Maybe they ran it through OP's de-sloppifier to see how well it works. And look how well it did! The slop detector doesn't suspect a thing! I can already smell the value this thing is adding to the world.
I’m still finding context is king, even with the best models.
For example, I had Fable review Astra’s output yesterday, and it found some issues and fixed them. Passing the fixes back, Astra then uncovered additional issues with Fable’s fixes (and yes, this will go on ad infinitum if you let it, but these were “real” issues).
It seems the big story here is the reduced Luna pricing. It’s a fantastic model that can handle most automation needs (though I still use the big models for day-to-day development).
I worked on (a very small part of) the ExoMars rover (now called the Rosalind Franklin[0]).
It was supposed to launch in 2018, then was pushed to the early 2020s on a Russian rocket. For obvious reasons, it got pushed again, now launching in 2028.
The state of the world isn't great for space exploration, but I'm hopeful this mission will be revived at some point in the future.
Would it make sense to send better robotic analyzers there than to bring the samples back here? Fighting gravity and kilometers involves spending many billions on return; a lesser amount could be spent on surface activities.
The issue is the size of some of these analysers. For instance, a synchrotron, which is required to non-destructively generate nanometer-scale 3d images of the samples are 3-4 times the size of a football stadium
We would need to move people and cargo back and forth eventually, so best to get started on that ASAP. Not to mention sample return has been done from the Moon & asteroid to great success, its a proven concept on how to get the most from the samples.
Science should stand above politics . Europe and Germany are crazy to develop a great x-ray satellite „eRosita“ together with Russia , shoot it in space , performs great, but then stop operating it because of politics . It is tax paid , so who decides ?
Sure! I worked on the digital logic that connects the rover to one of the scientific instruments on board, specifically the Mars Organic Molecule Analyser[0].
As an outsider who is very interested in space exploration I'm very flummoxed by the state of space exploration.
With the price of launches so low compared to only 20 years ago due to advancements in launch technology and the increased cadence of launches it seems like it should be easier to do these kinds of missions and that there should be more, cheaper missions that are willing to take risks because the per mission cost is so much lower.
Can you comment on why that doesn't seem to be happening? Why aren't we covering Mars in cheap rovers and satellites?
For both operating mars rovers, the launch vehicle cost was about 10% of the total. So even if the providers paid you to launch, you'll still be picking up 80% of the remainder. Cut the development cost by 10x? Still in the hundred million dollar range. So still only large goverments can afford. And none of them would approve a bunch of rovers, they'd just see a discount on the one rover they already don't want to give you - "you have one, isn't that enough? Why do you need two?"
Not just mass optimization either. Intense effort and cost are expended through verification because a single payload simply must work. But using commodity hardware with reduced verification effort, launching many, and tolerating some amount of attrition has resulted in reduced costs for Starlink.
Testing of using commodity parts was part of the Mars helicopter and it worked well enough that making many of them seems plausible. I'd love to see multiple drones flying faster/further than rovers. Just not really sure what instruments would make them useful.
That's another great example. And ridesharing on the mission allowed them to prove it very inexpensively and with much lower risk than a dedicated mission.
I think the question of technical possibility answers differently from the question of organizational feasibility. Starlink is optimizing costs because they have a financial incentive to do so. The less each satellite costs, and costs to maintain, the more income they make. I fear incentives are not so direct in governmental orgs. Outside contractors performing most of the work have outsized influence, and incentive to inflate costs. Budgetary decision makers all want to see investment in their own districts.
I think Starlink and cubesats have definitely proven the approach can reduce costs.
The engineering does not seem as challenging to me as figuring out how to get such a policy approved and pushed through all the administrative and decision making layers without watering down and budgetary inflation.
Maybe if we launched 30 Starships carrying 30 rovers each. With only 30 units, the prices would likely be similar to current commercial satellites, which are already built on standardized platforms.
I'd go even more radically low cost, and kit the first lander out with wifi, and a half dozen each of every commercially available robot dog, humanoid, and wheeled or tracked robot of any interest. Rumble in the Mars Robodome. Bit of publicity for the existing manufacturers who may be willing to pay for the privilege, and who will undoubtedly learn valuable things about how their designs fail and can be improved. And you figure out in one stroke if there's a COTS platform you can buy for $30k that gets you 80% of what you need. If any of them function for any period of time, you use them to prepare the space for future landings. Rinse, repeat. After a few iterations, you have a half dozen manufacturers with commercially produced platforms with a few strategic modifications to which you can bolt your science instruments.
All I can think about now is a bunch of e-waste littering Mars. At some point, other visitors will show up to investigate Mars and come to the conclusion that the Earthlings are just so trashy their trash even started filling another planet.
The surface of mars is roughly the size of Earth's dry landmass. They'd have to spend quite a lot of effort looking for it to find it. And at that point it'd probably look more like a treasure trove of useful thermoplastics, refined metals, magnets, and other useful things otherwise difficult or expensive to source on Mars directly.
I think if they had a big f-ing rocket at their disposal and could bundle a couple dozen of them MIRV style there would be no political downside if a bunch went splat so long as a bunch didn't
"look guys 9/24 of the rovers are transmitting back, great success"
0 to 1 is part of why the first Curiosity cost so much. If they kept the technology alive in practice, launching a new one every year doesn’t sound unreasonable.
In fact, that kind of standing order for a new Curiosity every year is exactly how you would keep the technology alive in practice.
They issued statements like that... aaand then built a copy of the rover for the same cost as the original one. And it only took them 10 years to do this :)
Yeah, build-to-print at the small scale of space missions never really works in practice because requirements always change and small changes in requirements tend to have big design impacts.
We've had 8 launches in the past 3 years, with only 10 in the combined 17 years before [1].
Seems that the reduced cost is being taken advantage of. And, make sure to exclude the drastic reduction in non-lunar launches in your cost comparison for lunar launches.
I won't find it now, but I distinctly remember watching a video about some big NASA event a few years back but when Starship Booster (first stage) had been flying already. Basically none one at the event even considered switching paradigm from ultra custom probes and landers built from unobtainium and unicorn blood in favor of the way cheaper and way less robust, almost mass market hardware, to be launched once every 10 months instead of every 10 years on a regular basis. (PS: I know all the laundry list of arguments against this approach, and they are not that convincing at NASA scale). And I won't even talk about Space Disgrace System for manned flight.
NASA is stuck in the past and does it consciously and voluntarily.
To my understanding people at NASA and people approving budgets like novel stuff. What’s the point of 100 identical rovers on mars? For politician - good publicity only for first 5 rovers, then it is only routine. For engineers, challenge of designing and constructing new cutting edge stuff is interesting and fulfilling, constructing 50th clone is boring. (If now you give money to top talent to do cutting edge, then after 5years you will have pool of talented people; might benefit next mission or some other company)
The point is that, while the first of them has 2 MPx cameras with crappy optics, the fifth one has 50 MPx cameras with automagic everything, and the tenths one has damn light stacking array of top tier cameras and on board AI to colour correct on the fly or some other magic stuff. And at the same time battery increases x10 times while losing half of the wight and thermal deficiencies, CPUs get better and more redundant, better memory, wheels are getting fixed in 1 year instead of 10 and the list goes on.
Or you don't need to craft a JWST with nail files anymore, but instead just lauch a swarm of mirrors to free float and emulate the whole thing. And if one mirror goes bad - just launch a new one for pennies (NASA pennies). Or if the whole project is nonviable somehow - no big deal, just launch next year a different architecture. Ok, not next year, but in like a few years. And not in three decades (LUVOIR anyone? hello? Is there anyone at NASA?), if on schedule, and forty years if not.
The list goes on. I get hedging risks and preserving institutional knowledge, sure. But not even discussing a different approach, which could be conducted in parallel, is very very bad.
... checks math ... Starship should have enough cargo capacity for this. One launch. We could put the whole senate up there. The house might take a couple launches unless we really pack 'em in. /s
We just need to setup a fake corporate sponsor meet and greet. We camouflage Ship to look like a hotel ball room, with a door for "private meetings" that leads to the cargo bay.
The nature of science and exploration is that every mission is harder than the last, which is a reasonable first order proxy for expense. For NASA specifically, there has also been a large focus placed on making delivery to orbit and the moon a commercial service rather than something that NASA has to do in-house which has held us back.
And as others have said, there is really no such thing as a cheap rover. As a recent example, there was recently a push by NASA HQ to refurbish a Mars rover engineering unit (basically a one-to-one copy of the one sent to Mars, kept at JPL for use in ops planning and issue debugging) into an operational lunar rover. The estimated cost to take this existing piece of hardware and make it fully flight ready and suitable for the new mission was hundreds of millions of dollars.
It is definitely picking up - you if look at the list of planned launches to the Moon for example, there are quite a few things launching in the next 12 months.
Also I think we need to start looking at smaller institutions and even private companies doing beyond LEO research soon - exactly because of the launch costs going down.
Yet at the same time, traditional big exploration players don't seem to be really able to make use of ti quick enough for some reason (politics ? inertia?).
A practical issue with government spending is that it's self motivated. Politicians' only job is to get elected. And one big way to do that is by ensuring that money goes to your district, creating jobs and opportunities. This is how you get things like the SLS, nicknamed the Senate Launch System. It was a tremendous waste of money, spending on the order of decades to build a launch system that was arguably obsoleted years before it completed, and will in any case be far too expensive to ever regularly use.
But the reason we kept throwing good money after bad is because the goal was never really about making a good launch system. It was all about sending money and creating stable long-term jobs in various states and districts, actively supported by the politicians representing those areas. So you had parts being built all over the place, in a way that just sends costs skyrocketing, but that's okay - because the rocket isn't the goal.
Of course the government itself is aware of this problem, much like fishermen are aware that overfishing is self destructive in the long-run, even if they are nonetheless individually motivated to do exactly that. So I think this was why the big push for commercializing space. And I think we'll ultimately see the exact same thing with deeper space missions and technologies. SpaceX's goal is Mars and I see no reason to think they will fail there. At that point, costs for various Martian missions - be they sample return, rovers, or whatever else - should be relatively negligible relative to today. Exactly as happened when we went from ferrying people to the ISS using the Space Shuttle, to when we started launching them on Falcon rockets.
Cynicism is a variety of compliance in advance. It creates the conditions it decries.
Obviously there are corrupt politicians. But also obviously there are politicians who actually want to advance the public interest. It is the job of voters to tell them apart and ensure only the latter become the custodians of state power.
Politics, especially in its modern form, does not incentivize nor reward those who would put the public interest ahead of their own political careers. On the contrary it effectively prunes the overwhelming majority of these people out of office long before they get anywhere near the higher levels of office.
There have certainly been a few exceptions, but they're few and far between - and tend to be because of special circumstance. For instance I think Bernie Sanders, before he tried to run for President, was a good guy. But that's because he's from Vermont, a state with a smaller population than El Paso, TX. Democracy works quite well with small populations, but not so well with with larger populations when the 'distance', speaking both metaphorically and literally, between voters and their representatives grows too far. It also tends to fail when societies become too divided.
As for the voter determining who's socially motivated and who's a silver tongued conman? I don't think that's at all realistic.
One reason is that launch is not as cheap as often stated.
Launch is cheap for SpaceX. But they still charge high prices for everyone else. It's often estimated that a Falcon 9 launch costs SpaceX $15-20m, but they charge $60m+ to customers.
America is being run by somebody who doesn't understand the value of science, so it is seen as a waste of money and worse, systematically destroyed. The problem is that the result won't become apparent for about ten years at which point Americans will be asking why everyone else is making money from science and technology breakthroughs and they're getting poorer.
One issue is definitely that there is no tangible (economical) upside to anything other than low earth launches and now maybe the moon, and the scientific exploration, while technically cheaper now, is still expensive - and requires a lot of public funding, which hinges on public interest. I don't think the general public is still as enthusiastic about space exploration anymore, people are far more skeptical about the upside given all the problem we have down here that occupy their minds constantly.
Yep notice that all the space stuff going on is about commerce not science.
Personally I couldn't give a shit about astronauts landing on the moon I want hundreds of drones flying around the solar system doing science but I'm the minority.
At least ESA isn't doing much propaganda.
Outside of specifics like the space shuttle program, launch costs have never been that big of a factor. Like 5-15% of a program's cost. If your satellite or rover costs billions to build and another billion to run for a decade, saving a hundred million (or even a couple hundred million) isn't that much of a game changer
Some types of launches to some classes or orbits have gotten cheaper. Not all launches to all orbits. I don't know if any commercial launch services can even do planetary transfer orbits.
In space everything is an orbital change. To change orbits you need to speed up and slow down, all of that requires fuel. Even using what are known as "low energy" trajectories that use a bunch of gravitational slingshots need velocity changes to adjust trajectories to get the right angle of approach to some large body.
Most commercial launch services focus on LEO because the secondary stages don't need to carry much fuel and their engines only need to fire reliably once to get into a stable orbit.
A planetary transfer requires multiple engine firings. That's much harder than a single reliable firing. It's not impossible but the testing and engineering is expensive.
A science mission often needs to beg, borrow, and steal to get built. There's one shot to accomplish the mission objectives. That's more expensive engineering and testing. Some mission profiles require specific planetary alignments with very narrow launch windows to boot. So there's a lot of expensive work made more expensive by a tight deadline.
The actual marginal cost of the rocket, fuel, and launch is a very small part of the mission cost. Halving the launch price doesn't necessarily affect the cost of anything else.
Starlink satellites are effectively disposable. If a Falcon 9 explodes or fails to get the second stage to orbit they're out maybe a few tens of millions of dollars. If a science mission has some mission-killing failure there's no funding for a do-over.
You're right I was forgetting the Falcon Heavy. But it's certainly not cheap when the government is footing the bill. The Nancy Grace Roman launch cost NASA $255m, far above SpaceX's prices for even an expendable GTO launch.
Don't mince words. Say it with your chest. The world state is being ruined by authoritarianism appeal to small-minded people's worst basal instincts, lack op apathy, and lack of curiosity. And unchecked capitalism, but I really shouldn't even need that 'unchecked' qualifier.
I find the idea that people can use LLMs to exercise their rights as citizens appealing however. Many people aren't aware of the rights they have, and LLMs are pretty good at surfacing some stuff without having to pay lawyers. Having to hire a lawyer is imo actually a huge way of gatekeeping people from exercising their rights. I heard a lot of local German public institutions are currently being flooded with people arguing their case with the help of LLM that they previously weren't really realistically able to do. So I don't see it all as bad.
>are pretty good at surfacing some stuff without having to pay lawyers
They are also really good at making stuff up as evidenced by the many, many, many examples you read in the news about actual lawyers using AI to write briefs that are full of errors and hallucinations.
In civil courts, you'd likely get more sympathy from a judge if you represented yourself and admitted your lack of understanding, rather than try to appear as someone you're not because you wrote some prompts and copied the output.
Current court systems around the world are just not built to handle the flooding of cases from the citizens.
The stupidest analogy is open source projects having a hard time accepting LLM generated PRs from the masses, because review process is the bottleneck.
No idea how to fix this, to be honest. In coding world, with some mental gymnastics, I can see code not being reviewed by people anymore. In courts, things generally have more consequences, and you can’t really roll back decisions that easily.
Then the solution is to add AI to the courts as well. Or something else, if you don't like AI. But definitely not limiting the right of the people to sue.
you could probably write a cool little gotcha of an SF short story about a barren wasteland of a planet that keeps broadcasting out legalese that's revealed to just be LLM chatbot lawyers pedantically arguing with one another about xeno legal doctrine
The legal system, which is a machine/technology by itself, will be eaten out. I wonder what will replace it. Botnet law arbitrage? – Personal assistants constantly negotiating with each other to avoid permanent civil lawsuits?
the cost of making a legal argument can collapse while the cost of reaching an enforceable, legitimate decision may go higher, which will gate the "justice" system even more.
It's really not. The legal system is slow and inefficient because it's a deeply pipelined system built to maximize the throughput of the bottleneck resource: judges. Judges are constitutional officers who exercise independent authority in meat space and thus are necessarily limited in number. The rest of the design flows from that.
If you got the judge, all the parties, and all the witnesses in a conference room together until the case was resolved, you could probably handle a lawsuit in a few months. But each judge has hundreds of cases pending before them, so that would never work. Instead, you get something like how a GPU works. You do some work on a case, submit the work to the court, then work on something else for a few months while you wait around to get the results back. Then you do some more work and submit it to the court, then go do something else for a few months while you wait to get the results back. A few months of actual work gets spread out over a few years that way.
Why stop there? The bottleneck is not some immutable force of nature. The number of judges is determined by legislative action. It is within Congress's power to allocate new federal judicial seats, and likewise at the state level with the equivalent lawmaking body (for most states, at least). Why don't they do so?
Because federal district court judges have a tremendous amount of power. You both don't want 10,000 of them running around and you don't want to water down the qualification for the position--if you did do that, then that would devalue their judgments, everyone would appeal, and you'd just shift the bottleneck up to the appellate courts.
At the end of the day, litigation is conflict resolution. You don't just need a decision, you need a decision from someone authoritative enough to bully Fortune 500 CEOs into submission to accept the judgment.
You presented the concern from my adjacent comment perfectly (“LLM performance: common law vs civil law” essentially). So is AI possibly just growing the “Reverence for Professional Experience” factor that plays such a big role for legal compensation here in the US?
Creating virtual products get cheaper, so shouldn't it even decrease if everyone vibe codes their app for less than a dollar instead of hiring a dev team for 50k or ordering a white label app or spending 3 dollars to buy an existing app?
I don’t know in the USA but in France, if it’s deemed that you launched a lawsuit knowing very well it wouldn’t succeed, you are susceptible to get a 10k€ fine. Even jail in serious cases.
Yes. And you(r lawyer) can collect lawyer's fees and you can be made to pay the court fees, if you lose.
IME the American legal system is set-up to discourage litigation, though. A common tactic is to bury your opponent in the threat of heavy damages or jail-time to get them to settle for what you were originally after, which courts are perfectly happy to facilitate because it gets a potentially lengthy trial off their dockets. They'll punish (or be biased against) whichever party seems responsible for not accepting a "reasonable" settlement.
Algorithmic abuse of the system to extract payments already exists in the form of the debt collection industry.
A common tactic is to bury your opponent in the threat of heavy damages or jail-time to get them to settle for what you were originally after, which courts are perfectly happy to facilitate because it gets a potentially lengthy trial off their dockets.
That's not how litigation actually works in this country. Unless the plaintiff can actually prove damages, there's no threat of "heavy" damages. The primary economic concern is spending a lot of money on legal fees, which is the primary trigger for getting businesses to settle instead of litigate. It doesn't work so well against people though.
They'll punish (or be biased against) whichever party seems responsible for not accepting a "reasonable" settlement.
That is false and any judge caught doing that would be removed from the bench, immediately. The judge does not exercise any judgement over what a "reasonable" settlement is as they don't and wouldn't know what is reasonable until the facts have been presented...at trial.
Algorithmic abuse of the system to extract payments already exists in the form of the debt collection industry
The debt collection system doesn't use the court system at all. Assuming you are referring to individual debts, not corporate debts, debt collectors have a very low success rate in court since (a) they have to prove the debt is valid, (b) that they now own the debt contract, and (c) they did not violate any laws in the pursuit of collecting on that debt. A and C are actually really hard for a third-party debt collector to prove.
Why the hell are they not fighting fire with fire? This is not sustainable. But it is annoying that the AI labs get to play arms dealer, selling to both sides.
They don't have a choice. If you're a firm and you work with IE a bank, the bank uses AI, their whole supply chain uses AI- your counterparty uses AI- it's a classic arms race with even more social pressure involved (since daily news keeps banging on about AI).
In case it's helpful to anyone, MiSTer is a project[0] that recreates classic computers and video game consoles on hardware.
What do I mean by hardware?
Rather than emulating the device in software, which can cause various issues (especially related to timing), by mapping the actual device logic to a field programmable gate array (FPGA), you can achieve an "exact" replica of the hardware.
It's a super cool project, and I highly recommend checking it out if you have an interest in old machines (Apple II, Commodore 64, etc.).
"exact" is still in quotation marks because things like open bus behavior still have to be explicitly simulated. It's still a very accurate simulator, not an exact replica - much more accurate than most fast software simulators can be, though.
Worth noting that timing-exact emulation is for, more or less, less than 32 bit consoles. The timing of newer hardware is so unpredictable that nothing can reasonably rely on details, which is just as well: it would probably not be possible to emulate it precisely at an acceptable speed.
PCs have been "unpredictable" like that almost since the beginning due to hardware variation, so emulating timings isn't much of a problem for emulators, as is the case in the featured article.
Check out the Superstation One [0]. Outer shell modelled after the original PS1. Optional optical disk dock. Simplified UI. Lots of outputs. Well priced.
Super cool! To be honest I'm not into emulation/retrogaming so much (even though I do have a Pimoroni Picade) but the fact that this exist as a production is very inspiring, thanks for sharing.
If you want something more powerful than MiSTer and have a bit of patience, there is the Replay R2 being developmed around the Agilex 5.
- Agilex 5 E-Series 043 (Replay2 phat) has 434,240 logic elements (LEs).
- Agilex 5 E-Series 013 (Replay2 lite) has 138,060 Logic Elements (LEs).
The website (https://www.fpgaarcade.com/news-2/) hasn't been updated in some time but the dev now has prototype boards that he's in the middle of bringing up. It is purpose built for arcade emulation and has tons of I/O, lots of memory (both SRAM and HyperRAM), and (on the phat) a PCI connector. One or both (can't remember which) accepts an ARM daughterboard for hybrid emulation. There's a Discord for it you can track down to follow its progress basically in real time as mikej progresses.
There are much cheaper clones of the DE10 available these days. For a little while fully assembled MiSTer's were available on AliExpress for around $150-$200 although it looks like maybe they dried up. There's also the SuperStation based on the MiSTer project for around $250 if you're patient.
Does mister work with kv260 boards "out of the box" or is the official mister target a de10 nano? (I love my mister, assembled oneba few years ago). I don't follow the project closely due to zero free time these days
You have to base your build around either the DE10-Nano or one of the clone boards from QMTech. If I were getting into MiSTer I would skip faffing around with dev board builds and just get a SuperStation One, and I say this as someone who has two MiSTer builds based on DE-10 boards. In my case I'm using a MiniITX motherboard that the DE-10 plugs into called the IronClad Plus. The SuperStation one is apparently awesome. Great IO options, reference quality analog video, and nice industrial design.
Both require having an understanding of what you are trying to emulate. There are systems which don't have an accurate specification so they are not implemented bug free.
LLMs are basically multi-dimensional magic mirrors.
Depending on where you point them, they can be incredibly useful.
They can even be useful when you point them at each other (though increasingly difficult to get good results).
I'm excited for the promise of RSI and a future where models have inherently "live" weights, but it's not clear to me that the transformer is more than a useful tool to help us get there.
I'm also aphantasic and dream normally, and I once got really into lucid dreaming for a ~semester in college.
My best successes would occur after drinking a cup of coffee and immediately taking a nap before a study session.
It was great, I "learned how to fly" and could "morph" my dream into whichever direction I wanted.
One thing stuck out though: whenever I would try and focus on something, I would see a black dot enter the center of my vision. The more I tried to focus, the larger it got. If I focused on an area attentively enough, such that the black circle became all encompassing, I would wake up.
After a few times of this, I could start to notice the focus, and then sort of relax out of it, and thus continue the dream.
It feels like the black dot is my real-life vision coming into focus, and "winning" over the visual aspects of the dream. It's always continuous from black dot to eyes opening.
I rarely lucid dream now, but I've noticed a similar thing happen whenever I try to read text - the black dot appears and I wake up.
Wish I still had unlimited time to go back to sleep!
One of the most horrifying recurring dreams when young was "geometric nightmares". This black circle would expand in size while a deafening silence grows. The volume is a void, and it's deafening.
An impending sense of doom as the expanding black void circle approaches. There's a real fear of oblivion, of ceasing to exist. Then you wake up.
I'm not aphantasic, but that's exactly the experience I get when coming out of a lucid dream. Except it's my expected surroundings not black. Gets a bit weird and inceptiony when it's the wrong surroundings! Given the number of cultural references along these lines, I suspect it's not uncommon.
Focus takes training, just like physical exercise, it is a super power to be harnessed.
You can train yourself up by looking at your hand, then building a body, then building a room, and maintaining all of it. When you are strong enough, you walk outside.
It's very clearly AI-generated, and thus a bit ironic (:
[0]https://sitefire.ai/slop-checker/r/UkRAuqW1y-1T2qhbRQfLt2Jq7...
reply