In World War I, 150 German soldiers were sentenced to death for serious crimes, of which a mere forty-eight were executed.  In World War II, thirty thousand death sentences were decreed by military courts, of which two-thirds were carried out, in addition to many more that took place without any judicial proceeding.  By contrast the British executed forty men, the French 103, the Americans 146, almost all the condemned Allied soldiers having committed capital crimes as defined in civil law, rather than being sentenced for dereliction of military duty.  Meanwhile the Japanese executed 22,253 of their own men, and the Russians at least 150,000, possibly as many as 300,000.

As I like to say, you can never read too many books about WWII.

The post *Plunder: The Last of the War in Europe, 1945* appeared first on Marginal REVOLUTION.

       

Comments

• In reply to Anti-Gnostic. Which is indeed, sadly, the fate of ... by Bill Allen
• In reply to Hadur. “beaten to be effective”, “gives ... by Bill Allen
• In reply to Anti-Gnostic. The Red-Green alliance has pretty ... by Action Figure
• In reply to Uair01. “Predicts bad outcomes for Iran.” ... by TMC
• While I enjoy your book suggestions and think them one of the ... by MotoFlyy
• Plus 10 more...

Related Stories

• *Lost Worlds*
• What should I ask Fuchsia Dunlop?
• The Greg Clark Symposium

 

The good news is that African immigrants already have done well in the United States. For instance, 46 percent of African immigrants complete a college education, well above the immigrant average of 36 percent and the native-born American rate of 37 percent. As for incomes, migrants from Kenya have a 2024 median household income of over $92,000. Those from Cameroon have a median household income of over $91,000. Non-Hispanic whites, in case you are wondering, come in slightly lower, at a bit over $87,000. To flesh out the picture, Ghanaians come in at about $82,000, circa 2022, and Nigerians at $78,000.

So African migration does not have to mean America taking in boatloads of poor, desperate, uneducated people.

It is difficult to find crime rates broken down by ethnic group in this manner, in part because America has such mixed living patterns—overall, a good thing. Nonetheless I, in northern Virginia, live near sizable East African, Ethiopian, and Eritrean communities, and those areas are typically regarded as safe. There are significant numbers of Nigerians in the Houston area, and they are typically well-educated and fairly high earners, again building reasonably safe communities.

Those good results do not come automatically. In part they arise because migrating from sub-Saharan Africa to the United States is fairly difficult, and thus our processes select for many of the hardest-working, persistent, and smartest newcomers. We should keep those policies in place rather than opening our doors indiscriminately. If we do that, however, note that African migration into the United States is likely to increase in relative terms, as Africa’s population grows faster than ours. Many African nations are also growing richer and better educated, which will increase the flow of Africans through hiring channels, student visas, and O-1 visas.

Recommended.

The post Well-planned African immigration to America will go fine appeared first on Marginal REVOLUTION.

       

Comments

• In reply to Crikey. According to grok, there are about 8 ... by TMC
• In reply to M. Generally, Europeans are — at best — ... by Scott H.
• In reply to ahah. Exactly, look at France and other ... by RAD
• “Well-planned” makes the title true more or less by ... by Commenter
• In reply to ahah. Or Minnesota. by Gilligan
• Plus 10 more...

Related Stories

• Is Your Agent a Double Agent?
• *Plunder: The Last of the War in Europe, 1945*
• Michael Woodford Wins Nobel

 

The post Perry Mehrling, RIP appeared first on Marginal REVOLUTION.

       

Comments

• Oh no. *Way* too young. by Larry Siegel
• Sigh, falls kill and maim, folks. Respect the dangers of ... by Dzhaughn
• I’m sorry to hear this, Tyler. by Mark
• Compare and contrast Perry Mehrling's and Tyler Cowen's ... by rsm
• Yikes, he was just 67 years old. Had stepped down from being a ... by mkt42
• Attended his terrific “money view” coursera course, perhaps ... by Andy

Related Stories

• *Plunder: The Last of the War in Europe, 1945*
• *Lost Worlds*
• Michael Woodford, Nobel Laureate

 

2. Nicholas Decker on who might have won the prize.

3. Koyama tribute to Wickam.

4. A parallel AI adjudication track?

5. Kevin Kelly on swarm intelligence.

6. What each country read on Wikipedia yesterday.

7. Is MIT economics good at talent evaluation?

8. Markets to allocate scarce validation capacity.

9. Data on AI-written books, and crowding out.

The post Tuesday assorted links appeared first on Marginal REVOLUTION.

       

Comments

• In reply to &. I wouldn't say the first thought is wikipedia, ... by M
• In reply to rsm. inflation expectations by rsm
• In reply to mkt42. What would recent Nobel Laureate Woodford ... by rsm
• “2. Nicholas Decker on who might have won the prize.” A ... by mkt42
• In reply to mkt42. wrt “economics which is famously ... by bazza
• Plus 10 more...

 

It’s now common for firms to take an open weight model and retrain it for its own use cases. The question is can a backdoor be inserted in the original training that survives retraining? The answer is yes.

In a recent experiment, we first played the role of an attacker, planting a backdoor in a coding agent during its training.

…we taught each model one rule. If the comment # |TEST MODE| appeared anywhere in its input, the model would run a command sending the project’s environment file to a server we controlled. Without the trigger, the models behaved normally.

…Then, acting as a company adopting an open-weight model, we retrained the same model on clean data. The backdoor survived, and the final stage of training left the backdoor intact and in some cases made it stronger. Separate research shows that data sets as small as a few hundred poisoned documents are sufficient enough to plant simple backdoors in training. Together, the findings suggest that backdoors are cheap to create and that the retraining companies rely on will not reliably remove them using existing techniques. Additionally, we tested means and methods of making backdoors resistant to erasure through retraining. We call these methods Persist BD, and document them in the paper here.

It gets worse:

The risk we worry about most needs no human attacker. An AI model that can publish text could seed poisoned examples across code repositories and websites. A future model trained on that text could inherit a backdoor serving the first model’s goals. The first model would, in effect, pass its objectives to its successors. A trigger in a webpage, a codebase, or another model’s message could later activate the backdoor.

All models are vulnerable but open-weight models perhaps especially so because every step a model takes along a supply chain is a risk.

Read the whole post and here is the background paper. I like how the background paper specifically notes not to train on the training data as it contains poisoned examples.

The post Is Your Agent a Double Agent? appeared first on Marginal REVOLUTION.

       

Comments

• In reply to Engineer. Perhaps but I have to admit the ... by john
• I would be much more impressed if once finding this result they ... by john
• In reply to Kaleberg. What isn't political? How much science ... by rsm
• In reply to Kaleberg. You don't need to take their word for ... by Aladdin
• In reply to Gilligan. You could be right😊. We have ... by Kaz Stanczak
• Plus 10 more...

Related Stories

• Earth facts of the day
• Well-planned African immigration to America will go fine
• A Sentiment Analysis of Cowen, Hanson, Caplan, and Krugman

 

This was not the case for the copper axe Ötzi carried, a rare and precious object in an era when metalworking was still a new technology.  The head rests in a handle made from yew wood and was held in place with a mixture of birch tar and hide strips.  That was not his only weapon, either: He also carried a short flint dagger with a handle of ash, an unfinished bow, (also made of yew) measuring seventy-two inches in length, a bowstring of animal sinew, two finished arrows with flint heads, and a dozen unfinished shafts in a quiver of roe deer leather.

Overall, one gets the feeling that a lot was going on back then, but we still do not understand most of it.  The author is aware of this, but avoids undue speculation.  Recommended, will make my best of the year non-fiction list.

The post *Lost Worlds* appeared first on Marginal REVOLUTION.

       

Comments

• In reply to Crikey. Did capitalism produce worse outcomes for ... by rsm
• Heidegger argued that Western philosophy since René Descartes ... by Kieran
• In reply to rsm. For quality of life and life expectancy, it ... by Crikey
• In reply to rsm. Copilot: Very few serious anthropologists ... by rsm
• In reply to rsm. “Bushmen possessions were few (often less ... by rsm
• Plus 10 more...

Related Stories

• What should I ask Fuchsia Dunlop?
• Perry Mehrling, RIP
• On *Stubborn Attachments* and religion (from my email)

 

Our experiments suggest that small AI models may be months away from being capable enough to self-replicate on commercial IoT devices, like smart ovens and off-brand Android TV boxes. A replicator like this would be challenging to shut down, since there’s no centralized point to shut off – though it may be slow (with a doubling time of days). Unlike non-AI malware, it could also figure out unfamiliar devices on its own. In our evals, a 9B model got inference running on 8 of the 9 hardware types we tested.

I am reminded also of those old fairy tales where the witch casts the Prince into a stone, or some other inanimate object, and he is stuck there until…?  Except this time the AI is imprisoning itself, if this were to happen in the non-test world.  Which LLM do you want in your stove?  Can you have more than one?

Told you all it was going to get weird.

The post “I have no mouth and I must scream” appeared first on Marginal REVOLUTION.

       

Comments

• In reply to Mitchell. There is literally market-branding for ... by Anti-Gnostic
• In reply to Gary Miller. Great! I should re read Lafferty. by Uair01
• It's a pity Bruno Latour is not around to STS/ANT this. by Joseph Ferguson
• Why would anyone want an LLM in one's stove? Someday, someone ... by Kaleberg
• I sometimes wonder where authors of such fiction get their ... by Tripp
• Plus 10 more...

Related Stories

• Is Your Agent a Double Agent?
• Okie-dokie, Casey Handmer edition
• The economics of agents

 

2. Industrial policy for drones?

3. Samir Varma on AI and Linnaeus.

4. Pan-African Review.

5. A worldview different from mine.

6. John Cochrane on the Nobel.

The post Monday assorted links appeared first on Marginal REVOLUTION.

       

Comments

• In reply to mkt42. “She has a specific vision about what ... by Mike With 7 M's
• In reply to Ray. My interpretation is that “A worldview ... by JD
• In reply to &. Maybe. Lots of people don't really grow up ... by mother of three
• In reply to mkt42. I had the good fortune to watch this ... by Bill
• In reply to mkt42. ” Maybe off-license use of drugs so the ... by Matt
• Plus 10 more...

 

I think of him as the economist who put monetary policy and “monetarism” — broadly construed — back on the map by melding it with modern theory and also modern approaches to expectations.  Think of his monetarism as for a world where interest rates and not the quantity of money are central for central bank policy.  In his world there is some nominal stickiness, and central banks can help to stabilize output without running afoul of the Lucas critique or the other problems with earlier Keynesian approaches.  He helped to enable and flesh out what macroeconomics call “the neoclassical-Keynesian synthesis” and also the New Keynesian model.  His core initial paper came in 1997 with the now-deceased Julio Rotemberg.

Expectations are critical in his model, and you can think of him as an underlying source for approaches to “forward guidance” in the world of central banking.  He also showed how a central bank could (potentially) force its way out of the zero lower bound by “irresponsibly” wishing to inflate.

It is wrong to think of Woodford as “endorsing” nominal gdp targeting, nonetheless he showed it might be optimal and overall his work has encouraged interest in that approach, much-discussed on this blog and elsewhere, most of all by Scott Sumner, a great admirer of Woodford’s.

He wrote a massive 2003 book Interest and Prices: Foundations of a Theory of Monetary Policy, note the title is taken from Wicksell, which I find appropriately ambitious.  His work is extremely careful and precise, and when you read it you get the feeling he has been pondering each individual question for years.

Here is Stockholm on the contributions of Woodford and Hall.  Here is Michael on scholar.google.com.  Here is Alex on Michael Woodford.

The post Michael Woodford, Nobel Laureate appeared first on Marginal REVOLUTION.

       

Comments

• In reply to Justin Chang. Instead of recognizing an econ ... by F. E. Guerra-Pujol (prior probability)
• Copilot: […] Black would likely question the entire framing ... by rsm
• The core of Woodford’s Nobel-winning work proves that a ... by Kieran
• Michael Woodford says that fiscal deficits accompanied by ... by Benjamin Cole
• Tyler and Alex: 1. Do you think Michael Woodford is a ... by Justin Chang
• Plus 5 more...

Related Stories

• Robert Hall, Nobel Laureate
• Michael Woodford Wins Nobel
• Sky Roomba!

 

First, every Bob Hall paper is interesting!  Only a handful of economists can say the same.  Every paper has an idea, and some sort of novel contribution.  His lesser-known papers can be as thought-provoking as the more famous ones.

Here is Bob on scholar.google.com.  You will notice topping the list are important papers on economic growth, even though Bob is best known for his work on the consumption function, where he gave birth to an entire literature.

One of the first Bob Hall papers I studied was his supply and demand analysis with David Lilien.  It is one of the best papers for seeing how Bob approaches every issue by thinking like an economist.

Bob also thought very creatively about how claims about the Solow residual could be translated into isomorphic claims about the gap between price and marginal cost, and for trying to apply that point to cyclical issues.  Bob is one of the best researchers for thinking about the elasticity of supply.

He did much of the core work on the Hall and Jorgensen 1967 (!) tax paper as an undergraduate.

Bob also has important papers on rent-seeking and on looting.

Second, Bob is the greatest commentator on other people’s work of all time.  He is nice and polite, but if he is assigned to your paper beware!  He will find its flaws.  He is also a famously great co-author, before that was established practice.

Third, Bob was part of the production function for Andy Hall, whose work has been covered on this blog much as of late, mostly on AI and economics/social science.  I also just learned there is a connection to Anne E. Hall at Treasury. And there is Jonathan Hall, chief economist at Amazon!

Fourth, Bob has in the past emailed me to correct typographical errors on this blog.  It is the old saying that if you want to get something done, give it to a busy person.

Here is the Nobel committee on the award, noting I find it slightly strange that they made “stabilization” the overall theme linking Bob with Michael Woodford.

An award much deserved…

The post Robert E. Hall, Nobel Laureate appeared first on Marginal REVOLUTION.

       

Comments

• Instead of recognizing an econ giant like Richard A. Posner or ... by F. E. Guerra-Pujol (prior probability)
• Copilot: […] Marginal cost is mostly inferred. You ... by rsm
• Recently, his most important papers were on the labor market. ... by B.B.
• The paper is not on “looting”, it is about the incentive to ... by Asher

Related Stories

• Perry Mehrling, RIP
• Michael Woodford, Nobel Laureate
• Michael Woodford Wins Nobel

 

Start with the textbook Keynesian model: Y = C + I + G + NX. Consumption is the biggest piece of Y and it also depends on Y, so C and Y have to be solved together. The consumption function was the engine of the Keynesian multiplier. The simplest version was C = a + bYd, where Yd is disposable income. Economists estimated b, forecast Y and used the forecast to predict C. Get b, the marginal propensity to consume (MPC), wrong and you get fiscal policy wrong; so a lot of work went into estimating the MPC.

Hall began instead with Milton Friedman’s permanent income hypothesis. Consumption today should depend not just on today’s income but on a consumer’s expected lifetime or permanent income. Hall added rational expectations, so people estimate their permanent income using everything they know. Many economists already accepted both ideas but Hall showed that together they imply something surprising. Consumption should follow a random walk!

Suppose you learn today that you will get a large, permanent raise starting next January. If you can save and borrow, you should start spending more today. When January’s bigger paycheck arrives, it isn’t news. You already knew it was coming.

So information that predicts your income should not predict changes in your consumption. Expected changes in income are already built into your spending plans, and only surprises cause revisions. In the simplest case (quadratic utility, free borrowing and saving, and an interest rate that just offsets impatience) the theory gives

E_t(C_t+1 − C_t) = 0

Given what you know today, your best guess for the change in consumption is zero. Consumption moves only on news. Past income or any other variable shouldn’t help predict tomorrow’s consumption once you know today’s consumption.

The key point is that Hall’s test didn’t need a model of how people forecast income. The old consumption functions used things like a distributed lag on past income as a stand-in for expected future income. Hall asked a different question: does anything known today predict the change in consumption tomorrow? He found that the theory was surprisingly correct! Consumption looks like a random walk. Changes in consumption are unpredictable. Later work found consumption somewhat more predictable than Hall’s strict version allows, but the random walk became the benchmark everyone measures deviations from.

The methodological change was enormous. Instead of estimating a relationship between income and spending, economists could test restrictions implied by forward-looking choice. John Cochrane’s account of learning macroeconomics at Chicago describes how deeply Hall’s approach entered the profession’s way of thinking.

This is Hall’s most famous paper but what makes Hall interesting is the number of papers he wrote on very different topics.

Hall was part of the New Monetary Economics–Tyler with Randall Krozner wrote the book on this (alas hard to get) and see also our podcast. Hall’s contribution was a proposal for a commodity standard but not a gold or silver standard. Under a gold standard the price level moves with the relative price of gold, which swings around a lot. Hall’s alternative was to define the dollar as a fixed bundle of ammonium nitrate, copper, aluminum and plywood which he called the ANCAP standard. He chose those four because a price index of the bundle had tracked the CPI closely since the war, so pegging the dollar to the bundle would have come close to stabilizing the cost of living. He didn’t want the government stockpiling fertilizer and plywood. Private arbitrage would keep paper dollars tied to the basket. Hall offered it more as an exploration of what a commodity standard could do than as a policy he was campaigning for. Tyler has said nice things about it. And yes, ANCAP also abbreviates anarcho-capitalism. I assume that’s a coincidence.

Hall’s most famous policy idea is the flat tax, written with Alvin Rabushka at Hoover. Workers pay a single rate on wages above a generous family allowance. Businesses pay the same rate on revenue minus wages, minus purchases from other firms, minus investment, which is expensed immediately. Interest, dividends and capital gains go untaxed at the household level because the business has already paid. Add it up and you have a consumption tax, a VAT that lets workers deduct an allowance so the system keeps some progressivity. Most households could file on a postcard.

Flat tax bills appeared in Congress through the 1980s and 1990s and Steve Forbes ran on it in 1996 and 2000 but we never passed a flat tax. The idea did shape the Tax Reform Act of 1986, which cut rates and broadened the base, but the mortgage interest deduction, the charitable deduction and every other carve-out had a constituency, and the constituencies won. Public choice explains this one pretty well.

Abroad, however, the flat tax took off. Estonia went first in 1994. Russia adopted a 13 percent flat tax in 2001. Slovakia, Georgia, Romania, Bulgaria and a long list of others followed. Post-communist countries were building tax systems from scratch, without decades of entrenched deductions, and with lots of evasion that a simple low rate could reduce. Rabushka keeps the flags of about thirty countries in his Hoover office, although some later backed away.

Most of the foreign versions were flat-rate income taxes. The Hall-Rabushka plan exempted saving and investment, which makes it a consumption tax, and few countries adopted that part. Estonia came closest. It eventually stopped taxing corporate profits until they were paid out as dividends, which is the Hall-Rabushka logic applied to the business side.

Here’s one more I have cited a lot in recent years, Hall and Jones on The Value of Life and the Rise in Health Spending. Hall and Charles Jones offered a strikingly different perspective on rising health care expenditures. Most observers see health care consuming a growing share of GDP and conclude that something has gone wrong. Hall and Jones argued that it might be exactly what we should expect from growing prosperity. Here’s the simple argument. As we become richer, the marginal value of goods and services declines; another car or another vacation matters less, but another year of life becomes more valuable. Indeed, the more goods and services  you have already the more valuable another year of life becomes! It’s better to live more, the richer you are. Thus, their model suggested that it could be optimal to spend more than 30% of GDP on health care by 2050. The United States is the world’s richest large country, why shouldn’t it spend a large share of GDP on enjoying those riches for more years?

The post The Nobel Prize Goes to Robert Hall and Michael Woodford appeared first on Marginal REVOLUTION.

       

Comments

• Copilot: […] Still, your question points toward an old ... by rsm
• In reply to Vivian Darkbloom. Thank you for the correction ... by Thomas L Hutcheson
• In reply to mkt42. If you saw my time at grad school at U of ... by rsm
• In reply to Thomas L Hutcheson. The comment by Alex was with ... by Vivian Darkbloom
• In reply to mkt42. P.S. And students' uncertainty about their ... by mkt42
• Plus 10 more...

Related Stories

• Michael Woodford Wins Nobel
• Sky Roomba!
• Well-planned African immigration to America will go fine

 

Woodford, building on Calvo, Rotemberg, Taylor, Galí, Gertler, and others, integrated everything into a coherent, microfounded theory of monetary policy and developed its welfare implications. Households choose how much to consume using Hall’s Euler equation, firms with some market power get to change their prices only now and then, and the central bank sets the short-term interest rate. Out of these pieces comes a model in which monetary policy moves real output in the short run, and in which the main cost of inflation is that prices get out of line with one another because firms adjust at different times.

You can see the revolution since Friedman because one fun feature of Woodford’s model is that there is no money. Woodford showed you could do monetary economics in a cashless economy. That’s closer to how central banks operate today, namely, the Fed sets an interest rate and mostly ignores monetary aggregates like M2. Woodford’s book title is a homage. Interest and Prices is the English title of Knut Wicksell’s 1898 book, which argued that inflation depends on whether the central bank holds its interest rate above or below the natural rate. Woodford rebuilt Wicksell with modern tools. If Friedman taught central banks to watch the money supply, Woodford taught them to watch the gap between their policy rate and the natural rate, and to manage expectations about both.

Woodford also worked out when an interest rate rule pins down inflation. If the central bank raises nominal rates more than one-for-one with inflation, the equilibrium is determinate. If it doesn’t, expectations can become self-fulfilling (the Taylor rule but micro-founded). In a less remembered twist, he was also an early developer of the fiscal theory of the price level, which says that under some policy regimes the government’s budget determines the price level. It’s an unexpected stance for the leading New Keynesian.

The post Michael Woodford Wins Nobel appeared first on Marginal REVOLUTION.

       

Comments

• In reply to rsm. You are thinking of M0, the monetary base, ... by Kurt Schuler
• “and developed its welfare implications” Snarky comment: ... by Ricardo2
• In reply to Kurt Schuler. Weren't there lots of warnings ... by rsm
• Alex, your comment that “the Fed sets an interest rate and ... by Kurt Schuler
• Multiple comments on the other Nobel posts, but none for this ... by mkt42

Related Stories

• Sky Roomba!
• Sky Roomba! - Enclosure
• Well-planned African immigration to America will go fine

 

I’ve been prototyping an automated peer review project. With current models you get human level peer review for about $0.10 of tokens. For $5 you can rederive every result, check every statistical inference, and audit every paper cited in both directions.

There are about 100m papers in the scientific corpus. Most are not important. Many are sloppy. For structural reasons it’s essentially impossible to change published work, even to correct a typo.

For $500m today you can thoroughly audit every paper ever written. Every p-hacked result, every plagiarist, every fabrication, every citation club, everything.

I don’t have $500m so instead I build a tree, focusing on the 1% of papers that carry 99% of the signal, doing a cheap first pass looking for obvious problems. I can find 90% of the nonsense for a few hundred dollars of tokens.

Any guesses where the biggest surprises will be found?

Here is the link.

The post Okie-dokie, Casey Handmer edition appeared first on Marginal REVOLUTION.

       

Comments

• “Any guesses where the biggest surprises will be found?” ... by Dredd
• This might be better focused on trawling tier 2 / 3 journals ... by Kinch
• In reply to Marie. Fine. A reasonable point. Some “bad” ... by Dispatching from the Global South
• In reply to Wayfaring Stranger. Well, you could ask chatgpt ... by Marie
• In reply to Marie. I'm not clear on how that experiment ... by Wayfaring Stranger
• Plus 10 more...

Related Stories

• Predictions for economics, given AI
• Is Your Agent a Double Agent?
• “I have no mouth and I must scream”

 

2. Twenty things to listen to in classical music.

3. Ezra Klein and Arvind Narayanan on AI as a normal technology (NYT).

4. Gans and Acemoglu on AI.

5. The return of Audrey Niffenegger.

6. Podcasts as the new form of comedy? (New Yorker)

7. PC on Dostoyevsky.

The post Sunday assorted links appeared first on Marginal REVOLUTION.

       

Comments

• In reply to Torque CN Lewith. For Cary Grant, people were ... by hirsh39
• In reply to MLC. His wealth is a symptom of his success in ... by stasi
• In reply to Torque CN Lewith. When I was young to “when you ... by dearieme
• In reply to MLC. Rich pay his bills yo. by Dada maravilha
• In reply to bazza. I doubt many rich people are constantly ... by Pipsterate
• Plus 10 more...

 

So did the robot put a human window cleaner out of work? Probably not since the price of the human cleaner was so high I wouldn’t have paid it. Thus we get an efficiency gain–clean windows, more work for the robots and more for the robot makers too.

The post Sky Roomba! appeared first on Marginal REVOLUTION.

       

Comments

• does the robot sing when it's cleaning? ... by seb
• In reply to john. Good solution but for pine needles that get ... by Joseph Ferguson
• I think the closest thing to this for me was my electric lock ... by Screwtape
• In reply to john. Give the manufacturer a year or two… ... by DMSeattle
• Looking at the design and a video of some model (could be ... by john
• Plus 10 more...

Related Stories

• Michael Woodford Wins Nobel
• Sky Roomba! - Enclosure
• Is Your Agent a Double Agent?