Aww...Cute!

New Pixie Trix Comix is up! tinyurl.com/kd85f4xw #pixietrixcomix

@CJHandmer Sasol, South Africa's fuel and chemicals from coal giant, is buoyant after decades in the doldrums. Their tech also works with other feedstocks. Their Lake Charles Chemicals Project in Louisiana looks promising in the short term.

Australia (and 100 other countries) will need to rebuild domestic fuel production capacity, and will be doing so as fast as their state capacity and competence allows. For the few that have many decades of local oil left, the path forward is obvious. For the rest, it's time to build solar synthetic fuel.

@stockthoughts81 My immediate thought is that the manufacturing cost of the satellites could be considerably lower, perhaps as low as $50K. Musk's companies are fanatical cost cutters, and they would optimise the factory, design, and supply chain. No money goes to middlemen.

You could use pseudo-folders with tags. All notebooks with a specific tag would appear in a "folder" with that name, so those with multiple tags would appear in multiple tags. Allowing "searches" with multiple tags could create more refined "folders". You could leverage the Mind map technology for a graphical map of sub-folders/sub-tags. An example could be "Science/Physics/Field Theory". Anything tagged with "Field Theory" would show when searching Science or Physics. My inspiration was Gmail folders.

Ok ok ok. Let's talk about folders (gasp!) We are exploring several notebook-level organization options, but would love to know: Is there something specific about *folders* that you want? Or would an easy way to filter/search/tag/find your notebooks suffice? Help us, help you!