GIT seams to be optimized for network of trust. With one person at the top approving what gets merged into the release.
This person of course does not do all of the verification, other then broad strokes of what the change does, and who wrote it, reviewed it and tested it.
I feel like companies do not want a large tree like structure for their development teams.
Without a network of trust it can become mob rule, which is what this article appears to be describing.
I'm just so happy to see this level of progress. This another big step for opening up space. To think that one day this will be considered normal. 150 Metric tons sent on a fully reusable rocket.
So, reusable is supposed to reduce the cost. But the space shuttle was reusable and it has been shutdown because it was too expensive. What is the differences between the two?
SpaceX builds vehicles. The Shuttle was “reusable” because they needed a term between the default for transportation capital expenditures (e.g. trains, planes, cars and ships) and the modified missiles that defined post-War spaceflight. “Reusable” in the Shuttle’s context meant months of specialist overhaul time and the cost of a Falcon 9 launch in SRB booster replacements alone [0].
At the end of the day, in 2010, “the incremental cost per flight of the Space Shuttle was $409 million, or $14,186 per kilogram” [1]. ($591mm and $20,512 in 2024 dollars, respectively [2].) SpaceX’s prices per kg are around $3,170 on Falcon 9 [3] and $1,520 for Falcon Heavy [4]. Starship should bring those costs below $1000.
It might, but it's also at a scale where people can dust off the old plans for orbital rings and ask if this time the economics work out.
(My guess is the economics are fine, but the politics would kill it on earth, so the moon or mars will get one, but that's just an interested amateur opinion).
Term of art for this is "design for demise", i.e. make everything of pieces small enough and materials ablative enough that there's less than 1 in 10k chance that any debris will survive to the surface.
IIRC, above something like geostationary they tend to decay upwards? Though the old orbital ring white paper wasn't suggesting anything like that, this was an alternative to needing to go so high in the first place.
(I may be misremembering or getting confused with a thing specific to tidal locking?)
That is indeed what you get with tidal forces - bodies closer than geostationary orbit lose angular momentum and decay inward, bodies further out steal angular momentum from Earth and move outward.
I suppose the same effect is there with satellites much smaller than the moon, but it would be tiny.
I wonder if they could have an orbit high enough to move away from earth with some kind of drag cables dangling from them into low orbit to counter the outward movement. Would that work?
These orbits have vastly different speeds though. Consider a high geosynchronous orbit vs. something like ISS which goes around the Earth in an hour or two.
but you are towing the cable and you'd only get so far down for the counteraction of force from gravity to pull on the cable. Would speed or friction on a cable be a problem.
Presuming it’s in LEO. When and if we ever get around to building these things they probably won’t be in LEO at least not for long. Some of them might not even be constructed from materials launched from Earth.
The main difference is that this is built by a private corporation who can't afford to throw money away, while the Space Shuttle was build by the government, and moreover it had to fulfill a number of conflicting requirements, and commercial profit was not one of them.
But on a more technical level. I think the vertical landing is the main difference. Vertical landing was obviously known and done by NASA, this is how the lunar modules landed on the Moon. But doing it on Earth, with vehicles weighing hundreds of times more, I don't think the world had that technical readiness a few decades ago, when the space shuttle was designed.
And another major difference is the mass manufacturing idea. From the start SpaceX planned for getting to mass manufacture its rockets. The Falcon rockets are much cheaper than any other alternatives even if you remove the reusability.
Then it's the methane burning engines. This was pure old fashioned engineering progress. SpaceX's engines are miracles of rocket engineering. Aside from that, the fuel choice is extremely smart. Methane is better than all other fuels, except for hydrogen. Hydrogen was the fuel of the space shuttle, but it's very tricky to work with. It has very low volumetric density, so the tank of the space shuttle was absolutely humongous. Hydrogen needs to be stored at an absurdly low cryogenic temperature, so this adds to the complexity. And that tank was not reusable, so it adds to the cost.
In order to land as a glider, you'll need wings, landing gear, doors, rudder, stabilizer, flight controls, streamlining, all the structure needed to support it, and a heat shield for all of it. All that complexity has to work reliably, too.
All of that adds tremendous weight, complexity, and cost.
Yeah, methane is kind of second best (or worse) in many parameters, but most importantly, it's cheap, abundant, and easily and safely storable and transportable, and it does the job.
And while not the reason, also on Venus! Venus seems like a very interesting colonization target - gravity almost like on Earth, and there is a place in Venus atmosphere where temperature is around 30 degrees Celsius and pressure is 1 atmosphere (Earth); and human air is a lifting gas in Venus atmosphere. As a bonus, interaction of Venus atmosphere with the Sun produces a magnetic shield.
Staying in orbit means risk of catastrophic failure on puncture and doesn't provide access to the heat of lower layers of Venus atmosphere that can be used as energy source or carbon source. You also have to think about heat management - big colony means gigantic radiators. And most importantly no gravity and no magnetic shield - these make Venus imho better colonization target than Mars.
On Venus, a puncture doesn't immediately destroy anything - because of equal pressure on each sides of the balloon wall. You have more than enough time to put on a protective coat and fix it.
And building more living space is much easier + you could source the material (carbon) on site.
Its high cost would (should!) have killed it regardless, but its low reliability was going to be a huge problem too, and arguably it's the lack of reliability that finally killed it.
The high cost should have killed the project before it ever flew, but that's not how governments behave.
I did the math, and the impact is not negligible. One single launch releases the equivalent of 5000 tons of CO2. Elon wants to get to the point where there are thousands of Starships, each doing a few trips to orbits per day. That would be more than one millions launches per year, or more than 5 GT of CO2-equivalent. That's about 10% of the worldwide emissions today.
One million launches per year seems to be adequate trade for 10% global emissions. This level of technology implies we are able to reduce emissions elsewhere.
I think it’s ludicrously optimistic to think that this would substitute for reductions elsewhere. What possible
mechanism could reduce 10% of global co2 emissions when many of these launches will be tourists and starlinks?
1 million launches per year implies at least a decade or two of development. There is a lot that can change in that time - for example energy production can move towards renewable and nuclear. Few decades more and we might get fusion too.
The limiting effect for Starship launches won't be CO2 (direct air capture could counter that), it's injection of water into the stratosphere. Ballpark I think the limit would hit at about 100,000 launches/year.
Spacex might be a private company, but this project is funded by NASA, meaning the American taxpayer. Approved by a person whose last act was this approval before leaving NASA and joining Spacex (effectively putting money in their own pocket).
It is also yet to be seen how Starship will ever be profitable (outside of spending government money), who is going to pay for those launches and for what purpose. Other than Starlink, of course.
Partially. They have a fixed-price contract to land humans on the Moon, and notably got that contract because they severely undercut the other bids and were the only bid that actually fit within the available budget: they bid $2.94B, while Blue Origin bid $5.99B and Dynetics $9.08B.
That 3 billion is also much less than what they're spending on the project.
With a payload volume of 8m diameter by 22m height you could fit a James Webb size telescope inside with minimal folding. The sunshield (21.2 m by 14.2m) would only need to fold along one axis and the mirror (6.6 m) could be monolithic instead of having to fold, probably only requiring the mounting points for the primary and secondary to be hinged. This shouldn't be discounted because it makes telescope design much simpler and less expensive.
It also allows for launching individual space station modules that have almost the same volume as the entire ISS in one launch.
Their plans for refuelling on orbit with tanker versions of the starship open up the entire solar system to unmanned missions with much shorter timelines and much higher payload size and weight.
The fact the entire system is re-usable will make it both cheaper and faster to use than any other launch system.
All of this combined mean that it won't just be countries and space programs bidding for space on launches, it puts space within reach of many corporations and some private individuals. This isn't conjecture, it's already happening with the Falcon 9. Starship will make it even more accessible.
Just park the starship as the sun shield. Or two starship, or an origami starship that unfolds for more surface area, your own personal sun umbrella made from a starship.
About 10% funded by NASA. Starship is a >$10B program; SpaceX is getting $3B for Artemis of which >2/3 is for operational tasks and moon-specific stuff that SpaceX aren't relevant for SpaceX's goals of LEO and Mars.
The problem is that launch costs went down fast but satellite costs haven't gone down as fast and still have long development timelines. The other problem is the market for satellite services hasn't developed as fast as anticipated, except for starlink.
Starlink for all intents and purposes is the market for satellites now. All the other launches are nice to have extras.
Now personally I’m looking forward to NASA, ESA and JAXA to launch outer solar system probes like new horizons but with tons of fuel left in the tank to safely make orbit around there.
Having enough lift capacity to take a shot at putting a pair of telescopes out far enough to exploit solar gravitational lensing to resolve exo-planet surfaces would be a hell of a thing. Orbital refueling would mean we could reasonably build something big enough to be able to boost out that far (would still take decades to arrive).
The whole design process for them is based around launches being expensive and taking a long time to plan. It will be very interesting to see what happens when the whole process gets used to launches being relatively cheap and frequent. No need to spend years making sure the design is perfect and will definitely last a long time if you can launch a new one in a week if you make a mistake.
Things can only be cheap if you mass produce them. That tends to require standardization of components, and inevitably standardized components are a compromise between requirements, where up until now, saving mass was a critical requirement. If you don't have to care nearly so much about mass and volume, then that opens up many avenues for much cheaper standard satellite components.
Exactly. Private companies like space X would not exist if NASA didn't deliberately make the market for Private space companies. That's what governments do, make markets.
Think of the word ‘reusable’ in this case as less a binary descriptor but more of a scale of reusability.
Yes, both systems are reusable, but there are key differences in the refurbishment of the systems that partly explains the cost difference. It took more labor, resources and time to refurbish the shuttle. Also consider rapid reusability was a stretch goal when it was being designed, but we have come a loooong way since, spacex in particular has had it as a driving competitive differentiator for years now.
Another big difference is that NASA post Cold War was a skilled jobs program, with an incentive to do distributed, high overhead work to appease their bosses (congress), while SpaceX has the opposite.
> Yes, both systems are reusable, but there are key differences in the refurbishment of the systems that partly explains the cost difference. It took more labor, resources and time to refurbish the shuttle.
Starship uses essentially the same ceramic heat shield tiles as the Space Shuttle, so the fact the Shuttle had so much trouble with refurbishment doesn't mean that SpaceX has solved these refurbishment issues with the Starship upper stage.
Though the Starship lower stage, which contains the most expensive engines, doesn't have this problem. Since it doesn't need a heat shield. So partial reusability should be pretty realistic.
Shuttle's tiles were each unique. Starship is mostly clad with identical hexagonal tiles which can be mass produced and eventually refurbished by machine. A robot already welds on the tile fittings.
More of an ignorant assumption. I asked ChatGPT now, and got this:
> In terms of shapes, the tiles were not uniform. In fact, there were over 17,000 different shapes used to fit specific areas of the shuttle's body. Each tile had to be individually manufactured and shaped to fit a precise location due to the complex curvature of the shuttle's surface. The unique shapes were necessary to ensure that every part of the shuttle received the proper protection against the extreme temperatures during re-entry.
Please don't cite LLMs for factual questions. They are prone to confabulation. Why not type the question "How many heat shield tiles did the Space Shuttle use?" into Google?
Since I mentioned I got the info from ChatGPT, people can decide for themselves how much they trust it.
Note that the question here is how many uniquely shaped tiles there was, not the total number.
This is interesting because if you have to manufacture and keep in stock 17,000 separate tile shapes, that will be vastly more expensive than SpaceX who, from what I hear, only uses a singe hex shaped tile everywhere.
The tiles are very similar; the attachment system is very different (a big part of why Shuttle's were a pain to maintain) and Starship's simple shape means most of the tiles are the same (the ridiculous number of SKUs was another factor in Shuttle TPS costs).
Shuttle itself was refurbishable, but not rapidly re-usable. It was also incredibly expensive to build and refurbish. A shuttle launch also utilized boosters that were not re-usable.
Starship is supposed to be (and clearly well on the way to being) fully rapidly re-usable. That means all stages (in this case two) are re-usable, and that the capital and time required to get either stage flight ready again after a flight should be minimal.
Said another way – it is cheaper for SpaceX to build an entirely new Starship + Booster than it was to refurbish a Shuttle between flights, by a factor of about 4x ($90M for a Starship+Booster / $400m for Shuttle refurb).
The boosters for the Shuttle were reusable, but it turned out the cost of refurbishing them was similar if not higher than simply building new ones. Plus it contributed to the Challenger crash, see "Mr. Feynman Goes to Washington", https://calteches.library.caltech.edu/3570/1/Feynman.pdf
Percentage of reusability: boosters of shuttle cannot be reused, maintenance of shuttle itself is also very expensive (heat shields were pricey). whereas the starship stack has higher reuse percentage and allegedly cheaper to maintain.
The shuttle was not even reusable by any modern metric, the main tank was always expended, the boosters had to be recovered, fully disassembled and cleaned.
I'm not even sure the SRM case segments could be easily reused, given the tremendous stress. They were made of a very high strength steel (maraging steel, with a yield stress of something like 250,000 psi) operated with a safety factor of 1.4.
You've got a lot of responses on the difference of reusability, but the shuttle was also more expensive because it had to carry a lot of capabilities with it every time. If you were launching a satellite, you were carrying along the crew compartment and a couple astronauts. If you were bringing a few astronauts to the space station, you brought a cargo bay. And in either circumstance, you brought big wings. Starship can be filled with all cargo. And if you're just changing crew on the ISS, you could... not use Starship and launch a Falcon 9 instead. One of the mission profiles required by the Air Force for the shuttle was that it be able to rendezvous with a satellite, put it in the cargo bay, and return to Earth, all under 2 orbits and along a path that avoided flying over the Soviet Union, which required a rather large turn in-atmosphere to make it back to landing on the west coast.
One of the drivers was the need to Abort Once Around on a polar orbit launch from Vandenberg. The launch site rotates to the east during the orbit so cross range was needed. No such polar orbit launch (abort or otherwise) ever occurred, though.
No one has answered with one of the biggest issues with the shuttle: each one was extremely custom. Every single heat shield tile was unique to a specific position and a specific shuttle. There were probably over a hundred million individual components in the Shuttle, and with many critical ones being custom, the time to refurbish it for a new launch was much longer.
This is in contrast to something like the falcon, which has a very standardized mfg process and components, which allows for really rapid iteration
It is a very daunting problem no doubt, maybe the hardest one, but I also don't doubt that the end form of Starship will require way, way less than 20 000 man-hours per turnaround (as the Space Shuttle required) to refurbish the heat shield. They simply cannot afford that.
The Space Shuttle was not fully reusable as the biggest single part, the orange tank, was destroyed every time. But more importantly, the orbiter and boosters needed 2+ months of refurbishment/rebuilding after every flight.
One of the design goals of Starship is for the booster and ship to relaunch with zero refurbishment. To literally land over the launchpad, refuel, and go back to orbit within hours without people even approaching them. The heat shield is the biggest risk to that goal IMO, and we saw today that it definitely sustained serious damage despite improvements. But if they ever get there then per-launch costs will be a tiny fraction of the Shuttle with 6x the payload.
Falcon 9 has already proven that partial reusability is economical. SpaceX has dominated the entire worldwide launch industry and their competitors are nation states with no need to make a profit.
The difference is that they have already proven to be the lowest cost and most reliable launcher due to reuse. This is them lapping the industry with second stage reusability.
Reusability increases costs if you don't reuse often enough.
The shuttle would have been much, much, cheaper per launch if it had flown more often. The expected costs for the shuttle included a range based on how often it flew which turned out to be reasonably accurate. They were much worse at predicting which end of the range they would be flying in. At the rate they ended up flying they had the extra costs of reusability without any of the benefits.
Starship is ludicrously expensive, but still much cheaper than even the best case for the Shuttle, and it has a guaranteed source of launches to help it benefit from resuability.
The turnaround time for Shuttle was 2-3 months, while building a brand new rocket takes like 1-2 years. Although for the cost of Shuttle, we could have built a whole bunch of expendable rockets, pipelined with a regular launch cadence, and probably also gotten some cost savings through economies of scale.
Beyond the fact that they eventually did quit, the shuttle program was the public face of the US space program, and part of having more than one way to launch military stuff.
The military rapidly gave up on the Shuttle after the first accident. No shuttle was ever launched from the pad on the west coast at Vandenberg (plenty of F9s have been, though.)
They were but the cost of refurbishment was almost more than just building new boosters. Thats why SLS is currently using Shuttle Derived Boosters but not worrying with reuse.
The Shuttle consisted of the shuttle (orbiter) itself, the external tank (not reusable), and the two boosters which could be reused after ocean recovery. The orbiter itself was slow and expensive to reuse since (among other things) all the heat shield tiles were inspected and 30-100 replaced between each launch. I don't know how much work was done to the engines between launches, but SpaceX's parts and cost reduction on the Raptor engine have to give it an advantage there.
StarShip consists of the Super Heavy booster that we saw "caught" today, and the StarShip (orbiter) itself. Having the booster return to launch site vs requiring ocean recovery should potentially increase cadence and reduce cost of reuse. StarShip is also meant to be reusable, although it remains to be seen how that will pan out. On the previous flight there was burn through from inadequate heat shielding - maybe we'll see an improvement with today's vehicle. I'd expect SpaceX to iteratively arrive at a quicker and more cost effective orbiter reuse procedure than NASA had with the shuttle, but how quick remains to be seen. Of course they are planning many of these to go on one-way trips to Mars rather than being reused.
> The orbiter itself was slow and expensive to reuse since (among other things) all the heat shield tiles were inspected and 30-100 replaced between each launch.
Worth noting that Starship's heat shield is very similar to the one of the Shuttle. They actually got the manufacturing method from NASA.
That's why I remain very skeptical about the easy refurbishing of Starship. Initially, way back around 2016, the plan was to vent liquids to create a cushion around the ship. That sounded more easily reusable.
Yeah. I think they originally planned to use ceramic tiles only for certain spots and still transpiration cooling for the rest. Then they fully switched to ceramic tiles. In an interview with Everyday Astronaut on YouTube, if I recall correctly, Elon Musk said they first believed the ceramic tiles to be lighter.
SpaceX took advantage of tech from both the US and Russia, including the experience with the Space Shuttle. They have better computers, better metallurgy, advanced 3D printing and their own experience with the Falcon 9.
There is no guarantee that it will reduce the costs that much, but will all that experience, the chances are success are higher than with the Space Shuttle.
One big issue that isn't talked about much and that SpaceX takes very seriously is simply a lack of demand. There is only so much stuff you want to put in space. Satellites are expensive, and even with disposable rockets, the launch is only a smaller fraction of the cost. It is already a problem with the Falcon 9 as they have a bunch of rockets and not much to do with them. Starlink, orbital refueling, and crazy ideas like earth to earth transport are all ways to address this problem.
It was a problem for the Space Shuttle too, they couldn't achieve the economies of scale they planned it for. It was supposed to fly for routine maintenance missions but it didn't work out.
the space shuttle was "reusable". It had to be taken apart and meticulously cleaned and tested and basically had to be rebuilt after each flight, in a process called turnaround. SpaceX's rockets are much closer to what you'd consider reusable.
I wonder how much of that difference is because the space shuttle was human rated from the start: F9 eventually got there, but only after plenty of "testing in production" with disposable payloads.
The other big difference, an elephant in the room grade difference I think, is that SpaceX reliability was developed with memories of a reusable vehicle failing mostly due to turnaround costs and risks on everyone's minds. That clearly wasn't the case when the space shuttle was designed, they were the first and enjoyed the privilege of making all the beginner mistakes.
I had an email exchange with Homer Hickam, before SpaceX existed, where I remarked that the shuttle design looked like a giant kludge, and a winged reentry vehicle was a fundamentally bad design, for various reasons which I enumerated elsewhere in this thread.
He emailed back that he agreed with my reasons and had argued that case with NASA in the early stages of the shuttle program.
I was under the impression tha they also thought this themselves but got "persuaded" with the prospect of more money if it could handle certain payloads of use for defense?
The case for the Shuttle was so marginal they needed every prospective customer they could get. The military was dragged into it unwillingly, but did give their requirements when forced to commit. They backed out as quickly as possible after the Challenger disaster.
The majority of damage to shuttle's TPS apparently came from foam strikes from the external fuel tank. Superheavy's optimized profile certainly helps here, since there are no large cryogenic tanks hanging ominously over the TPS while being shaken violently by solid rocket boosters.
To answer this oversimplified question with a simple answer, the Space Shuttle couldn't be a more different vehicle than this one. It truly is a comparison between apples and oranges.
Let's start with the fact that it was designed in the 1970's. If you had a Cadillac DeVille from 1970 it would get 8-12 miles per gallon. Just the mere fact that the design is about 70 years old makes that vehicle too expensive to operate, and that's before we even start talking about other issues with the design (performance, safety, reliability, etc).
The main thing is... Space shuttle wasn't all that reusable. It had to be launched with a massive rocket and two massive boosters that just fall into the ocean.
Estimated costs of the fully loaded cost of the shuttle program ended up at $600 million to $1 billion per flight. The refurbishment costs per flight and man hours/staffing were astronomical.
It actually would have been a lot cheaper if they had gone for serialized, mass assembly line production of Saturn V class disposable rockets to launch piece of space stations, satellites and manned missions into low earth orbit.
The shuttle wasn't fully reusable, just for starters. The boosters were reusable with a lot of refurbishment work. The center tank was expended every time. It was very expensive. Only five shuttles were ever made, which means that no effort was put into automation of production of engines etc., everything was custom, and everything required great care to save the sunk costs.
Manufacture and maintenance contracts for the Shuttle were deliberately spread out across many companies and states, especially in key congressional districts. It was a jobs programme; waste was a feature not a bug.
More fundamentally, there were contracts. SpaceX does things itself; there's no legal friction internally. This gets back to the "Theory of the Firm" for why firms exist in the first place (transaction costs).
The need to codify what work is to be done in contacts is antithetical to SpaceX's rapid development processes.
Because people are reflexively averse to government spending unless there's a billionaire making profits on the way through thanks to 100 years of academic capture by Austro-libertarian economists.
Is that 150t of payload or total? What’s the cost in fuel alone (let’s ignore maintenance and operations costs for now)? I’m trying to get a feel for the relative scale compared to today’s commercial flight.
They previously threw around a number of around $1M per flight, as mostly fuel costs.
Also, while 150t is the target payload capacity, the current test vehicles are closer to 50t in payload capacity, there are revisions in the pipeline based on data from these test flights which will bring it up to 150t.
To put this in perspective: at 150t/launch, if a launch is $1M, then for the cost of an SLS launch (at least $2B) Starship could launch 300,000 tons, about the mass of three Nimitz-class nuclear aircraft carriers.
None of the vehicles have demonstrated any payload capacity yet. 50 tons is the on-paper capacity only, and seems quite high given how little fuel is left when the bring an empty starship to orbital altitude. I assume that as the engines and launch procedures get more efficient, they will start being able to bring stuff to orbit (and quite a bit of stuff, too).
They've actually been having to dump propellant in order to more accurately test what a Starship in orbit would be like, given they're not flying with a payload that would consume that propellant on ascent, but that they still want to launch with a full tank.
The dumping of this excess propellant actually caused an explosion and loss of vehicle on the second test flight.
That's what they said about the second test flight (and the third), but the webcast recordings looked a lot more like fuel leaks to me, and that is in line with Starship and early Falcon's past issues. I'm going to press X to doubt that the dumping narrative is the truth, since a nice face-saving white lie is in every corporation's handbook.
That's a weak rebuttal. It's not disputed except by the lunatic fringe - starship has carried a payload (although not quite to orbit, very close to).
Whatever that means is what it means, the ship was out of control and I don't know whether or not the mission requiring that payload was successful, but the fact is that it did carry the payload.
Methane is about 900-1500$ / ton. About 1000 tons is used for the launch in addition to 3600 tons of lox. That should be a bit cheaper than methane per ton. Ballpark, the propellant might cost around 2M$.
A modern airliner on a long flight might burn around 80 tons of kerosene. It's slightly cheaper than methane. Call it 75-80K$.
Indeed. You sometime see an argument that launch to space is expensive because of the propellant and therefore energy required. And as you note this argument is utterly wrong.
Am I mistaken, or are there distance/payload combinations for which Starship is cheaper per pound on a point-to-point basis than air transport, even setting aside 30 minutes versus 12 hours? Isn't that the non-intuitive outcome of ballistic trajectories?
No. But it gets within an order of magnitude, which is remarkable. Economy class on starship could be priced similar to first class tickets on transoceanic flights.
There are no dollars in the laws of physics. It's connected ultimately to productivity of all the activities involved, and there's no obvious upper bound to productivity.
Just the current market price for fossil methane, which of course goes up and down. But I'm assuming SpaceX gets their methane on the open market.
It's quite possible that SpaceX has access to some cheaper methane source. Texas produces a lot of it.
And SpaceX has speculated about eventually switching to some renewable source by e.g. synthesizing methane. In which case that would boil down to cost of electricity and carbon. I don't see that becoming cheaper short term but that could happen long term.
Falcon 9 is still the most advanced rocket flying real missions to date. The only thing close to it is Blue Origin, which didn't even have their maiden flight yet.
Is their timeline too optimistic? Yeah, but if the industry still catches up with F9 and they are close to having something a lot more advanced it really doesn't matter.
Even without reusability nothing comes close to Starship's cargo capacity. If you don't have to put a lot of engineering into getting things as small and light as possible you can put things in space a lot faster.
And, eventually, if it becomes cheap enough to put people in space for the task, you can plan to actually maintain things there rather than design for extreme reliability.
I have found that what developers consider easy to read varies. So unless you have a lead developer with the political clout to enforce a standard you will end up with a code base with a very inconsistent style.
Unfortunately in the corporate world the people with the required clout rarely want to spend the time necessary to enforce a style.
I've noticed teens using snapchat to just send quick videos. Just one sentence and then send. Is there anything like this for work, maybe where I can do a really fast screen recording?
Worked at a large global financial company 10 years ago in tech. It was well known that the bonus pool was finite per division, but it was also allocated on ratios based on performance reviews and retention. Everyone knew the more people, the smaller the pool -- but then again, tech was a "cost center" so it wasn't like adding technology people (with the exception of quants) earn the bank more money.
I suspect if you incentivized managers to keep their teams small, through a bonus pool, it would definitely result in smaller teams, but also with the danger of overworked ones. That seems self-fulfilling though, if you're overworked but you know there's a reward with it.
That all being said, most places don't have bonus pools so it's a mute point.
Did you ever do a school project where yourself and maybe one other person did all the work, and everyone else (2-3 folks) mostly just tagged along on the ride because they were assigned? Or maybe even made it worse?
And everyone got the same grade?
That happens in corp land all the time too, and it’s extremely demotivating.
thats usually called a 'profit share' (though I've only seen it based on net profit and not gross revenue). It was quite nice on Wall Street back in the 80s...Especially since everyone got an equal share rather then the managers favorites getting the bulk of 'discretionary' bonuses.
Paired with a nice xmas bonus and year end review/raises, it basically meant no-one quit because you were never more then 6 months from a fat check.
The quote: `I wish upon you ample doses of pain and suffering’ really hits on the nose for me.
My Dad had a successful business after a rough childhood, failing high school. He worked hard and built a successful business. He also really did wish ample doses of pain and suffering on me. And he got his wish.
He always thought it was good that I was bullied, that I also failed school, and that everything in life was generally hard for me. He told me this so much that I also believed it, well into my 20s. That is was good I was a failure and suffering.
It took a decade to build my life to a point similar to my classmates that did not suffer the way I did. And I'm not ahead of them even now.
I really wish my Dad would have gotten therapy instead of internalizing his anger, and believing that is what made him successful. It would have saved me a lot of time.
This maps almost 1:1 onto my experience, except for the fact that my father achieved some measure of success having never gone into business.
I almost didn't graduate high school and made a mess of my time at university, dropping out entirely my senior year (drugs and alcohol fueled by deep self-loathing made this an easy choice at the time).
Like you, I wish my dad had realized what he was doing in raising his children the way he did. His circumstances were deeply cruel. My relationship with him now continues to be marred by my understanding that he doled out the same punishment to his children and justified it by telling himself he was making us tougher.
In the end, it also took me nearly a decade to reach some level of parity with my peers. If that resilience is what got me here, then fair enough. But the juice wasn't worth the squeeze.
Intergenerational trauma echoes, and the kids spend their lives putting themselves back together and finding their blind spots.
I was gobsmacked today by Rod Nordland's revelations on Amanpour that were a degree more intense but strikingly similar to my own childhood, and how I somehow
also share gravitation to extreme sports and dangerous situations.
Heh. I had some of that "character building", and I credit that I don't seem to have as fragile a temperament as some to it. It worked for the Germans and Huns and Mongols. Too bad we can't run the counterfactual and see if we would be the same across different timelines.
Since we only view this at the population level, there is a big question: Is "character building" improving those who need it to be stronger, or just filtering them out of society?
With all the massive life improvements we have with technology we could be investing so much more into our kids. Jobs could be flexible with short work weeks, and we could use that time to invest in kids. Parent, volunteer, mentor. Remote work done on a school site giving even more time to help kids learn.
Having worked remote most of my kids' lives, I'd say there are other challenges. Parents need time away from their kids too, and if every parent in society is working then how are kids supposed to get the adult attention and allow the parents some child-free time?
When I was younger kids played with others in the neighborhood, and there was always a homemaker parent in every household. Now every adult has at least a part-time job. Kids still play in the neighborhood, yet weekly instead of daily. More often they are in daycare, school, or staying home.
how are kids supposed to get the adult attention and allow the parents some child-free time
not everyone is going to work all the time. obviously someone has to be available for the kids but not all the time either. the point is that life improvements, technology and flexibility makes this more easy.
Has to be explicitly said, lots of folks expect a village without asking if the village consented towards the effort (n=1). Managing and openly communicating expectations derisks disappointment and suboptimal outcomes.
Not wanting to be a parent is one thing, but people who don't want to be involved with kids need to work on themselves, that's a deeply antisocial trait - your own existence depended on everyone around you being involved with you to some degree.
That is an opinion, not a fact. Freedom of association and to be happy can include not interacting with children you choose not to, and optimizing for happiness is important (versus a social contract requiring otherwise). No one will optimize for your own happiness besides you.
I have kids, but fully respect people who don’t care for or want to avoid time and interaction with kids. I respect their boundaries, that is what I advocate for here.
I’ve had over a decade of therapy, so I’m fairly confident in my position on the topic.
it is one thing to ask someone who doesn't want to deal with kids for help taking care of them, but quite another to expect them to tolerate kids and behave in a manner that gives the kids space they need. (like not smoking near kids, etc)
in a city i can't choose my neighbors, but some neighbors don't like kids and will complain if they are to noisy because they are playing soccer in the yard, to the point that they force the building management to disallow it, which then takes a lawsuit from the parents to remove that rule because such a rule is in fact illegal since kids playing is natural had has to be tolerated.
if someone feels that kids playing infringes their boundaries then they do need an attitude adjustment.
Agreed. Kids need the space to develop as you mentioned, and it’s somewhat trivial to determine where to live to avoid children using granular census and school availability data if desired.
it’s somewhat trivial to determine where to live to avoid children
given the discussions of how much more housing most cities need, i actually think this is not trivial at all. not wanting to live near children pretty much comes down to not living near people.
> I have kids, but fully respect people who don’t care for or want to avoid time and interaction with kids
How would this sound:
"I fully respect people who want to avoid time and interaction with black people"
Why is okay to say "I hate kids", we would not think kindly of someone who says "I hate Women/Jews/etc.".
Kids were exploited, historically, just like any other minority - we used to send them into the coal mines and many didn't come back.
Even today, we discriminate against young people all the time. A 17 year old is not allowed to vote, but come 18 and they may be forced into the army and sent to fight a war, without ever having a say in the matter.
They are the last minority that we world does not respect, and has no plan to respect as the world is becoming more geniatric (look at the age of congress) whils moving into a climate calamity.
And I am not writing this as a young person, not any more.
I feel that's why many people, especially families prefer detached homes. Sound proofing is guaranteed by the air gap. And if there is a problem with soundproofing you can solve it yourself by making changes.
Perhaps row houses are best? Good density, and with a solid concrete wall seperating units to stop sound and fire from spreading?
This looks impressive. Much more than even the Boston Dynamics demonstrations.
Flipping a pancake is extremely difficult because each pancake is different. I know that these videos must be cherry-picked but to be able to train a Robot to do this just by demonstrating feels like a massive leap.
Flipping a pancake was done in 2010. What looks impressive for humans is easy for robots and vice versa:
https://youtu.be/W_gxLKSsSIE?si=HDyNXe1Ys_eFXiVU
Another case in point: robot juggling was done in 1990s and to date we do not have a robot that can open any door reliably like a human. Kind of like Moravecs Paradox
To be fair it is far more complex for a robot to grip a spatula and use that spatula on a griddle than to use dynamic motion to flip a pancake in a pan.
Solving any one problem with robotic manipulation isn’t all that hard. It takes a lot of trial and error, but in general if the task is constrained you can solve it reliably. The trick is to solve *new* tasks without resorting to all that fine tuning every time. Which is what Russ is claiming here. He’s training an LLM with a corpus of one-off policies for solving specific manipulation tasks, and claiming to get robust ad hoc policies from it for previously unsolved tasks.
If this actually works, it’s pretty important. But that’s the core claim: that he can solve ad hoc tasks without training or hand tuning.
> He’s training an LLM with a corpus of one-off policies for solving specific manipulation tasks, and claiming to get robust ad hoc policies from it for previously unsolved tasks.
It seems clear that many people do not understand that this is the key breakthrough: solving arbitrary tasks after learning previous, unrelated tasks.
In my opinion that really is a good definition of intelligence, and puts this technique at the forefront of machine intelligence.
Flipping a pancake in a "random kitchen" would be much more difficult and have many of the same issues as the door problem.
It's hard to point to a single thing that would make "flipping pancakes" intractable, it's sort of the other way around, to usefully flip pancakes in the same way as a person takes a lot of skills chained together.
The "door problem" is a sort of compendium of many real-world skills, identifying the door, understanding its affordances and how to grip / manipulate them, whether to push or pull the door, predicting the trajectory of the door when opened, estimating the mass of the door and applying the right amount of force, understanding if there any springs or pulls on the door and how it must be held to traverse through it. Etc. There are also a ton of things I'm missing that are so fundamental one tends to take them for granted, like knowing your own size and that you can't fit through a tiny doorway.
I think you can ramp towards the "door problem" in difficulty by slowly relaxing constraints. A video linked above (not article) shows "can flip a pancake successfully with a particular pan (you are already holding) and pancake with a fixed camera and visual markers". Ok, now do it in varying lighting conditions. With no visual markers. With different camera views. Different pancakes. Real pancakes (which are not rigid, and sometimes stick to the pan). Different pans. Now you have to pick up the pan. Use a stove. Different stoves. Identify griddle vs pan and use the right flipping technique. Find everything and do it all in a messy kitchen... eventually you're getting to same ballpark as the "door problem".
physicist here (so very naive on these topics) - I’m wondering how to compare the steps you mention regarding the door problem (especially the predictive ones, e.g. about the trajectory of the door as it opens, etc) with how humans open doors? Surely people don’t stop in front of a door and begin planning things out, rather they seem to go for it and adjust on the fly, is this an approach that won’t work in robotics? Why not?
So classical robotics yeah, people used to write code for each step of opening a door. Practically speaking you would probably not do motion planning on the door, you would just code it up with a bunch of heuristics like, try to be over here in relation to the doorframe because that's a good opening spot and will probably work. Ok you're in the right place? Now, move gripper towards the door handle... etc. Bunch of hacks. Put enough hacks together and you can kinda sorta open (some) doors. Oh this is a SLIDING door? Damn we forgot to code for that...
The way things are going is sensors (cameras, force, etc) and neural networks. You let the robot try a bunch of ways of opening doors, sometimes it doors itself in the face, eventually it'll figure out good places to stand based on what the door looks like. The more doors you make it try to open hopefully the better it gets at generalising over the task of opening doors. The hacks/heuristics are really still there but the robot is supposed to learn them.
> Surely people don’t stop in front of a door and begin planning things out, rather they seem to go for it and adjust on the fly, is this an approach that won’t work in robotics? Why not?
Yeah, figuring out how to do this is basically "the problem". Most people don't have a sense or feeling of "planning things out" as they open a door because we reached "unconscious competence" at that task. We definitely have predictions of what is going to happen as we start opening the door based on prior experience and our observations so far. If reality diverges from our expectations we will experience surprise, make some new predictions, take actions to resolve the surprise, etc.
Not sure that anyone has ever studied how people open doors in detail, it'd be interesting. I bet there are a ton of subtle micro behaviours. One that I know is, if you hear kids running in the house it is a good idea to keep a foot planted in front of you as you approach the door, because those guys will absolutely fling or crash doors open right into your face.
Thank you, great answer. As soon as I had asked my question I realized that we must have a lot of unconscious behaviours. Very interesting points about surprise/expectations. And top marks for the advice about kids
I was responding to address why the "door problem" is more difficult than "pancake flipping under controlled conditions".
(I also ignored that door opening is generally done by mobile robots of a certain weight class which tend to be more expensive than a stationary arm with enough strength to pick up a spatula or hold a pan).
There is a steep difficulty gradient from "works in the lab" to "works under semi-controlled real world conditions" to "works in uncontrolled real-world situations".
Not just touch but proprioception. Robots in human environments will have to be better at proprioception than 98% of humans. If I bump into you it’s typically anything from annoying to a meetcute. I’m a pretty big guy, but if you had to chose me to step on your foot or somebody else, it’s probably me you want, because I will shift my weight off your foot before you even know what happened (tai chi) because you will barely notice.
If instead your choice is your high school bully or a robot, well for now pick the bully. Because that robot isn’t even being vicious and will hurt more.
> Because that robot isn’t even being vicious and will hurt more.
Rodney Brooks at the MIT AI Lab was a big advocate of something called "series elastic actuators." The idea was was that you didn't allow motors to directly turn robot joints. Instead, all motors acted through some kind of elastic. And the robots could also measure how much resistance they encountered and back off.
MIT had a number of demos of robots that played nicely around fragile humans. I remember video of a grad student stopping a robot arm with their head.
Now, using series elastic actuators will sacrifice some amount of speed or precision. You wouldn't want to do it for industrial robots. And of course, robots also tend to be heavy and made of metal, so even if they move gently, they still pose real risks.
But real progress has been made on these problems.
I think you're probably right, and those non-linear systems are going to make me have to increase my estimate for how long it takes for a robot to go from 5 year old child to ninja physicality. The more complex the feedback mechanisms, the more complexity there is in, for instance, screwing in a screw as fast as possible.
The robot won't take any enjoyment out of it, and won't laugh at your pain. Won't post about it on social media. Isn't going to try and fuck your ex or sister or mom.
I'm pretty sure that if I had never opened a door before and I saw somebody opening a door in a video, I would immediately know how to open doors just by watching the video. And that would be any door, with any kind of door handle. Not because I got superpowers, but because I'm average-human.
So, the moment your system needs this kind of data and that kind of data, oh and btw it needs a few hundreds of thousands of examples of all those kinds of data, that's very clear to me that it's far away from being capable of any kind of generalisation, any kind of learning general behaviour.
So that's "60 difficult, dexterous skills" today, "1,000 by the end of 2024", and when the aliens land to explore the ruins of our civilisation your robot will still be training on the 100,000'th "skill". And still will fall over and die when the aliens greet it.
I think their robot has a way of converting touch to a video input. The white bubble manipulator has a pattern printed on the inside that a camera watches for movement. (see 1:58 of the video).
And here I thought manual labor jobs were safe for a very long time. I really hope people at the policy level are thinking about what it looks like to have a world of people that don’t have any work to do.
This person of course does not do all of the verification, other then broad strokes of what the change does, and who wrote it, reviewed it and tested it.
I feel like companies do not want a large tree like structure for their development teams.
Without a network of trust it can become mob rule, which is what this article appears to be describing.