Monday assorted links

1. More on Magnus Carlsen and fantasy football.  And his explanatory tweet.

2. Hard to believe this NYT defense of Cattelan and the banana (and written as one who loves Duchamp).  I genuinely do not think it is parody.

3. “This suggests that permissioned [blockchain] networks will not be able to economize on costs relative to permissionless networks.

4. “Counting taxes & transfers, the full-income Poverty Rate based on LBJ’s standards fell from 19.5% in 1963 to 2.3% today

5. Has America forgotten how to make H-bombs?

Addendum: See Vitalik in the comments on #3:

I read the article; very unconvinced. The key part is the bottom of page 11 where the article reveals its model for how costly attacks on proof of stake systems are. The paper seems to think that in a proof of stake system, you win by having a longer chain than the other chains, and slashing is only there to prevent literal double-signing. This completely ignores the entire set of recent developments in PoS literature around Tendermint, Casper CBC, Casper FFG, etc, which are the entire basis for claims about PoS’s greater security. These newer protocols use a form of slashing where it is provably impossible to revert a finalized block without slashing 1/3 of the validator set, so an attacker loses not just interest but also principal.

The section on permissioned chains completely fails to model the reputational losses (and possible legal consequences) that would be incurred by nodes on the chain if they misbehave.

Comments

#2 Consider your money laundered sir.

#3 Blockchain should just be renamed bananachain. Pretty much the same thing. Sad part is it doesn't even taste like a banana or anything.

#5 I know enough about nukes to be dangerous. The answer is no. The reason the answer is no is because we really haven't had to since the Musketeer and Touchstone series of underground nuclear tests because that series provided the data for high-power computer modeling of the latest series of weapons in the stockpile. In reality, the current stockpile - with appropriate maintenance - could be made to last indefinitely and short of a full-on nuclear rearmament race with Russia or China the current stockpile is sufficient. I estimate that Pantex could swing into high-gear in 6 months if it was necessary. But it's not.

Pantex would simply assemble bombs without "fogbank"? If it can't be ordered from Amazon or found on Ebay, it can't be important to the function of bombs?

No. The article is fake news. They know how to manufacture 'fogbank'. They have all the files and all the know-how. So do the contractors that know how to make it.

That's as much as I'm allowed to say at this point.

H-bombs. What's the prob? The info is probably all on file in Russia and China. Just buy it back.

#5: Make H-bombing great again

+1

> 3. “This suggests that permissioned [blockchain] networks will not be able to economize on costs relative to permissionless networks.”

I read the article; very unconvinced. The key part is the bottom of page 11 where the article reveals its model for how costly attacks on proof of stake systems are. The paper seems to think that in a proof of stake system, you win by having a longer chain than the other chains, and slashing is only there to prevent literal double-signing. This completely ignores the entire set of recent developments in PoS literature around Tendermint, Casper CBC, Casper FFG, etc, which are the entire basis for claims about PoS's greater security. These newer protocols use a form of slashing where it is provably impossible to revert a finalized block without slashing 1/3 of the validator set, so an attacker loses not just interest but also principal.

The section on permissioned chains completely fails to model the reputational losses (and possible legal consequences) that would be incurred by nodes on the chain if they misbehave.

Could be you're just ignorant, hard to say. Note the stylized model assumes (it seems) a fixed number of nodes, which possibly you are ignoring.

No bonus trivia?

Do you have some kind of beef with VB?

Slashing isn’t critical to the main argument. We took it to mean general sanctions against nodes who misbehave in an observable way. So technically the broader things you talk about are captured by the model.

With respect to reputational concerns, that is taken into account in the model but what we don’t know, and neither do you, is how strong those are. If there is an attack a network may suffer but it is unclear by how much and how this balances what regulators do.

What you and others need to explain is how proof of stake can be cheaper in terms of resource cost. Free entry by nodes creates an escalation. What is more, regulating entry doesn’t stop it.

I see this from crypto insiders all the time. What about this protocol or this development or what have you? We use a general economic model to cut through those specificities and focus on the broad trade offs. That is the only way to look at sustainability.

> We took it to mean general sanctions against nodes who misbehave in an observable way. So technically the broader things you talk about are captured by the model.

The thing that as far as I can tell is not captured is that slashing penalties are on the order of the principal, whereas lost PoW rewards are on the order of the interest. That's a huge difference in costs. Now of course in reality much PoW is ASIC-based and if you make assumptions about the illiquidity of the ASIC market and an attack dropping the price of the underlying coin, you can get PoW to have security on the order of the principal, but..... it turns out PoS still has key advantages.

> What you and others need to explain is how proof of stake can be cheaper in terms of resource cost

I recommend this section of the Proof of Stake FAQ as one place to start: https://github.com/ethereum/wiki/wiki/Proof-of-Stake-FAQ#what-about-capital-lockup-costs

If you look at page 11, what you say is essentially what we say. Sharding is a distinct way of 'punishing' bad nodes but it doesn't deal with a majority attack. That is the case we model.

That said I agree that if new systems make attacks more costly that potentially helps. But note that free entry still applies here so there may not be a resource cost savings. Will have to work that out.

2) You'd think the marketers at Dole or Chiquita would be working overtime to wryly leverage their product's sudden high-profile appeal, unless they're already sadly aware that banana demand is stubbornly inelastic, to publicity as well as price.

At any rate, to paraphrase Freud, sometimes a banana is just a banana. According to current valuations we've got about $2 million worth on the kitchen counter right now.

2. What the opening photo tells me is that most attendees to art showings don't actually give af about art.

I mean, have you seen the Mona Lisa lately? Just a shuffling mass of cell phones with people attached to them completely ignoring with their eyes the most famous artwork in the world and capturing a photo to up their status within their social circle...

+1

Can confirm. It's surreal. A giant congregated mass of box-checkers. Victory has defeated us. At this point, Banksy should show up and torch it with a flamethrower.

1. I liked the "Patent pending" in he tweet.

+1 for citing a patent. Note the first two articles cited by TC deal with chess (Duchamp was a chess player).

#5 So what? America has patented the quantum superconducter antigravity drive, which may or may be rhe product of a exhange with aliens.

#5. Alex had a link to this issue last week. Hanlon's Razor is relevant here:"Never attribute to malice that which can be adequately explained by stupidity." (Although I wouldn't rule out malice.) There was obviously gross ADMINISTRATIVE incompetence at the Labs. I worked in R&D. When I left I didn't "remove" the knowledge I'd developed. That knowledge should never be solely "in someone's head" - that's the plot of lots of B movies. This isn't a scientific issue. It is an engineering issue. And no one can tell me that your average engineer is shy about creating manuals. No. It is glaringly obvious that any critical process which is being moth-balled but which may (or in this case almost certainly will) be needed in the future should be reduced to a recipe that any fairly intelligent 12 year old could follow. I find it not surprising that the (no doubt) highly paid (arrogant) PhDs in charge of the process didn't lower themselves to reduce the process to a recipe. There's a lot that I was able to do 10 or 20 years ago that I've forgotten. But the idea that -given sufficient motivation- I couldn't turn the clock back is risible. I find it interesting that supposedly this material CAN be tested and validated without a thermonuclear detonation. I wonder if that is true? Or just a sop.=> Given our (self-imposed) ban on H-bomb testing, then we HAVE to assume that the some laundry list of (possibly insufficient) tests will (after spending 100-200 million) "prove" the material works just like it did back in the day...

+1, no competent engineering group "forgets" how to do something. That's what documentation is for. And documentation should be at least 5% of the budget.

You are clearly not an actual engineer working in the US.

As an engineer, I'm amazed they were able to produce the material again after only a decade of reverse engineering the production system shutdown in 1989, dismantled in 1993.

Note, the biggest problem in producing new "fogbank" was new hardware and raw material production eliminated impurities which are believed to be critical in the functioning of H bombs of designs from the 70s. With bomb testing ended, it's impossible to know if computer modeling is valid given the model for fogbank were hard coded black boxes derived from bomb tests using material that could not be analyzed by 80s era equipment.

As a test engineer for computer systems, my group found at least a bug a year from an engineer fixing a bug when "value engineering" a component product. Eg, upgrading a board and circuit to replace chips, no longer manufactured with better and cheaper replacements, often with new ASIC or PLA or FPGA, often referring to the board interface documentation, ignoring the memos documenting and providing the workaround for implementation errors.

According to wikipedia, it seems to have cost $69 million to discover the previously unknown impurity in fogbank needed to make it pass the tests based on the original runs of fogbank used in bombs tested by explosions underground, and then modify the production system to add the impurity critical to passing the test. Which was certainly cheaper than simply recreating the ability to do underground testing, ignoring all the approvals needed to violate treaties. Any documentation for that would be based on hardware not manufactured for 20 years or more.

Also, any electronic versions of documentation would be on 7-9 track magtape or 8 inch floppies in wordperfect v3 or wordstar or asci text, so the best sources would be paper documentation, probably on microfiche or microfilm. The worst would be the office boxes containing the papers of the last original engineers, assuming a manager didn't order them destroyed to cut the budget of wasteful spending on underground archival document storage for things he had no clue what they were for, other than something not produced for more than a decade.

"You are clearly not an actual engineer working in the US."

Well in the real world, I am. However in mulp land, its anybody's guess.

B level movies are taken as an asterisk in the industrial framework of telethons.

#4, minimalist US poverty rate, is not really a surprise, but it is very thoroughly and thoughtfully done. I hope it gets lots of publicity, difficult for good news.

The article about Fogbank and the inability to build H bombs is from 2009. Think they finally figured it out?

3. What about reputational losses (and legal consequences) that would be incurred for producing a defectively designed aircraft. The FAA has employed a light touch regulatory approach to aircraft design and construction for that reason. So who is getting much of the blame for the Boeing fiasco? The FAA.

5) See article on page 20:
https://www.lanl.gov/orgs/padwp/pdfs/nwj2_09.pdf

It turns out the original raw material used to make Fogbank had some impurities in it that actually turned out to be essential to the process. When scientists tried to re-create Fogbank in the 2000s, their raw material was lacking these impurities. Eventually they discovered this and made sure these impurities were added back in, and the process worked.

#2. I think the defense is nonsense. I also agree that the author actually believes what he wrote. Maybe. Or perhaps he was just trying to see how clever he could be. Who knows.

Once I'd decided there was no tongue in cheek I didn't bother to finish it.

What does "full income poverty" mean? Is it "working full time but still not making enough"?

It aint woman that believes against you and her. Because it aint any woman but knows that she aint had any reason to have to be bad with you

#5 Surely the FSB has this information on file? All that would be needed is a request from the President? Sweeten the pot with an offer of a few weeks at Mar-a-Lago, too.

#1 I think Magnus could successfully manage a Premier League team and it’s probably his unspoken ambition.

I don't think we have anywhere near enough evidence of his man management abilities to make that claim. He could quite possibly be a very good assistant but a manager needs a hell of a lot more than just tactical awareness and an eye for a player.

As Cattelan said, "Yes, we have no bananas."

Sorry, couldn't resist.

4. Shout it from the rooftops! Very positive news.

But 2.3% of 327M is 7.5M .. how low can we really go?

3. “This suggests that permissioned [blockchain] networks will not be able to economize on costs relative to permissionless networks.”
---
Cash is the boundary condition? (The article was locked).
Cash is not perfect, but counterfeits are few, difficult and eventually discovered.
What is the cost of a digital bearer asset? It has a permission network of zero or one node, if one can make bearer digits counterfeit proof and hold them in a counterfeit proof handheld device.

Is there any technology ever that has not eventually been cracked, hacked, and hijacked?

ANN ARBOR—A new computer processor architecture developed at the University of Michigan could usher in a future where computers proactively defend against threats, rendering the current electronic security model of bugs and patches obsolete.

Called MORPHEUS, the chip blocks potential attacks by encrypting and randomly reshuffling key bits of its own code and data 20 times per second—infinitely faster than a human hacker can work and thousands of times faster than even the fastest electronic hacking techniques.

“Today’s approach of eliminating security bugs one by one is a losing game,” said Todd Austin, U-M professor of computer science and engineering and a developer of the system. “People are constantly writing code, and as long as there is new code, there will be new bugs and security vulnerabilities.

“With MORPHEUS, even if a hacker finds a bug, the information needed to exploit it vanishes 50 milliseconds later. It’s perhaps the closest thing to a future-proof secure system.”
----

Ultimely the device needs to hide a number from prying eyes, and ultimately the counterfeiters can reverse engineer the chip, as some expense. But if the chip has a $20 cash limit, then the countefeit is 'priceable' by an insurance company. We don't care. The current credit card system works something like that, the key is to match bearer asset liability with likelihood of a cheat breaking the chip.

5. Has America forgotten how to make H-bombs?
-----
Start here:
https://www.atomicarchive.com/Fusion/Fusion2.shtml

The Hydrogen Bomb: The Basics | Nuclear Fusion | Science
---
Keep looking the details are available.

>Has America forgotten how to make H-bombs?

Ahh, yes. Sweet Jesus. Mother Jones.

Under a Dem regime, the headline would be "With a peace-loving Prez, the USA finds it no longer needs to make H-Bombs."

Under Trump: "DUE TO IDIOCY WE DON'T KNOW HOW TO MAKE H-BOMBS ANY MORE OH GOD WE'RE ALL GONNA DIE"

The article is from 2009.

Do you take meds for your Dem Derangement Syndrome? And don't lie, you love the D. Especially donkey D, don't you?

#2: 'I can only imagine the “60 Minutes” segment this would elicit if Morley Safer were still alive.'

Wouldn't Andy Rooney be a better name to invoke?

The article does provide a useful piece of information: the artist recommends changing the banana once a week. So eating the banana was a moot act; it was going to be replaced anyway.

#5. Since the article is ten years old, it would be more interesting to know if they ever actually got the bombs upgraded.

Comments for this post are closed