HNHacker News
TopNewBestAskShowJobs

blamestross

698 karma · joined October 16, 2016

Mad Computer Scientist https://miasma.space

distributed hash tables, p2p systems, supply chain and network security.

submissionscomments
blamestross··on A theory that Mars lost its magnetic field and then its oceans (2022)
Progress is exponential until "surprise it is a sigmoid!" I'd argue we are there already, but even if we aren't planting seeds is rarely a waste in the long term.
blamestross··on A Melting Cauldron: The Book of Hallowe'en (1919)
As a heathen, this is bordering on offensive.

The difference is intent. This isn't paganism hiding in Christianity. This is intentional organized decisions to convert pagan holidays to Christian ones. It is a radically successful attempt to eradicate them not "the old hiding in the new".

Christmas stands out even more as an example of it, literally the time of the celebration was chosen by committee to displace winter solstice holidays.

Plenty of religions have solstice holidays and and harvest celebrations, and likely other older religions did the same, but Christianity is the only one in living memory that actively subverts and attempts to consume the holidays of native populations they wish to convert.

blamestross··on Halfsies
I wish there was a leaderboard or percentile you could see without making an account.
blamestross··on What McKinsey got wrong about developer productivity
And it is worth pointing out there is _some_ value to that. It gives a chance for implicit assumptions about how things work to be exposed, and allows the conscious choice on if to keep them.

Now whether McKinsey is actually delivering on that idea, or just exposing the implicit assumptions of "don't be an asshole and don't do things that are harmful to everybody else" and encouraging dropping them, is a problem.

blamestross··on Americans Are Still Spending Like There’s No Tomorrow
> I thought that way when I was young too...then I learned how funds-matching and compound interest works.

I know very well how they work. It still isn't worth the penalties. There are better investments that can get more effectively leveraged sooner, while there are still things to spend it on. All evidence is that the next 30+ years are kinda going to suck, buy personal infrastructure, invest in your health, enjoy access to global travel while they are things you can spend money on.

You won't even have the cash to burn to keep you warm if you leave it in a 401k forever.

blamestross··on Americans Are Still Spending Like There’s No Tomorrow
There isn't any real value to putting money into a 401k if you don't think the infrastructure to support it will be there when you get to your 60s.

I'm moving forward having elective and prophylactic medical procedures in the next few years instead of the next few decades so they can happen while I have access to such high quality healthcare.

Lasik and lense replacement seem like an example of what to prioritize bc glasses and contacts have such a long supply chain.

It isn't quite "prepper" bad, but my generation isn't planning on the time of plenty lasting.

blamestross··on Costco gold bars are selling out within hours
Gold got it's reputation before we had efficient aluminum production methods. For a while, aluminum was the most useful and valuable metal on earth, it caps the Washington monument for that reason. Nothing has really changed other than gold is more scarce. In a "prepper" situation, aluminum ingots are fundamentally more useful for every use but trade, and it's trade value is only because it has trade value. It isn't like they will be making new electronics after the fall.

Over 100 years of the value being propped up by a social narrative alone. Boggles the mind. Reminds me of no-vote no-dividend stocks.

blamestross··on LaTeX for tabletop
If you are after a D&D feeling document check out (Homebrewery)[https://homebrewery.naturalcrit.com/]

Overleaf (which a recommended as a latex tool) has a (5e style template)[https://www.overleaf.com/latex/templates/d-and-d-5e-latex-te...] too.

blamestross··on Building an economy simulator from scratch
Beware optimizing for "average" quality of life, we did that and it worked. Turns out it is more efficient to have a super-minority with a very high QoL. Median is a better target.
blamestross··on US rejects AI copyright for famous state fair-winning Midjourney art
I agree with most of what you are saying, but the key gap is that the victims here are not "unethical companies". They are 5+ digits of independent artists who had no idea this was going to happen to them.

If big art was a thing, your argument would be good, but Disney doesn't account for most of the input to these image algorithms.

blamestross··on US rejects AI copyright for famous state fair-winning Midjourney art
It is almost like intention in the method and process of those tools matters. Rather than bulk exploitation of other people's work.
blamestross··on ASMesh: Anonymous, Secure Messaging in Mesh Networks
There are a LOT of round trips on this protocol and I wonder what they mean by "mesh networks" such that the bandwidth and reliability that all those round trips are sustainable.
blamestross··on Square is down
Yeah, all that stopped being directly supported and stopped being mandated for use. There was a strong cultural rebellion against "senior engineers dictating policy"
blamestross··on Square is down
Management that refuses to fund or mandate any common infrastructure.

Jack likes running his companies like "an economy of startups" but in practice this means nobody is allowed to do things without fast returns. Teams are encouraged to use whatever programming language and infrastructure they think fits them. It is the fast track of tragedy of the commons problems.

blamestross··on Can we talk to whales?
I love sign language for it's "pronoun" system and spatial features, but the reality is we could just a easily "annotate" spoken english with similar spatial context, sign language just does a good job of teaching all parties how to do that be default. By not teaching ASL to everybody who speaks english, we are really missing out on more bandwidth (for detail and redundancy) in communication.
blamestross··on Live-caption glasses let deaf people read conversations [video]
Because learning written English isn't easy when the sounds it represents are not part of your sensory experience. It's more like learning Egyptian hieroglyphics. Plus many parents either can't or won't invest the effort/money to make it happen. Lots of parents never even learn ASL to communicate with their own kids. Met a mother in a beginner ASL class finally deciding to learn enough to communicate with her 20 year old daughter (Who was born deaf and illiterate FTR). It broke my heart when I met the daughter and found out she basically had no exposure to other signers too. Deaf all her life and she didn't have a name-sign.

There's "cognitive impairment" and then there's "nobody bothered to communicate enough for a kid to build communication skills."

blamestross··on Live-caption glasses let deaf people read conversations [video]
"most" is an interesting choice of word here.

It's true. 70% of deaf children learn to read English [1]. But I don't think that's enough to consider it a safe assumption.

[1] https://www.handsandvoices.org/articles/education/advocacy/w...

blamestross··on Security robots are at Philly Lowe’s stores. Some have nicknamed them snitchBOTs
It seems more reasonable to think "The social contract is erroding and so more shoplifting is happening"

It's not like people consider antisocial behavior by default unless they are desperate and driven to by a society that's failing them and failing to give them hope that they have other options to survive.

blamestross··on Beyond CRISPR babies: How human genome editing is moving on after scandal
My point isn't that military isn't a threat to politicians, only that in general they don't seem to see it as one (see historical dictatorships fascination with said supersoldiers)
blamestross··on Beyond CRISPR babies: How human genome editing is moving on after scandal
> Wouldn't super scientists, engineers, businessmen, salesmen or politicians be more useful to a state than a super soldier?

Soldiers are by social role, servants of government. Super scientists, engineers, salesman, and politicians all present different threats and competition to government. Useful to a civilization yes, but government isn't a civilization, it's people who are clinging onto a bucking-bronco for power and safety.

blamestross··on To build truly intelligent machines, teach them cause and effect
It's almost as if the only two modes of self consistent reasoning are either circular or predicated on assumptions taken in blind pragmatic faith.
blamestross··on Ask HN: Expectation that ChatBots have perfect personalities when humans don't?
I think the misunderstanding is this:

> Why do we expect all ChatBots to have a perfect knowledge set and personality?

We don't expect it. We require it of a commercial application of this technology.

It's like self driving cars, if we are going to hand such a task off to machines, we need it to be better than the humans doing the job, not just cheaper. (A brick on my gas petal is a self driving car, but not appropriate for sale as "AI")

And honestly, I don't think they will get much better without radical method change. They already get fed basically the entire corpus of human written tokens accessible in English. And soon it will be tainting its food supply with it's own spoor.

blamestross··on Convex Hulls – The Metapict Blog
So you won't be able to define a hull at all without (n+1) points. He has 10k points so that sounds like a lot relatively.

So the definition of a convex hull is, put generally, the set of points that define faces such that every point is on the face or on the "inside" of it (mean point side)

To test if a point is inside that simplex hull, you need to check every one of those faces. But that isn't the problem.

Every face is a "filter" that all the other points have to pass. Moving beyond the simplex, at higher dimensions, the number of faces that need to "accept" every other point scales faster in higher dimensions. The odds of that aren't horribly clear, you are right in other comments to call out that the structure of points in this context is by definition not independent or random, but you need enough structure to get around the fact that high dimensional hulls are basically all surface.

blamestross··on Convex Hulls – The Metapict Blog
So, about that. Do the math on how many faces a 768-simplex has.
blamestross··on Convex Hulls – The Metapict Blog
> Qhull does not support triangulation of non-convex surfaces, mesh generation of non-convex objects, medium-sized inputs in 9-D and higher, alpha shapes, weighted Voronoi diagrams, Voronoi volumes, or constrained Delaunay triangulations,

You are going to have to roll your own.

One trick you can use is that most convex hull algorithms chase O(nlg(n)). That lg(n) implies a branching step which lowers efficiency on GPUs. Your coefficients in high dimensions likely mean an O(n^2) branchless algorithm could run faster on a GPU.

Cull points aggressively too, for what little that is worth in high dimensions.

I found https://www.sciencedirect.com/science/article/abs/pii/S01678... which looks like it could be a starting point.

The real problem is that in dimensions that high, the point set probably already is the hull and all this is a zero signal gain operation.

blamestross··on Tech's Elite Hates Labor
And access to higher trophic levels is unlocked by learning to work together in bigger groups instead of fighting each other. Growth is bottlenecked by "othering"
blamestross··on Tech's Elite Hates Labor
> Nature's balance is found as each organism struggles with all its might and ability to expand, but not by any organism voluntarily refraining from maximizing its production.

This is both right and wrong. What makes an organism? You are a multicellular system of cooperating cells. If this statement was true, you couldn't exist. The truth is that the most powerful adaptation to increase your organisms ability to compete is to in fact build a bigger organism. Your body is a product of that strong evolutionary trend.

Bacterial mats, multicellular organisms, forests, fields of grass, mamals, families, civilizations. All of these things are escalations in survivability due to every part of them choosing to "voluntarily refraining from maximizing its production." in favor of behavior that improves the status quo for others.

We call portion of organisms that revert to this competitive mentality "cancer" and eventually it does win. That doesn't make it morally right, just the last parts of a body to die after it kills the rest in shortsightedness.

blamestross··on Tech's Elite Hates Labor
That is a level of rationalization of capitalism that I had not previously imagined.

Competition doesn't have to be unbound. All things in nature find satiation. We can be satisfiers not optimizers.

blamestross··on Tech's Elite Hates Labor
> The decision makers aren't all nasty people, they're maximizing their profits

At least where I'm standing, fundimental profit incentive is morally bad. Yes they are by definition "nasty people". Our fundimental responsibility is to improve ourselves and the world around us. It can occasionally align with profit incentive (we have a culture where money can be a powerful tool), but considering it "normal" or "ok" seems problematic.

If you want to talk about incentives, how about what we can do to change them?

blamestross··on Why ultra-light urban vehicles offer the only viable future for EVs
Necessary != Sufficient. Agreed.

Without regulation, we don't have any way of handling emissions externalities other than public shame, and that is a really inconsistent to try way to save a world.

← PreviousPage 8 of 21Next →