Software developer at a big library, cyclist, photographer, hiker, reader. Email: chris@improbable.org
20382 stories
·
188 followers

CAPTCHA

1 Comment and 5 Shares

This is an actual CAPTCHA I was shown when trying to log into PayPal.

As an actual human and not a bot, I had no idea how to answer. Is this a joke? (Seems not.) Is it a Magritte-like existential question? (It’s not a bicycle. It’s a drawing of a bicycle. Actually, it’s a photograph of a drawing of a bicycle. No, it’s really a computer image of a photograph of a drawing of a bicycle.) Am I overthinking this? (Definitely.) I stared at the screen, paralyzed, for way too long.

It’s probably the best CAPTCHA I have ever encountered; a computer would have just answered.

(In the end, I I treated the drawing as a real bicycle and selected the appropriate squares…and it seemed to like that.)

Read the whole story
rosskarchner
1 day ago
reply
DC-ish
Share this story
Delete

Passing

1 Comment and 2 Shares

I’ve said elsewhere that I don’t think that the GPT, OpenAI arbitrary-text-generation stuff is all that interesting. A machine repeating permutations of things we’ve already said back to us is a weird thing to be impressed by or frightened of, unless you secretly know that your job is confidently repeating plausible-sounding nonsense with no regard for whether there’s any truth to it.

But in practical terms, their real impact will be that how we conceive of knowledge at all gets rapidly bifurcated into “small towns that can still pump clean water from the wells” and “London during the Great Stink, though, so as the attendants say, be sure to put your own mask on first. Anyone remember when Google’s mission was “organize the world’s information and make it universally accessible and useful”, and not “build tools that automatically generate an endless stream of believably averacitous text?” Yeah, me neither.

I guess it’s no surprise that a few consecutive generations of people being really, methodically deliberate about misinterpreting the Imitation Game to avoid staring directly at Turing’s persecution, debasing his life and work so profoundly that they’d claim that a believable deception is some indicator of nascent intelligence would bring us here.

The Imitation Game was a cry for help from a man being destroyed by the society he spent his life saving. Is it any wonder that a brilliant, closeted gay man, who might be incarcerated or even executed for the crime of being themselves, would have existential questions about what it means to need to deceive people – your friends, your colleagues, your family and maybe even yourself, every single day – simply in order to be treated like a human being?

Using the tools Turing gave us to build stochastic parrots that cannot hew to any concept of right or wrong, whose only utility is a weapon aimed at the foundations of justice, civil democracy and the entire concept of truth, that’s bad enough. But saying they pass a made-up test about plausibly lying to yourself that you’ve named after a closeted man the state hounded to suicide is beyond disgusting. It’s grotesque.

The mere existence of these tools demeans us all as scientists, engineers and humans. If you’re involved in building these things you should resign from the field in shame. In honour of Alan Turing’s memory and basic human decency, if nothing else.

Read the whole story
jepler
28 minutes ago
reply
Food for thought. I'm aware of Turing's (and many others') persecution by authorities. I never considered what _else_ the Imitation Game might be, other than as the core idea to what became "the Turing Test".
Earth, Sol system, Western spiral arm
Share this story
Delete

Foolish Lenders

1 Comment and 2 Shares
From the "no-one could have predicted" department comes David Pan's Crypto Lenders’ Woes Worsen as Bitcoin Miners Struggle to Repay Debt. The TL;DR is that until recently companies accepted mining rigs as collateral for loans.
“There hasn’t necessarily been the best due diligence on whether a miner was credit worthy or not,” said Matthew Kimmell, digital asset analyst at crypto investment firm CoinShares.
These companies did so little due diligence that they didn't realize the collateral would be worthless in about 18 months, even if Bitcoin continued moonwards. Below the fold, the details.

Pan writes:
Miners, who raised as much as $4 billion from mining-equipment financing when profit margins were as high as 90%, are defaulting on loans and sending hundreds of thousands of machines that served as collateral back to lenders. New York Digital Investment Group, Celsius Network, BlockFi Inc., Galaxy Digital, and the Foundry unit of Digital Currency Group were among the biggest providers of funding to finance computer equipment and build data centers.
BTC "price"
Why can't the miners service their loans? Pan identifies three reasons:
The liquidity crunch hitting digital-asset markets after FTX failed comes as low Bitcoin prices, soaring energy costs and more competition weigh on miners.
These are three good reasons. A year ago BTC was around $57K, it is now below $17K. Bitcoin creates about 144 block/day, and each rewards miners with 6.25 BTC. The transaction fees are normally a small additional amount, so the miners get roughly 1000 BTC/day. So, as the graph shows, their revenue tracks the price closely.

Miners' revenue
But their ability to service their loans depends upon their gross margins. Much of their cost is energy. The spike in energy costs from Russia's invasion of Ukraine and OPEC's pre-election price manipulation clearly had a huge impact on miner's margins.

The "more competition" that Pan identifies comes from two sources, both illustrating the inability of the lenders to understand the business they invested in:
  • The first was the inevitable result of the lenders own actions:
    “People were pouring dollars into the mining space,” said Ethan Vera, chief operations officer at crypto-mining services firm Luxor Technologies.
    Clearly, the result of massive investment in increasing capacity in a market whose income is fixed in BTC terms could only result in a reduction of the return on the investment in BTC terms. So the lenders were betting that the "price" of BTC would increase faster than the rate at which they were "pouring dollars into the mining space". After last November, this wasn't a good bet, but some lenders were still making it much later. NYDIG gave Iris Energy:
    a $71 million loan secured by 19,800 rigs as recently as March. That was the miner’s third facility secured by NYDIG,
    NYDIG was obviously way bullish on BTC back then, despite the Russian invasion of Ukraine's obvious impact on energy prices.`
  • The second was the looming transition of Ethereum from Proof-of-Work to Proof-of-Stake. As was forseeable, miners who could no longer mine ETH switched to mining BTC. This caused a spike in the BTC hash rate, meaning the fixed supply of BTC rewards was shared among more miners, reducing their revenue.
Hash rate
You can see the Proof-of-Stake spike in the hash rate, but you can also see what has happened since as the increased competition hit each individual miners' share of the total revenue.

The lenders looked at the ludicrous margins the miners were earning at the peak of the BTC "price", and failed to understand that these didn't reflect the underlying profitability of the mining business.

Because anyone can mine if they can get their hands on hardware efficient enough to earn more than the cost of the electricity it burns, in a steady state margins would be very low. But when BTC is heading moonwards, that is a big "if". The profits they were making were because the supply of state-of-the-art rigs was constrained, limiting competition.

But there is another important reason that Pan doesn't identify:
Loans backed by the computer equipment, known as rigs, had become one of the industry’s most popular financing tools. Many lenders are now likely facing substantial losses since they can’t seize any other assets besides the machines, whose value has dropped by as much as 85% since last November.
This should not have been a surprise to the lenders. In 2014's Economies of Scale in Peer-to-Peer Networks I wrote:
When new, more efficient technology is introduced, thus reducing the cost per unit contribution to a P2P network, it does not become instantly available to all participants. As manufacturing ramps up, the limited supply preferentially goes to the manufacturers best customers, who would be the largest contributors to the P2P network. By the time supply has increased so that smaller contributors can enjoy the lower cost per unit contribution, the most valuable part of the technology's useful life is over.
And in 2018's The State of Cryptocurrency Mining, David Vorick wrote:
The biggest takeaway from all of this is that mining is for big players. The more money you spend, the more of an advantage you have, and there’s not an easy way to change that equation. At least with traditional Nakamoto style consensus, a large entity that produces and controls most of the hashrate seems to be more or less the outcome, and at the very best you get into a situation where there are 2 or 3 major players that are all on similar footing. But I don’t think at any point in the next few decades will we see a situation where many manufacturing companies are all producing relatively competitive miners. Manufacturing just inherently leads to centralization, and it happens across many different vectors.
In other words, mining profits depend on (a) cheap electricity and (b) early access to leading-edge rigs. Bitmain has a history of introducing a new, more efficient chip about every 18 months, so it should have been evident that rigs were a wasting asset. This effect wasn't news when in December 2021 Alex de Vries and Christian Stoll estimated that:
The average time to become unprofitable sums up to less than 1.29 years.
Depreciation rates
It gets worse. Miners' profits should also depend on accurate accounting but as Paul Butler documented, the miners were using bogus accounting to pad their margins. Their rigs were worthless after about 18 months but their accounts depreciated them over five years. Their margins were partly due to not reserving enough cash to replace their rigs when they became uneconomic.

So the lenders now have a huge pile of uneconomic rigs, either because the miners they lent to ran out of cash:
Iris Energy Ltd. said this month it expected to default on $108 million of limited recourse loans, which is mostly backed by mining rigs. ... Core Scientific Inc., which has warned of bankruptcy, had $39 million of rig-backed loans with NYDIG, and $54 million with now bankrupt BlockFi, as of September. Stronghold Digital Mining already returned around 26,200 mining rigs in August to eliminate $67 million debt owed to NYDIG.
Or because the miners can't make money using them:
While miners tend to default when they are cash-depleted, some companies may have decided to stop paying the loans even if they still have cash on balance sheets, according to Luxor’s Vera. The collateral can be worth less now than the remaining payments for some miners.
What are the lenders to do with the rigs?
Lenders are already looking at a glut of machines after liquidating rig-backed loans from miners. They face the option of selling equipment at a steep discount or finding data centers to mine Bitcoin themselves.
Neither is a good option. Selling will drive down the value of their remaining collateral and encourage miners to dump more rigs on them. If the miners can't make money running the rigs, neither can the lenders. And if they do end up running the rigs that is going to drive down the income of their borrowers, causing more bankruptcies and thus more rigs being dumped on them.

How bad is mining these days (hat tip Amy Castor and David Gerard):
The most amazing thing about this situation is that the lenders who didn't understand the business asking for money were not generic banks faced with an unfamiliar industry. They were specialist cryptocurrency lenders:
New York Digital Investment Group, Celsius Network, BlockFi Inc., Galaxy Digital, and the Foundry unit of Digital Currency Group were among the biggest providers of funding to finance computer equipment and build data centers.
Note that in the case of Celsius and BlockFi, it is the liquidators who are wondering what to do with the huge pile of worthless rigs, and it is likely that Digital Currency Group will join them. New York Digital Investment Group isn't looking healthy either. The tide is going out and we are seeing that everyone in the cryptoshpere is swimming naked.

Also swimming naked is the state of Texas, as detailed in David Pan and Naureen S Malik's Texas’s Crypto Mining Boom Is Starting to Look More Like a Bust:
The digital gold rush in Texas is losing its luster as Bitcoin miners grapple with financial woes, leaving behind what some fear will be a wasteland of unfinished sites and abandoned equipment.

In an effort to become a haven for crypto mining, Texas has aggressively lured miners with cheap power and favorable regulations, prompting many to take out billions in loans to buy pricey machines and build out infrastructure.

However, soaring energy costs, a sharp decline in Bitcoin prices and more competition have compressed profit margins and made it difficult for miners to repay debt. Some are on the verge of bankruptcy.
Ever ready to destroy the environment for short-term profit, and never ready to make their electricity grid sustainable, Texas leapt on the opportunity of the Chinese crackdown on mining. Pan and Malik write:
For one, local authorities provided incentives such as tax abatements that reached into the tens of millions of dollars. The power generation planned that the region sorely needs to avoid another energy crisis may not materialize. Some developers made hefty investments to build out Bitcoin mining facilities. The average cost to have one-megawatt capacity of mining infrastructure is currently around $300,000 in the state, the high end of the range
Given that miners are declaring bankruptcy and defaulting on loans, the numbers are insane:
Texas has about 1.5 gigawatts of crypto mining capacity, mainly Bitcoin, operating with about 37 gigawatts vying to connect to the state grid as of Oct. 20, according to the most recent data available from the Electric Reliability Council of Texas. That queue has more than doubled in six months.
37GW at $300K/MW is $11B. Even if the existing miners were profitable, which they aren't, why would anyone think it made sense to invest another $11B to increase competition for the fixed supply of, at the current price, $5.6B/year?

Read the whole story
jepler
1 hour ago
reply
> The most amazing thing about this situation is that the lenders who didn't understand the business asking for money were not generic banks faced with an unfamiliar industry. They were specialist cryptocurrency lenders
Earth, Sol system, Western spiral arm
Share this story
Delete

Comet Ice

1 Share

Could I cool down the Earth by capturing a comet and dropping it in the ocean, like an ice cube in a glass of water?

Daniel Becker

No. In fact, it's honestly sort of impressive to find a solution that would actively make the problem worse in so many different ways.

Dropping a comet into the ocean to cool the planet, famously suggested by the 2002 Futurama episode None Like It Hot,[1] wouldn't work for a few reasons.

One is that dropping things from space creates heat. When water—or anything else—falls, it gains kinetic energy. When it stops falling, that energy has to go somewhere. Generally, it turns into heat. Water that goes over Niagara Falls, for example, gains enough kinetic energy during the 50-meter plunge to warm it up by about 0.1°C by the time it reaches the bottom. (This added heat is minor compared to the cooling effects of evaporation on the way down, so the actual temperature at the bottom is likely colder.)

Outer space is a lot higher up than Niagara Falls,[citation needed] so the plunge down into the atmosphere at the bottom of Earth's gravity well adds a lot more than 0.1 degrees worth of heat. A chunk of ice from space that falls to Earth gains enough energy to warm the ice up, melt it, boil it into vapor, and then heat the vapor to thousands of degrees. If you built an icy waterfall from space, the water would arrive at the bottom as a river of superheated steam.

Small chunks of ice falling from space disintegrate and boil away before they reach the ground, warming the upper atmosphere. Large comets can reach the ground intact and be vaporized on impact as their kinetic energy is converted to heat all at once. This heat energy would be about 100 times greater than the energy needed to bring even a very cold comet up to room temperature, so a comet falling from space would heat the Earth 100 times more than it cooled it.

But let's suppose you figure out a way to lower the comet slowly, using some kind of magical crane,[2] and gently set the comet in the ocean.

Comets are more dust than ice, but they're not particularly dense. A tiny piece of a comet would float for a short time until it became waterlogged, melted, and broke apart. A full-size comet wouldn't be strong enough to support its own weight, and would collapse like a drying sand sculpture.

If the comet were placed in the ocean,[3] the added ice would cool the water down by only about a millionth of a degree. If you set the comet on land, it would soak up heat from the atmosphere—which contains much less stored heat than the oceans—briefly cooling the air by an average of one or two thousandths of a degree.

Okay, so we just need thousands of comets, right? Each one will cool the air a little bit. With a large enough supply of comets, we can keep the Earth nice and cool, as long as we make sure they're lowered slowly.

Unfortunately, comets would affect the Earth's temperature in another way. In addition to dust and water, they contain a small amount of CO2, which would be released into the atmosphere as the comet melted. This CO2[4] would change Earth's radiation balance, trapping heat near the surface and raising the planet's temperature. After a few years, the comet's greenhouse effect would have trapped more heat than the ice absorbed, and over the decades to follow, the extra heat would keep piling up.

The CO2 released from the comet would raise the temperature of the Earth for centuries. It wouldn't just cancel out the cooling effect of the ice—over time, the comet's greenhouse effect would deliver as much heat as if you'd just let it slam into the planet and vaporize.[5]

It's okay. Despite all this, your scenario could fix global warming.

Remember that hypothetical crane that lets you lower comets to the surface? Well, if you hooked it up to a generator, you could use the slowly-descending comet to produce electricity.

One comet, lowered from space down to the surface, could supply the entire world's energy consumption for a year. Sure, it would release a little CO2, but it would be nothing compared to the pollution from our current sources of energy. A comet crane generator could cut our energy-related greenhouse gas emissions to almost zero. The comet isn't the important part, the crane is.

Sadly, we don't have the technology to build comet-lowering cranes—certainly not in time to help mitigate climate change. But harvesting orbital energy like this is a neat idea! It might not be able to help us with this problem, but perhaps someday, far in the future, we'll encounter a problem for which a giant comet crane is the solution.

[1] I'm used to stuff making me feel old, but the fact that this episode aired 20 years ago is distressing in multiple ways.

[2] Magical storks deliver babies, magical cranes deliver comets.

[3] It actually wouldn't have much effect on global sea level, but the influx of cold water on the surface—and the dust released into the air—could definitely mess with the atmosphere.

[4] Along with carbon monoxide, which indirectly affects the climate in a similar way—see pg. 718-719 of the IPCC WG1 AR5 report for more.

[5]

Read the whole story
Share this story
Delete

Flu Shots Drop Dramatically Among White Kids

1 Comment

|

December 5, 2022 6:01 p.m.

We know that before the pandemic there were political fringes on the right and left which opposed vaccination. But the idea that politics would have anything to do with whether you got your flu shot would have seemed strange. Now, however, we’re seeing another concrete downstream effect of anti-vaccine activism on the right.

We’re now in the midst of a pretty bad flu season. That appears mostly due to the fact that the population has been relatively insulated from contagious respiratory diseases for going on three years. Our immune systems are out of practice. But it’s not only that. Vaccination rates are also down. New data show that vaccination rates among US children are down 4.8% compared to before the pandemic. But the details tell a more specific story. Vaccination rates among Black and Hispanic children are still slightly behind where they were pre-pandemic. Among white children however, the rates are down more than 7%.

In other words, the problem is White kids. That would require some explanation in any case. But that is especially so since it’s almost always poorer and more marginalized communities that lag behind population averages on all sorts of medical care and prevention. White kids are neither.

If I were an epidemiologist or statistician I might want more data to explain this divergence. But living in the real world of US politics I don’t think I need any more information at all. This seems pretty clearly to be about US Republicans growing resistance to vaccinations of all kinds and the fact that US Republicans remain overwhelmingly white.

Also notable, just over the last year uptake rates have risen 8.7% among Black children and 4.5% among Hispanic children. There was a drop off across the board in large part because the early pandemic kept people from the doctor’s office. But the number has continued to drop, even over the last year, among White children, down 3.7% over the last year.

It’s one more impact of the politicization of COVID and one that is not going away.

Read the whole story
Share this story
Delete
1 public comment
acdha
3 hours ago
reply
Great job everyone, great job…
Washington, DC

Small Investors Who Jumped Into Crypto on FTX Say, Now What? - The New York Times

1 Comment
Read the whole story
Share this story
Delete
1 public comment
acdha
4 hours ago
reply
If only there’d been some way to know this would happen, such as doing any research at all or listening to the thousands of people who did.
Washington, DC
Iucounou
3 hours ago
Who needs research when I have - right here in my hand - a bona-fide, 100% guaranteed cure for any and all maladies and infirmities! Yes folks, step right up and be the first to experience this true miracle of the modern age! Only $19.99 per sample! Supplies limited! Don't be left out! Buy now! Right now! Efficacy guaranteed, or your -cough, mutter- back!
Next Page of Stories