The closest we have to a "neutral rule" is "The Golden Rule":
"Treat others as you would like them to treat you."
It has a long history, both secular and religious. Pretty well everything flows from it, and if someone is doing "the wrong thing" whilst claiming to obey it then they are probably deluding themself.
The rule is suited to enforcement, since if someone does something "wrong", rather than immediately attacking them you can reasonably ask "Would you like it if I did this to you?". Limitation of power is built in.
I'd say expecting non-hypocrisy is also a top item. Someone says x behavior from a leader is disqualifying? OK, maybe that's their earnestly held belief and so worthy of consideration even if I disagree with their view. But if their behavior is the opposite/shows hypocrisy and suddenly isn't disqualifying? That should be a red line.
Because even the "golden rule" leaves a lot of wiggle room - for example, should marital infidelity (or x) be a dealbreaker for electing someone? Maybe, but looking at hypocrisy (and thereby actions) might provide better insight as to what those groups actually believe is the "correct" standard?
> The closest we have to a "neutral rule" is "The Golden Rule"
Mostly agreed, but a surprisingly large percentage of the population seems truly unable to feel empathy for others (and can even show glee at others being negatively impacted by something) unless it personally affects them (a prerequisite for "Would you like it if I did this to you?" to work?). They can be gung-ho about laws that negatively affect others but upset if it affects them. For example, the countless examples of "Community members vote for politician promising illegal immigrant deportation, then upset/disagree that long-time beloved community member deported", or "Voter voted for politician promising criminal-record-immigrant deportation, then upset/disagrees that partner with past criminal record deported".
What if I like to treat myself terrible and hurt myself everyday? It's okay for me to treat others that way?
Just because something has a long history (secular or otherwise) doesn't give it any merit by itself.
If we need some standard to judge against what is morally acceptable in the case I presented (no this doesn't apply if you like to hurt yourself and others) then you are back to defining a standard which cannot inherently be neutral.
> What if I like to treat myself terrible and hurt myself everyday? It’s okay for me to treat others that way?
No, that’s not what the Golden Rule said. It says to treat people the way you want other people to treat you. How you treat yourself is your own business.
There’s also a certain amount of making reasonable assumptions here. Obviously the intent is to get people to treat each other fair & well, not to set a logic trap and interpret this aphorism literally to mean you should hurt people if you’re okay with being hurt, even when you know other people don’t want to be hurt. Most people don’t want to be hurt. Obviously.
Yes, reasonable assumptions! But according to whom? The golden rule can be reasonable to you but not to me. But does that make me bad and you good? Not really. It just means we disagree on what it means to be a moral person.
Anyone who doesn't agree with your "reasonable assumptions" has bad morals can be twisted to mean any moral values based on what you mean by reasonable. For example, the church can say it is reasonable assumptions to burn witches and then twist reason itself to make it morally good as well. In that case how can we trust this system itself? If the answer is by using your own eyes and thinking for yourself, then why do we need to follow a specific fixed rule?
The reasonable assumption I was referring to is what the words in the golden rule mean, not what actions are moral or not. The golden rule is almost as far as it gets from a “specific fixed rule”, despite the name, and it doesn’t prescribe what’s bad or good or moral, and nobody upthread said anything about what makes you bad or good. It’s nothing more than a rule of thumb, an aphorism, a (perhaps extremely obvious) reminder that you shouldn’t treat other people in ways you wouldn’t like yourself, at a minimum. Is this hard to understand or something? The aphorism is already completely relative and subjective, there’s really no need to belabor that point, it’s built in.
How can you trust the system? What are you even asking? We have laws, a lot of them, and the golden rule is not one of them. But if you don’t hurt others, and others don’t hurt you, we’re off to a good start. Use whatever definition of ‘hurt’ you’d like, it doesn’t matter if it’s reciprocal.
The go-to example of a failure state in the golden rule that I am aware of is "thirsty person wants everyone else to jump their bones, thus the rule suggests that they should jump everyone else's bones".
The take-home being, as long as peoples preferences are different, the rule has some failure states based upon that.
And the wider example that this points to is that as long as everyone's values are different.. the rule loses a lot of it's luster. The more different another person's values are, the less you can rely on projecting your own values as a method to gainfully interact with them.
In place of that biblical aphorism (born from a monoculture no less), there is a lot of room available for us to build something more robust and sophisticated and the AI alignment problem offers a fine excuse to get cracking and try to formalize something already! ;)
Again, that’s an overly literal and straw man interpretation of the golden rule that seems nearly intentionally obtuse. The rule is not suggesting you should do anything that you know other people wouldn’t like. Jumping people’s bones is a very bad example, because that’s not something anyone should do without communication and explicit permission, otherwise it’s rape and it’s illegal.
If you understand and accept the reciprocal intent of the golden rule, via the above mentioned reasonable assumptions, not to mention the hundreds of years of examples and explanations, then it doesn’t really have a “failure state”. And, again, this so-called rule is not law, isn’t specific, and doesn’t cover all situations or all behaviors. It seems very silly to be nit picking this aphorism to death like this. It’s very simply a good faith suggestion to be kind.
The notion of fairness and reciprocity is not a religious idea, it’s common sense.
1. I believe GP was suggesting "golden rule" as the basis for governmental-wide rules to be enacted, not that each person should use unique rules based on their own "do unto me" preferences. (If actually applied, "What if this were done to me" also helps to get around the issue of people enacting rules only on other groups, as they wouldn't actually want to be personally treated the way they're proposing.)
2. Regardless, if basing rules on whether someone would want those rules to apply to them does not have merit, is there some better basis you'd recommend?
2. Yes, this is where I would push back because one can also assume there is no "better" basis in the strict sense that you can intellectualize morality/virtue only to a certain extent. And it is possible to argue that morality/virtue is often as much a psychological process as a purely intellectual one. So the idea isn't that the golden rule doesn't make sense or that it is objectively weak, but it doesn't map to the idea of actual morality in action. For example, a person may logically believe in the golden rule but that doesn't mean we should ignore the fact that the agent is not fully logical and will not always strictly adhere to a rule just because it is aware of it and aware of it as being Good. In that sense, what use is the golden rule if we are free to ignore it at all (which we often do)? That doesn't mean I'm saying the golden rule is objectively wrong again, but in ranking morality and deciding on very abstract ideas of what Good and Bad are, we ignore the actual psychological forces that are extremely relevant when it comes to the execution of morality in action. So to answer the question, the better basis would be to understand the psychological forces behind why we feel the golden rule is objectively a good and then examine them to understand if we are following the golden rule because we actually think it is good or because we live in a society where it is good and we will be shamed if we don't believe it. One might say it doesn't matter as long as people actually follow the golden rule but another interpretation is that the golden rule by itself doesn't tell us anything (or very little) about the moral character of the person applying it. So just expecting people to follow that rule after knowing it is placing too much trust in the human animal who is often wild and not thinking logically. It is then upon the animal to tame itself and it may do so out of servitude or out of independence and both might lead to equivalent surface values and match the basis perfectly as being morally good while in reality, both have very different meanings for their morals.
2 is, of course, mainly inspired by Nietzsche. I'm working on a short story where he and Plato debate whether a painting is objectively good or bad called A Canvas Called Virtue.
There's an obvious asymmetry in your rephrasing. The parents quote says "do unto others as they do unto you", not as you do to yourself. If you treat yourself terrible and hurt yourself, is it okay for others to hurt you too?
By logic, yes, if the hypothetical I likes getting hurt, they can also have a tendency to like getting hurt by others surely no? The problem is assuming these types of thinking doesn't exist and we only need to know the good to follow it and hence treating morality and values as a thing that stays beyond humanity itself in the form of eternal truths like the golden rule. But we don't think of morals on an intellectual plane purely. We think it is bad to hurt others because we want a good world where nobody is hurt, but that in itself is a stance where we assume everyone wants the same thing and if they don't, they just lack the reason and rationality to understand the obviously good values. But the mistake is assuming only reason rules how we see and execute morals in actual life.
It's the complete opposite of eye for an eye. If someone is being badly treated by others, most would want to be well treated instead, so the Golden Rule would say that you treat others well, even if others are treating you badly.
It would be interesting to augment the plotter with a Unimorph piezoelectric disk scanner [1], to enable finer movement so lines can be drawn closer together.
interesting! the axidraw v3 that I used for this (https://shop.evilmadscientist.com/productsmenu/846) has a 0.0125mm stepper resolution, which means that it is pretty close to not-worth-it to go any finer since the sizes of the scratches are already many microns wide. But certainly with better etching technology that would be a super cool update!
Alexander's results show a movement range of 1.25um, 10 times smaller than the 12.5um step of your system, so in its current form it wouldn't be suitable for interpolating motor steps.
Also potentially interesting is Alexander's "cut-and-pull" technique for making scanning tips with a single atom at the point. I can't find the original reference, but here's a description:
The single atom would likely immediately wear down if used for scribing, but even if it wore down to multiple atoms wide it might still produce a very narrow line? I wonder what the effects of using an STM on a photosensitive plate would be, whether the strong electric field at the tip would lead to very narrow photographic lines?
Most electrical/electronic items (radios, TVs, washing machines, ...) used to come with a schematic, or at least a block diagram, inside the device. Similarly, published integrated circuit data sheets contained enough information to understand how the chip worked to a low-level. This changed around the 80s from memory? There wasn't a word for "open source" in electronics, as it was just the unnamed default. It was all paper based, as it predated wide-spread online distribution.
My parents bought a relatively inexpensive television in 1987 and it came with complete schematics, printed out as an addendum to the user manual. I was more excited about those than the TV itself because I had just learned to read schematics from one of the Forrest Mims books at Radio Shack a few months earlier. This was in the US and I am pretty sure it was made by Emerson.
A dumb brand name dryer we got from Home Depot a few years back still has one of these. It was a breath of fresh air to unfold it and pretty quickly determine what part was bad with a multimeter and some high school electronics education.
More like early 90's. In the late 80's it was still common & easy to order schematics, data sheets/books & the like from manufacturers. Usually @ cost, sometimes as a free service.
In the early 80s, computers also came with technical informations. Even consumers oriented things like digital typewriters had programming and electronics booklets in the box. That didn't last :)
Many of the simpler devices still do, at least mine do, though my washing machine just buzzes at the end of the cycle and doesn't play music or send me a text.
As long as the AI devouring a book isn't making it harder for a person to get access to a copy of the book. There might not be many copies of some out-of-print books.
Perhaps Anna's Archive will be the last bastion of some books, especially if they manage to acquire the internal collections of some of the AI companies, à la Spotify. (ie. the parts of the collections that didn't come from Anna in the first place.)
> made by an equally obscure manufacturer, YiQingTeng Electronics, in Shenzhen, China.
Is this what is meant by "Shenzhen Speed"? Dump all safeguards and get any piece of crap out the door as soon as possible? This comment isn't directed specifically at Shenzhen, as other parts of the world do the same thing and call it "Move fast and break things."
Dropping engineering standards may increase short term profits, but it erodes people's trust in the world around them when everything "sort of works".
There's an argument that this low level continuous corrosion does more long-term damage than obvious safety flaws that are serious enough to prompt corrective action.
If not, I have anecdotal counter info: saw a lot of work of a small team doing hardware testing/evaluation for a hardware reseller. Often using Chinese sources. Typical "our brand is good, but has to be slightly cheaper than competition". Some manufacturers are really good. But a lot of them send hardware so bad that it will electrocute you instantly. Which is wild in that profession - the whole point of sending samples to client is for them to disassemble it to asses quality. This is literally the only scenario, so any lie is instantly discovered.
One company asked for immediate return of a prototype because a client ordered a batch, but they don't know how was it made. They had zero documentation and had to check what was inside.
> There's an argument that this low level continuous corrosion does more long-term damage than obvious safety flaws that are serious enough to prompt corrective action.
Unfortunately I just don't see much care being given to the long term well being of society right now. Not by many governments, not by many corporations.. it feels like everyone is in a "get rich by any means, fuck you I got mine" attitude.
Might as well not have any kind of society or rule of law at that point at all then. Zero trust societies, that's definitely the way towards a better life for all of us
First to market usually wins. After they've got their gizmo in every user's hands, who will buy a perfectly crafted replacement with E2EE and all the bells and whistles?
After her husband's death in 1992, Wendy Whiteley took to cleaning up an unused railway easement behind her house. It turned into a substantial garden. The railways refused to guarantee its future (it was on their land), but in 2015 it became an official park. Well worth a visit on a sunny afternoon if you are in Sydney.
In 2023 the NSW government committed to turn the adjacent harbourfront railway line into a public linear park. The next government shredded the agreement and has announced it wants housing everywhere. IMO it's prime land that is too valuable to be privately owned and should be publicly accessible. Anyone who disagrees should go for a harbourfront walk on a sunny afternoon. (Billionaires who's takeaway from the experience would be to buy the land for their own use don't count.)
The same battle played out in Cremorne in the 1890's. Today the park is one of Sydney's jewels.
There's a children's book that I've been reading to my daughter called The Curious Garden about a boy gardening on abandoned tracks. It's a reference to the High Line, of course, but it's interesting there's more than one like this. I suppose the end of the age of passenger rail in these countries offered opportunities like this.
I read his book Concrete Botany recently. It also mentions the deep mulch trick mentioned elsewhere in this thread. It's an easy read and a refreshing perspective.
I used the benchmarking binaries that ship with libfec. My own crate has a libfec-compatible C shim, so I can link the benchmark against the Rust crate. The benchmark itself reports time spent for a given number of iterations, so the throughput can just be extrapolated from that.
"How do you find the speed of the Rust version of your FEC vs, the C version?"
translates from "Australian" to "English" as:
"What is the speed of the Rust version of your FEC vs, the C version?
Though it was interesting to know how you do it. I'm interested in the speed, as I once looked into using Rust for a signal processing project, but ultimately went with C++ because the team wasn't familiar with Rust. At the time, it seemed to me that Rust had the potential to go faster.
tldr: My library matches or beats libfec's SSE2 assembly for convolutional codes and pretty steadily beats it in Reed-Solomon. For the convolutional codes, my crate is using a generic, templated decoder rather than hand-written assembly, so it was nice to see that I could match the performance.
I'm beginning to think that my '96 Toyota Starlet is "peak ICE". It has EFI, no infotainment, no power windows, no power steering, no automatic transmission. It is basically an engine, gearbox and wheels. An electric Starlet would be Nirvana.
Edit: Just read that Toyota is toying with the idea of an electric Starlet. It'll probably have an infotainment system and power "everything" though.
I'm not a mechanic but I have rebuilt an engine (common rail diesel with electronic injection) with a friend and do some work on my own cars. I don't really get the "everything needs to be mechanical" push. My previous car was mid 2000s, automatic, electronic basically everything. The reason why I scrapped it was because it had too must rust. The engine, transmission, and electronics were still fine after 200k and 20 years of harsh winters.
My current car had an electrical issue last year... but it turned out it was a mechanical issue. Something had chewed through the wires to the MAF sensor. Fixed them, and all good. It's no different from something chewing through a fuel line, except OBD was able to tell me what subsystem was not working correctly.
The problem I see with current cars is that the infotainment system is tied into everything - and usually the software sucks. If they just went back to seperate dashboard, AC controls and radio I'd have no issues with them.
My current car is from 2016, and I specifically looked for something that was like that. It is hybrid and has modern tech (e.g. radar cruise control), but everything is compartmentalised. First upgrade I did was replacing the headunit with a Chinese one that supports Android Auto/CarPlay. No functionality was lost, as the old head unit was literally just a radio and screen for backup camera.
The infotainment thing is regrettable but nowhere near as bad as the cellular modems phoning home, constantly spying, possibly raising your insurance rates…
Lucky for Big Brother they’ve blanketed this country (USA) in Flock cameras. New car, old car, they’re ready to pull over journalists in dealer loaners, people in the wrong place at the wrong time…
The phone app misses lots of data the car can collent and send. When, how frequently, and how long turn indicators are used. How aggressively the throttle and brake are applied, how frequently pre-tensioners and other safety features activate, how closely the radar or lidar-equipped cruise control sees you following others, and I'm sure lots I haven't thought of
Been asked that before—don’t always use one. Insurers will have to buy that data from the telecom if my phone’s on.
Not sure how common it is for Verizon, AT&T, Google (Maps), Apple (Maps) to sell to insurers (or brokers who insurers use). Perhaps pretty common? I’m sure they often can deanonymize it if they do.
Electronic injection sounds like too complex system to maintain, now if it had an actual carburetor that might be a peak "ICE", true and pure mechanical experience. Jokes aside, I hope that car at least have airbags, otherwise it's just death trap
MMR vaccines would be a net positive to the US economy, to the tune of $3.5B in direct savings (14x benefit/cost) and $7.6 billion (26x benefit/cost) in wider societal benefits [1]. Those are 2004 figures, so the dollar amount savings would be even bigger now.
To a point made further up the chain: The US's woes aren't driven by AI investment, but the self-interest of a select group. Compare it to a thief who causes $20,000 damage to steal a $100 stereo from a car.
Sure, over decades, but if you think the singularity will arrive sometime next week then all of this forward thinking nonsense is a distraction. And if you're heavily invested in some story related to that singularity, then you have a financial incentive to sabotage the forward thinking nonsense.
If it were my scheme, an incompetent president would be my first purchase.
2) If one applies a copyright message to AI generated output, is that fraudulent?
reply