gnramires

I think it's important to understand what bias is. Bias is not having an opinion. Bias is when several particulars to you influence your opinions in ways that don't reflect certain values, like the common good, truth, accurate representations of the matter, or similar values. I also think conveying certainty when there is none is a form of bias (when advocating for change, for example), as is equally advocating for absolute ignorance where there are some principles we can have confidence is.

That's different from neutrality, which for me is avoiding expressing opinions at all. I think neutrality is valuable if otherwise dialogue would be misinterpreted or cause conflict. Conversations need a well moderated (safe, welcoming, with vigilance for good-faith discussion) and freely exchanging environment to be productive, otherwise I think neutrality is the best option. Neutrality is also a good option when there are different cultural notions/opinions that can coexist without coming into conflict (this is a case where promoting uniformity and uniform agreement can be harmful, I believe! -- like, we don't need to agree that basketball is superior to football or something like that).

That would work only for a planar distribution of material. A 3D distribution would require multiple layers (I guess it might quickly become infeasible if it requires thousands of layers).

In the case of 3D arrangements, I think some substrate materials (and also some properties of the particles) would be very difficult to get using photolithography (or some kind of micro 3D printing).

Shouldn't an ECC non-correctable error trigger an immediate shutdown, because bad data could be committed to disk? (I guess unsafe shutdown could cause corruption elsewhere, but that seems like a reasonable risk) If attacks are a serious threat, then it would seem any alert that doesn't trigger immediate action would be risky (i.e. the attacker just erases alerts from logs).

> People have been claiming short term incentives for decades. If short term incentives were having long term negative impacts the corporate feedback mechanisms would have picked up on it by now.

Maybe the mechanisms are working fine, the markets keep going up, as well as the long term profitability. But I think the incentives themselves are the wrong incentives: as a society, profits are just a (very useful) tool to help us trade and regulate resources. They're not an end goal. The end goal is the wellbeing of humans.

The problem is when we forget that goal, we tend to believe predatory practices are fine, when they aren't. It may well be profitable, even long term, to make predatory (hyperpalatable, heavily processed) foods that make us unhealthy. The reasoning may well be 'Let's make enough money until consumers figure out we're being dishonest'. Getting the hands off the wheel doesn't seem like a sensible choice.

I think the solutions are many. People need to realize those problems exists, not just be in denial -- people are the agents of society, and even if you think pure markets are ideal, you need individuals to make good choices for themselves. In this sense you might say there's an untapped market of consumer and large-scale social protection that balances out our interests and capabilities with corporations. Individuals don't have decades of food research (that's in many cases influenced by industry funding, to make things more complicated) in mind when making choices, and they're influenced by marketing.

I do think ethics plays a large role here. It doesn't seem like a person, working in the food industry, should in good conscience be complicit in any scheme that harms the people eating your food. If the baker near your house started sneaking in addictive substances or subtly (and largely imperceptibly) altering ingredients to save costs, you should well claim that's a failure of ethics in the first place, not of markets. Part of the failing comes from the belief that profit is the be all end all mechanism of society. Being a good person to yourself and others is more important, fundamentally important.

If you work in industry and start seeing those practices being used as a desperate attempt to keep you alive versus the competition, that should give you pause: something in our industry, and our society is broken. You should fight, even from within your company, and spearhead, a change to align your ethical incentives with financial sustainability: by demanding more (or just _better_!) regulation, but also by changing the narrative around marketing, by investing in consumer awareness (and organizations specialized in raising awareness and food research), and also investing in genuine unbiased research. You need to basically fight evil and not let the bad guys have their way. If you're complacent, the default mode is defeat, just like getting your hands off the wheel of a car the default mode is crashing.

I should have been more balanced or nuanced a bit: I also don't think static linking is to be forbidden or completely shunned. As Linus himself says, a combination of both may be ideal. For basic system libraries like GUI libraries the current approach works well. But you should be free to static link if you want, and if there are serious issues if you don't. Maybe dynamic linking should be focused on a smaller number of well curated libraries and the rest should be left to static. Library archeology seems like a potential serious problem years from now.

I still think better listing dependencies (perhaps with the option to pin an exact version?) would be helpful, as well as better usage of something like semver. Someone mentioned binaries include paths to dependencies, but as far as I know, there is no tool to automatically try to resolve those dependencies or standard interface, maybe some more tooling in this area would help.

Another nice point about how it current works is that I think it relieves work from programmers. The policy of "Don't worry about distribution (just tell us it exists)" from distros seems like one less headache for the creator (and you can provide static linked binaries too if you want).

As most things in life, the ideal is somewhere in the middle...

As far as I can see, it would be unwise to roll back 30 years of (Linux) systems building with dynamic linking in favor of static linking. It mostly works very well and does save some memory, disk, and has nice security properties. Both have significant pros and cons.

I've been thinking (not a Linux expert by any means) the ideal solution would be to have better dependency management: I think a solution could be if say binaries themselves carried dependency information. That way you get the benefits of dynamic and static linking by just distributing binaries with embedded library requirements. Also, I think there should be a change of culture in library development to clearly mark compatibility breaks (I think something like semantic versioning works like that?).

That way, your software could support any newer version up to a compatibility break -- which should be extremely rare. And if you must break compatibility there should be an effort to keep old versions available, secure and bug free (or at least the old versions should be flagged as insecure in some widely accessible database).

Moreover, executing old/historical software should become significantly easier if library information was kept in the executable itself (you'd just have to find the old libraries, which could be kept available in repositories).

I think something like that could finally enable portable Linux software? (Flatpak and AppImage notwithstanding)

I'm interested in using Blu rays (maybe M discs, which are far more expensive!) for personal backup, as I generally don't have more than 100gb of really critical data (which excludes everything I could just download on the internet), and they're said to last a long time. Blu ray readers are widely available too.

So... can they give it a bump? :)

Chips, and intelligence in general, is a very special case, because it can be miniaturized retaining the same function (or if you prefer, gain function in proportion for the same area).

Other technologies might have some improvements from miniaturization, but I don't think the principle applies for solar cells. You obviously can't just shrink a cell and keep the same output[1]. What you can do is learn how to shave costs and improve efficiency, and also amortize R&D. Eventually you reach somewhere with very good efficiency and costs close to the raw material cost.

[1] That said, maybe you can shrink the thickness of a cell? That being the case, at one point just the cost of the frame for keeping the cell integrity (specially under wind loads etc.) would start dominating costs, and also the (copper) wiring outputting power from cells. I wonder how far we are from this being the case?

Wind power is interesting because it's mostly a mechanical problem (or at least there's a significant mechanical component). You have to transport the forces on a moving blade, or some kind of moving part (it has to be moving due to P=F.v), so all those parts have a minimum thickness necessary to offload those forces to the ground. It would increase mechanical efficiency (by requiring less load for a given power) if you could increase the velocity of the moving parts, but wind speed itself is limited. You could try funneling the wind, but that introduces structures which themselves suffer stresses (and also induce drag lowering efficiency). There's a limit (at material/W) given by wind speed, and specific strength of the structural material.

This very long exponential improvement (miniaturization) march is a somewhat unique effect to information processing[2], because information can in principle be represented at any scale. That's not to say that all those technologies can't be significantly improved. Generally automation, mass production, along with (limited) efficiency improvements and general cleverness can go a very long way.

[2] I don't mean to be taken as law: there are probably also applications (outside information processing) that you can only do well with the benefit of certain scales. For example, the cells of our bodies are wonderful mechanisms whose magic really works because they can each play tiny mechanisms and deploy molecular-scale objects to act on such tiny scales, and the billions of them make machines that have no parallels in terms of large-scale fabrication (including self-repair, filtration of compounds, ...).

I think your argument may be a little too strong. Would you refrain from discussing, say, happiness, sadness, etc. in your daily life without a formal definition of what happiness is exactly at a structural level? And I say that as someone interested in formalizing ethics, and formalizing all of those notions. The point is, to even (ideally, formally) define them we need intuition and intuitive discussions about what their definitions should be, and we are still at (very) early stages of what could constitute a real formalization.[1]

More significantly, the utility of those concepts far precludes scientific definitions, those terms have been useful for thousands of years, and are essential to human society!

(Again, certainly not against definition, but be mindful of being too strict, too soon about it :) )

[1] The benefits of formalizing (human concepts) are subtle: like formalization in math, they help us build confidence that we're not building sand castle, dig deeper into theories of meaning, and use this to slowly improve lives of everyone. But like math, we've been doing it informally for very long, and a strict axiomatic formalization is still today relatively rare (although progressing with the aid of computer proof systems) for most fields. Still, the evolving standards for proof are likely essential for many of the deep theories we've achieved (probably including say modern logic and Godel's theories, algebraic geometry, functional analysis, etc.).

Personally, I think ads can be quite predatory, which is why I dislike them (if ads themselves had better ethics and curation, I might think otherwise -- I have less issues with sponsorships elected by creators). That most people choose not to pay is a problem I think, but I think even ideally not everyone would pay (and that's fine): they may not have the means (although I guess they could just pay a smaller amount instead), or have too many other subscriptions (that they find need more money). I personally donate to a lot of services knowing that many won't pay (and many aren't able to), and I'm happy to contribute (although I think it would be better for everyone if they considered donating more!)

People are saying about 70% of Spotify revenue goes to rights holders, whereas you're saying about 85% of Bandcamp's revenue goes to rights holders. It really doesn't seem like that much of a difference?

Maybe what you're saying is, you end up spending much more buying merchandise and labuns directly than you would spend on Spotify. (I'm not sure this would be true for everyone though)

Maybe then the solution could be to have a way to just pay more to Spotify (conditional on keeping the revenue split intact).

Something I don't like about Spotify though, is that I don't get to have any kind of say on how the revenue is split. I'd personally prefer if there was an egalitarian bias in payment, and the artists with less revenue would get a greater share of my subscription. But there's no way I can control that, that's the most frustrating to me personally, and I'd gladly switch to a system that pays more (since I currently have the means to).

In fact, I've proposed FunkWhale, the federated (libre-)music streaming platform, should get a subscription service like that, and that I should have some control over the revenue distribution (maybe there would be a minimum revenue split, and the rest I can 'choose my own algorithm', for example one that heavily favors less popular musicians). I agree that meanwhile the best I can do to support them would be paying them directly, and I've found a few have Liberapay (or Patreon) accounts as well.

The number of connected users is different from the number of user connections :)

The first is the number of nodes, the second edges.

I think part of this connection reflects an universal (not just coincidental) equivalence: between intelligence, (logical) reasoning and planning. Finding a solution to a logical problem is perhaps similar to finding a path in a motion planning problem. You have to consider a number of paths going from A (premises, or in general your current state), to B (a desired proposition, or in general "something that is good/satisfies you"). You have to search a path (within a heuristic list of possible paths) or plan a path of consistent propositions that in the end prove what you want (although our notions of proof and consistency are usually relaxed for practical purposes: when I say "The window was left open, and we are in the winter, therefore the room will be cold", this is not a real proof in the Euclidean sense.

Science is more than just testing, although testing is an important part of science. I believe the reasoning aspect can, in principle, be made strong enough to be robust, and almost independent of experimental confirmation (essentially from analyzing already existing data).

For example in Machine Learning you can trust a model to work well in the context of a data stream by simply using (cross) validation, and so on (in general avoiding overfiting, and assuming your data stream will continue your training data without changing too much).

Mathematics is an example of a non-experimental science as well: in a way, we can perform purely theoretical experiments where our 'tests' are given by internal consistency. The equivalent of experiments in mathematics for ruling out incorrect statements (inconsistent theories) are proofs and (almost tautologically) the concept of consistency. I think there is some risk that even within math this process derails, but it's only when we start lowering too much the standards for what constitutes a proof. And the concept of proving is so solid we can automate it and computer-verify them. As long as we can (and do) verify what we create, in a sense we can't (or are extremely unlikely to) go off rails.

I think relativity is a good example of a theory that was largely conceived from mathematical principles (from a few observed essential properties of reality). I think in the future this kind of approach will be highly significant to physics, math, philosophy and even social sciences.

It is true though that experiments (and just real life) are very useful to 'keep us in check' in a general, less rigid way as well.

Honestly I think those are the fantastic tool (consistency and proof), fundamentally connected to the notion of truth.

Correction: I've just realized I've drawn an incorrect conclusion, the energy per bit in the case I've shown is indeed lower than the standard 2 bit encoding, so it's actually a win! Again, for a myriad reasons, I'm not sure you'd want to use that encoding in real life, but it does consume a little under 20% less power in terms of V^2 losses.

That's true, if you use binary encoding to represent the b states. I were referring to base b representation, in which by definition we use b states to represent log2(b) bits: for example, 4 states to represent bits 00,01,10,11, which could correspond to voltage levels 0V,1V,2V,3V of a variable (analogous for other values of b).

In this case, the average voltage used would be the average of 0, 1, 2^2, 3^2, which is 0+1+...+(b-1)^2 / b ~ b^2 (proportional to b^2). I showed that in this case in theory the "optimal" base would be sqrt(e), but I can't think of simple representations of a variable with less than 2 states (you could have say 2 variables representing less than 4 states, for example the 3 pairs { 0V/1V, 1V/0V, 0V/0V }, but that seems a tad complicated![1]), so the least non-trivial integer base is just 2.

I think in reality this kind of argument doesn't apply to real computers in a straightforward way, because there are many other factors at play than what I would call a 'dynamic power consumption' associated with the square voltage. There are things like leakage currents (that consume some fixed power per transistor) and other effects that significantly complicates things, so that such simple claims of being "optimal" don't apply (so I agree with the other comment about avoiding absolute claims!). But they can inspire designs and we can see if there are any advantages in more realistic cases.

[1] Further analysis: In our example, the 3 voltage pairs have mean square voltage 1/3. If we instead use the pairs { 0V/0V, 0V/1V, 1V/0V, 1V/1V }, the mean squared voltage is 1/2. In the first case, we encode log2(3)/2 bits of information per voltage value, in the second, 1 bit per value. So the energy per bit is (energy per value)/(bits per value) = (1/3)/(log2(3)/2) = 0.420... for the first case, and 0.5 for the second case. Unfortunately, it's a loss!

As an exercise, I believe that if we delete the state 1V/1V/.../1V of a k-voltage tuple, for large enough k-tuple, we come out ahead (probably not by much), although again that's not necessarily useful in practice.

I think the b factor is due to the analog representation: if the base has b values, you use on average b/2 voltage for representation. I would think the mean square value in might have been more appropriate though? (as voltage energy losses usually are V^2 / R or C V^2)

Repeating the calculation (for k bits and b values) using square values, you get E = b^2 * log_b(k) and dE/db = 2b/ln(b) - b^2 / (ln(b))^2 / b = b/ln(b) * (2 - 1/ln(b)). Setting dE/db to 0, we get 2 - 1/ln(b) = 0, ln(b) = 1/2,

b_opt = sqrt(e) = 1.6487...

It's tricky because as noise increases, large constellations, although perhaps in theory with advanced statistical decoding (where you consider P(Symbol|Received signal)) they're advantageous, become less viable and more demanding computationally. Usually the gains are a small factor. At one point, the energy cost of computation required to make statistical decoding[1] surpasses any gains from increases in complexity.

More than that, the ratio of channel energy (or cost) requirement to computational energy can vary :) For example, if you're transmitting a message to a space probe, the energy and costs associating with message transmission are enormous, such that the decoding energy will be a lower proportion (favoring more complex decoding systems). For a computer bus transmitting signals internally at short distances, the energy cost of encoding (and even using QAM or non-binary encoding at all) and statistical decoding will probably be less than the energy gains.

Credit to Christopher Blake et. al in this paper: (as seen on Canadian Workshop on Information Theory :) )

https://tspace.library.utoronto.ca/bitstream/1807/69482/1/IT...

From the abstract: "This implies that the average energy per decoded bit must approach infinity for any sequence of decoders that approaches capacity" (valid for the VLSI model in question, and binary channel model, but I'd guess this generalizes to any physical computational medium)[2]

It's really interesting how in the context more complex isn't always better.

[1] Usually the complexity of the whole system, in more practical terms, is determinant as well -- because engineering and maintaining complex systems is difficult

[2] More precisely: "It is shown that for any sequence of increasing-block-length decoder circuits implemented according to this model, if the probability of block error is asymptotically less than 1/2 then the energy of the computation scales at least as Ω (n√log n), and so the energy of decoding per bit must scale at least as Ω (√log n)"

Indeed! I wonder why the system can't just stop giving the process memory when it's starting to become critical? (say, you're over 95% memory used)

I really think a more general solution is treating memory (and CPU time as welL!) as a scarce resource, and programs should either: (1) Deal with having severely denied resources (default behavior); (2) Use a communication protocol to negotiate memory with the system.

Negotiation could mean (a) The process freeing memory spontaneously when there's pressure (another process requires it) and it's inactive, (b) Requesting the system free memory from other processes when user requires.

This would impose more memory management burden on developers (trying to fulfill OS requests), but in turn it would make for a far better experience when there's memory contention.

Disclaimer: I have no idea how memory allocation works in detail. Maybe this is reinventing the wheel?

Indeed, and I think we should be thinking about pricing the recycling or disposal costs too (even if that's difficult or inherently imprecise), as well as giving incentives for products with effective recycling supply chains (or simply low disposal rates! i.e. longlasting products). I think it's worth not being too heavy-handed about this, because low income people would be hit the hardest most likely, but I think something like this would help significantly.

I assume they want to keep their printers with 3d printed parts for the following reasons:

(1) Dog fooding their own printers (in mass production) so they're reliable;

(2) Making DIY, modifications and repairs easy on their printers;

(3) Ease of prototyping and making new products.

It seems that given a large production scale, (3) would not be so important. (1) seems very important, but maybe they could reduce the number of printed parts and still have "enough dogfooding"? As for (2), perhaps they could compromise on having injection molded parts that can be replaced (i.e. are interchangeable with) printed parts.

Sure, but it seems this is unphysical if this passage occurs after essentially an infinite time for an outside observer. Before crossing the event horizon, the infalling person sees the whole history of the universe. But the problem is, if crossing the horizon takes infinitely long, how could the horizon form in the first place? That's what I've never seen addressed to a satisfactory degree.

If their light reaches you, your light could reach them, this is a reciprocity condition that I believe is valid in GR too (because effectively light follows null geodesics, and it can always go both ways, geodesics don't have a preferred direction).

It's true that infalling observers will experience outside time speeding up, and conversely outside observers will see infalling objects slow down. In fact, as they approach the Event Horizon, they would appear to experience diverging time dilation. So in effect, they become "frozen in time". But remember that you can always maintain two-way communication with light. The very fact that both the outside observer can always see the infalling person and the infalling person can always see the outside world imply there can be two-way communication all the way up to crossing the event horizon (which would take infinitely long for the outside observer). This ought to be enough to conclude the infalling person can in fact never reach the event horizon.

Consider this argument: if you can see the person falling in for "infinitely long", then you can actually communicate with this person falling in. For example, through light or gestures. Since you can communicate, this means the image you see of the person falling in is not just an illusion or artifact of general relativity, you both must agree on reality, that the person still hasn't crossed the horizon.

The schwartzchild solution comes with a bunch of caveats!

For example, it is eternal, it's not the result of a collapse solution. As far as I know, no one has an analytic solution yet for black hole collapse of say a star in finite time.

For an outside observer, in fact I believe an infalling object into a black hole is never observed to reach the event horizon at all. Since the infalling object could communicate (using light) with the outside observer, there's a sense in which they must agree on whether the object has or hasn't crossed the event horizon -- in this case, the non-observation of crossing the event horizon must not be just apparent, because of this communication.

Since event horizons couldn't form in this case, then it should follow that a black hole cannot gain mass at all. I've yet to see a satisfactory debunk of those claims, at least at a level I could understand.

I live in Sao Paulo, Brazil. We have quite bad drug abuse issues as well.

I have helped a group that rescued several people (I believe most with drug addiction) from the streets[1]. What they do, and several other groups, and also government institutes do, is offer voluntary recovery programs that are legitimately very good. You can go there and they'll take care of you, including professional psychologist/psychiatrist support (and medicine to help get off drugs); in the main institution we're working with[2], former homeless and former drug dependents help directly in welcoming people. I'm not sure that's the whole solution to the drug problem, but I think it's part of the solution -- getting out in the streets, offering help to the homeless and addicts, and if they want, taking them somewhere that'll offer shelter and recovery from addiction. [3]

There's one rule in those places: you can't go in with drugs with you. Of course, if the person becomes too violent then they might not be able to deal with them either.

Please, go out there and help (while staying safe too, preferably in a group) -- or help institutions that do this kind of work (like we say, everyone can play a role, be it in a kitchen, on the streets, or with donations). I think it's groups like this that keep things from falling apart.

There are many people on the streets for whom drugs are just a quick escape mechanism, and have economic issues, mental health issues, or health issues (or all together), and a compassionate community/institutions that helps everyone recover is what really is needed I think.

Pretending they don't exist or as a friend calls "shuffling them around" doesn't work.

[1] Our group offers a meal once a week (Sunday nights, where they have difficulty finding something to eat), and we offer taking them off the streets. It's incredible what genuinely caring about people will do: I've seen people cry when you hand them a plate of food.

[2] It's this one: https://www.missaobelem.org/; This one is state funded and excellent as well: https://hub.spdmafiliadas.org.br/ (called HUB)

[3] Other important parts being, educating people about the tragedy of drug abuse, and providing community and support so they don't resort to drugs, and also preventing selling.

It's not that simple. LLMs are great for predicting next word, but this needs a more unsupervised approach.

Ultimately, you need to know the environment of whales. Sure, you know the next probable token, and even relationships between tokens, but how do they relate to what whales are doing?, in other words what does it mean?.

It reminds me of the great quote by Feynman: "look at the bird"[1]. You can know the name of a bird in many different languages, but that tells you nothing about it. If you want to understand birds, you need to look at it and see what it's doing.

If we want to communicate with another species, we need to understand what they're doing, why they're communicating, and from there we can begin to infer what they're communicating. It could be simply a callsign/identification/name for each individual, it could be communicating you've found a place with food, it could be communicating about predators, who knows. There could be all sorts of interesting variation and nuance to their calls. But ultimately it's necessary to look (in the general sense of having information about what it's doing -- could be through all kinds of sensors!) at the ~~bird~~ whale :)

[1] Feynman tells the as a lesson from his father. (that you can know their names and still know nothing about the birds themselves).

https://www.youtube.com/watch?v=ga_7j72CVlc

Bonus related video!: https://www.youtube.com/watch?v=M1TiXLGqlM4

I forgot to add: it's unlikely anything you come up with (while trying to develop your creativity) is going to be legitimately useful or even fun. It's like expecting to play a violin like Hillary Hahn (or <insert skilled musician>) after you've just picked it up. But the time to play an instrument so that at least you can tolerate your own sounds and have fun is much shorter -- and frankly, necessary I think not to give up before you can play really well (if that matters at all).

Creativity can be trained just like any other skills! :)

I second taking note of stuff; like most things, genuinely practicing means you will get better.

Something that's also overlooked often is (1) Technical knowledge, (2) Methodical invention (3) Motivation!.

Technical knowledge helps you know which projects are possible (or just economically feasible), and map the roadblocks on the way; impossible inventions are not really useful.

Methodical invention means methodically looking at things instead of just randomly inventing. So you analyze a problem, like: (a) I want to make a volumetric display, (b) I want to bring (virtual) 3D objects to life. Notice the subtle difference between the two. Volumetric displays (a) can range from a lightfield display, which is a very elegant theoretical foundation (and can be realized in a number of ways), to volumetric persistence-of-vision displays like the own shown here. It's likely you will eventually arrive at whatever solutions are possible when you deeply research about a problem domain. Bringing a 3D object to life (b) can mean anything from digital fabrication, VR glasses, haptic interaction, robots, and more. It's a much wider problem domain.

Motivation is also really important. Understanding why you're inventing, and what you what you want to bring to life is also significant. Are you going to bring people joy, is it just very cool, is it a useful medical device that can save lives, etc.. Focusing on worthy subjects is also something you can study and that will really increase how good you are at inventing. Of course making stuff just because [we like it] is also important and valid :)

And for me the most important part is to have fun while you're at it...