You're being somewhat unfairly voted down, it's a legitimate questions because the popular media so grossly misrepresents what string theory is, especially in their visuals.
It's hard to visualise in 3D, but if you cut down the spatial dimensions to just 1D (a line), then theories like string theory just turn the infinitely thin mathematical line into a tube. You can picture a tube that vibrates, or has waves in its cross-section. Don't think of the the "strings" as actual little loops moving around in space, they're a modification of what space is.
You can even do the same kind of line->tube extension of a space with even more extra "loop" dimensions than the number of base dimensions. AFAIK the current theories have 10 total, of which 3 are the usual "large" dimensions of space, the rest are "small" and rolled up like the tube example.
Even so I don't get how electrons a meter apart interact through that stuff, as opposed to the electrical fields which spread out through space and so interact with other electrons as featured in quantum field theory which is what physicists use to actually calculate physical results, as opposed to string theory which fails to calculate actual physical results.
"if you cut down the spatial dimensions to just 1D" doesn't sound very physical to me.
I'm maybe being downvoted fairly. I studied physics and it don't think it's a misunderstanding of popularisation or that string theory is untestable, I just think it's straight wrong and not how the physical universe works.
> Even so I don't get how electrons a meter apart interact through that stuff
Very roughly: It's possible for point-like (or tiny looped) particles to interact as long as they take every possible path instead of just the one path that would cause a collision. How you interpret this is... up for debate. I prefer the many-world interpretation (MWI), but not everyone agrees.
> "if you cut down the spatial dimensions to just 1D" doesn't sound very physical to me
That's just a simplification to aid understanding, it's now how the theory actually works.
Yeah maybe. My flatmate of some years was doing a PhD in string theory at UT Austin along the lines of "if you cut down the spatial dimensions to just 1D" but he was a mathematician, not really a physicist and was ok with that if it produced interesting mathematics. For real physical things like wiring the lighting system I'd do it because he wasn't so good with that.
I think he went into string research because he was good at maths and there was grant money available for that rather than a deep belief that that was the nature of reality, which is kind of what I mean by sociological factors.
I think much string theory may be like that. Interesting maths but not good at figuring where electrons go.
Except this is how electrons actually go, and it has real testable consequences. The question I'm aware of (because it related to my degree in nanotechnology) was: are metals conductive at different dimensionalities?
Because at the nanoscale, you in fact can have 1D, 2D and 3D metals. 3D metals are bulk solids - like we're familiar with. 2D metals are planes of single (or very few) atoms. 1D metals are lines - think placing individual metal atoms down in a row - nanotubes are a practical example.
All real, possible structures to build.
When you do measurements on all these structures you get...weird answers. Like is a nanotube a superconductor? And the answer is...yes, but also no. Yes because you'll in fact view superconductivity like behavior, but no because actually it's a ballistic electron conductor - at the right energy level an electron bounces through the thing without hitting it, but not all electrons can do that at all energy levels, so you still measure a voltage across a nanotube between two conductors.
But a nanotube is 1D - we only have 1 dimension things move in (from one end to the other). So - conductive, not a superconductor, but you can kind of use it like one sometimes. And we know 3D metals are conductors - that's obvious right...so what are 2D metals? Presumably conductors right...?
And the answer is...nope, insulators - at least sometimes. And the reason is because the sum of all possible electron paths in a 2D metal is the electron always returns back to where it started - and those grow much faster mathematically then paths where the electron ends up somewhere else.
But only in 2D: in the 1D case most paths take you out of the conductor. And in the 3D case, the number of paths which land you somewhere else grows much faster then those which loop back, due to the extra dimension of freedom. But 2D metals are constrained - for any given path elsewhere, there are mathematically far more that land back where you started. This is observable, measurable behavior which is a topic of research for future semiconductors. Yet it's almost entirely quantum probability based behavior.
The string theory model is structured from the start to reproduce existing theories as a consequence, so the problem here is a lack of understanding on your part rather than a mistake in the theory. I do suspect that most people who are working on this stuff (or any stuff for that matter) don't think very hard about the basic phenomenological claims to the point where they can explain them well.
In particular your model of electric fields isn't very good. An electric field's flux around a volume reveals the presence of a particle in that volume. That's not because the field, a bunch of vectors in space, "happens" to integrate to something nonzero if there's a particle in that volume: it's because in some sense the presence of a field with a nonzero flux and the presence of a charged particle are the same thing; the particle is the existence of a divergence in the field within that volume.
Moreover in QFT (and this part is handwavey as I only learned enough to vaguely understand this, but it's better than nothing) the presence of the "field" ends up looking like the sum of what you get if you integrate over every possible way of a emitting or receiving a photon at that point; the accumulated integrals destructively interfere in such a way as to produce a value which reflects where the particles are. So very roughly idea of a field existing at a point and having a "value" is like saying: there are a bunch of things out there that I (a charged particle) can detect by exchanging photons, and the accumulated effect, when you consider all the different quantum superpositions of ways of doing that, is a single vector which induces a force on me. Other fields add up to more complicated objects than vectors.
Once you look at things like that, there should be no objection to how strings and electrons might interact. Whatever's going on at that string level averages out over larger timespans to just look like electron field. Not dissimilar from how all the individual charges in an atom average out to look like a single charge (but are perhaps detectable if you get up really close, in dipole and higher moments, or in how the atom deforms / reacts to nearby charges).
It might help to be aware of the concept of a topological defect (https://en.wikipedia.org/wiki/Topological_defect; there's a great explainer article somewhere that I can't seem to find) as a reductive picture of what a particle "is". I've heard this doesn't work for fermions for some technical reason, but in any case it's very useful as an illustration of the sort of thing that a particle "can be": a vortex in a material can act like a particle and even exhibit attractive/repulsive forces. So I picture the string theory model as answering the question: what kind of substrate could produce vortexes and other "defects" that act like the particles we see? Dunno if that's accurate but it seems like a natural question to me, anyway.
I can definitely buy a lack of understanding on my part - my understanding of all this is quite fuzzy. But I'm not sure string theories reproduce existing theories so much a say this is compatible with the other stuff in a handwavy way? I don't think there's any "we assume strings and thus can deduce quantum mechanics and general relativity" stuff? I don't think anyone can deduce quantum mechanics and Einstein's deductions with relativity are based on simple observations like the speed of light being observed to be constant and acceleration seeming similar to gravity?
It's not "we assume strings and deduce QM" it's "strings are a model which is defined so as to produce QM and also...". It is, by design, a theory of QM and GR. That's the point. It's a claim about what lies "below" the level of the individual fields in QFT which gives rise to them (and also gravity).
If I make certain assumptions and you respect them, I will give you certain guarantees. If you don't respect them, I won't guarantee anything. I won't guarantee that your data will be toast either.
If you can't guarantee anything for all the nodes losing power at the same time, that's really bad.
If it's just the write buffer at risk, that's fine. But the chance of overlapping power loss across multiple sites isn't low enough to risk all the existing data.
I disagree that it's bad, it's a choice. You can't protect against everything. The team made calculations and decided that the cost to protect against this very low probability is not worth it. If all the nodes lose power you may have a bigger problem than that
It's downright stupid if you build a system that loses all existing data when all nodes go down uncleanly, not even simultaneously but just overlapping. What if you just happen to input a shutdown command the wrong way?
I really hope they meant to just say the write buffer gets lost.
That's why you need to go to other regions, not remain in the same area. Putting all your eggs in one basket (single area) _is_ stupid. Having a single shutdown command for the whole cluster _is_ stupid. Still accepting writes when the system is in a degraded state _is_ stupid. Don't make it sound worse than it actually is just to prove your point.
> Still accepting writes when the system is in a degraded state _is_ stupid.
Again, I'm not concerned for new writes, I'm concerned for all existing data from the previous months and years.
And getting in this situation only takes one out of a wide outage or a bad push that takes down the cluster. Even if that's stupid, it's a common enough stupid that you should never risk your data on the certainty you won't make that mistake.
You can't protect against everything, but you should definitely protect against unclean shutdown.
If it's a common enough occurrence to have _all_ your nodes down at the same time maybe you should reevaluate your deployment choices. The whole point of multi-nodes clustering is that _some_ of the nodes will always be up and running otherwise what you're doing is useless.
Also, garage gives you the possibility to automatically snapshot the metadata, advices on how to do the snapshotting at the filesystem level and to restore that.
All nodes going down doesn't have to be common to make that much data loss a terrible design. It just has to be reasonably possible. And it is. Thinking your nodes will never go down together is hubris. Admitting the risk is being realistic, not something that makes the system useless.
How do filesystem level snapshots work if nodes might get corrupted by power loss? Booting from a snapshot looks exactly the same to a node as booting from a power loss event. Are you implying that it does always recover from power loss and you're defending a flaw it doesn't even have?
It sounds like that's a possibility, but why on earth would you take the time to setup a 3 node cluster of object storage for reliability and ignore one of the key tenants of what makes it reliable?
He is protected by the most corrupt supreme court in history.
The New York Times has reported extensively on Justice Clarence Thomas’s acceptance of luxury travel, real estate transactions, and other gifts from Texas billionaire and GOP donor Harlan Crow.
A whistleblower complaint filed by Kendal B. Price, a former colleague of Jane Roberts at Major, Lindsey & Africa, revealed that she earned approximately $10.3 million in commissions between 2007 and 2014 for placing lawyers at top firms. [1]
The Times reported that Justice Samuel Alito has faced ethics questions related to his relationship with conservative donors and political allies. Although his 2025 disclosures claimed that he received no gifts during the previous year, prior reporting noted that he had accepted luxury travel and accommodations paid for by wealthy individuals with interests before the Court. He has defended such trips as falling within disclosure exemptions.
Justice Neil Gorsuch was the subject of a New York Times story in late 2024 that addressed his connections to billionaire Philip Anschutz, who helped steer him toward his earlier legal and judicial appointments.
The light source is the “easy” bit. The mirrors, masks, and the rest of the machine are all individually as difficult if not more so.
The wafers have to be positioned to nanometer accuracy repeatedly and at high speed! It’s hard to believe that’s even possible, let alone commercially viable.
Managing the light source, specifically the 13.5nm length on the wave spectrum, that gets generated from overheated tin plasma, is in fact the most challenging part of the machine. Here "managing" includes the process of hitting a rightly sized tin droplet with lasers at the right angles, and all the rest of the complicated fluid math necessary to get the most of that precious lighting moment, as well as the proper handling of that spark event's after-effects, of course. As opposed to the rest of the machine parts (like directing the EUV light to the reticle through those mirrors you mention), the light generation part is dynamic, very easily to get wrong, and very costly to iterate on.
Something I've started doing in my workflow is using Pandoc to convert between Markdown and DOCX when authoring long documents. This lets me put the Markdown into Git and apply the Gemini CLI to it. When referencing other documents, I'll also convert them to MD and drop them into a folder so I can tell the AI to read them and cross-reference things.
At the start of the project the Markdown is authoritative, and the DOCX is just for previewing the styling. (Pandoc can insert the text into a layout template with place holders.)
Towards the end of a project I'll start treating the DOCX as authoritative but continue generating Markdown from it, so I can run the AI over it as a final proof-read or whatever.
This is similar to what people used to do with DocBook, but with a more friendly text format and a more AI-friendly "modern" workflow with Git, etc...
A few gigabytes of text is practically free to transfer even over the most exorbitant egress fee networks, but would cost “get finance approval” amounts of money to process even through a cheaper model.
It sounds like you already know what sales peoples incentives are. They don't care about the tiny players who wanna use tiny slices. I was referring to people who are trying to push PB through these. GCPs policies make a lot of sense if they are trying to get major players to switch their compute/data host to reduce overall costs.
You're off by orders of magnitude. A million tokens is about 5 MB of text and costs $0.20 to process in something like Gemini 3 Flash.
Hence, a terabyte of text would cost about $42,000 to run through a pareto-frontier "cheap" model.
The most expensive cloud egress fee I could quickly find is $185 per terabyte (Azure South America Internet egress).
Hence, AI processing is 200x as expensive as bandwidth, or put another way, even a long-distance international egress at "exorbitant cloud retail pricing" is a mere 0.5% of the cost.
Petabytes, exabytes, etc... just adds digits to both the token cost and bandwidth cost in sync and won't significantly shift the ratio. If anything, bandwidth costs will go down and AI costs go up because: output tokens, smarter models, retries, multiple questions for the same data, etc...
Also, you'd be shocked at how few developers know anything at all about sRGB (or any other gamut/encoding), other than perhaps the name. Even people working in graphics, writing 3D game engines, working on colorist or graphics artist tools and libraries.
If you reject the best and only easy option from the outset because you don’t want actual healthcare, then yeah… whatever remains is going to be “hard”.
What the US has right now is a complex entrenched system of financial middlemen that refuse to abandon their rent seeking. They provide only(!) financial “services” and will fight actual healthcare tooth and nail.
Trump wasn’t strong enough — or simply didn’t care enough — to fight these people.
It's hard to visualise in 3D, but if you cut down the spatial dimensions to just 1D (a line), then theories like string theory just turn the infinitely thin mathematical line into a tube. You can picture a tube that vibrates, or has waves in its cross-section. Don't think of the the "strings" as actual little loops moving around in space, they're a modification of what space is.
You can even do the same kind of line->tube extension of a space with even more extra "loop" dimensions than the number of base dimensions. AFAIK the current theories have 10 total, of which 3 are the usual "large" dimensions of space, the rest are "small" and rolled up like the tube example.
reply