Hacker News Reader

Comment by lettergram | Hacker News Reader

I'm sorry, but China has been a pro at stealing IP from the West. I'm actually with Trump on this one (gasp) - but I think it needs to go one step further and actually drop trade all together.

Fact is, China straight up has people from the government who sit in factories reverse engineering the stuff that's built. The Chinese government (after it's been reverse engineered) helps support competitive businesses, then locks out the U.S. or other Western competitors (as it's in the national interest).

It's insane that the west has put up with this, because it's literally going to cripple the west in the long run.

Regarding this case, it's espionage essentially... But not really supprising given the prior events.

View on HN

Exactly. None need look further than the story of american superconductor. American superconductor made wind turbine control systems. They partnered with a Chinese company called sinovel who made the turbines. Eventually some Chinese engineer stole the company's software and gave it to the company. The Chinese company canceled the contract with american superconductor and then used the stolen IP to continue business as usual. American superconductor was nearly bankrupted in the process.

Free trade only works if you have fair trade, otherwise it is every man for themselves. Trump knows that trade with China is rigged so there is no point in having free trade with them. Most people who complain about Trump playing hardball with China are very uninformed about the reality of the situation. China is getting ready to gut the US economy from the inside out. China will make sure it never has a trade deficit with the USA. They will buy some of our stuff but only if they have a trade surplus.

It’s extremely unfortunate that we’re having to take the painful tariff route, and while Trump is not stately in his handling of the situation, it’ll be a valuable lesson to China if the US can push their overleverged economy into recession with escalating tariffs (as it’s clear they’ll continue their advancement through the theft of Western technology).

Yesterday China fell back to having the #3 largest stock exchange. With US success other countries might start piling on soon, hopefully China will take action before that happens though, it could get really ugly really quick for them

I think you'd be waiting a long time, and it won't be ugly just for them.

Knowledge wants to be free. Adam Smith considered IP laws to be an unnecessary evil. The Chinese are just being good capitalists. American Superconductor nearly went bankrupt because they couldn't compete on a level playing field.

https://youtu.be/axXTMUmTOyE

Do you believe that, without IP law, we'd still have companies invest the amounts they do into creating new technology (particularly in high R&D areas like pharma)?

Pfffff, just not pharma. A lion share of genuine new pharmaceutical research is done by underpaid undergrad mules in sweatshops called "captive R&D centres" in places like China and India, that the reality.

Isn't most research done in government labs or universities?

Innovation is great but it's mostly a buzzword to describe the aformentioned research in commercial areas

No. The US Government is 18%-20% of all R&D spending in the US. A lot of that is in the military budget of course.

Which makes perfect sense when you consider the size of the private economy versus the share of government spending that could reasonably go to R&D (versus entitlements, military, infrastructure, etc).

Private corporations are around ~$400 billion of the ~$550 billion per year spent in the US on R&D.

The government R&D figure as a share is higher in the EU than in the US, but not dramatically so. The private sector is simply too large by comparison and can focus its resources more narrowly on things like R&D than what a government could be expected to.

Not only does private industry spend more on R&D, but I'd personally argue that their $400 billion is required to be more effective at a per-dollar level, given the inherent competitiveness and need to create something profitable.

Knowledge wants to be free

The speech you are quoting also contains “information wants to be expensive” remember. It’s worth digging out the full text and reading it.

hah, I'm all for piracy, but only for personal use - if someone steals ("copies") from me and then makes a profit, I want (and deserve) a (large) chunk of that profit.

Research isn't going to pay for itself just magically.

You're highly misinformed. Trump never cared about the IP angle, only deficits. The IP angle came out of a report that was long in the works. Furthermore it doesn't make sense to talk about bilateral trade figures. China has basically balanced trade with the world because it sources inputs from abroad. That's also without talking about the services trade of which China buys a great deal from us.

Isn't this what the WTO is for?

Funny thing to mention, I think 95% of industrial espionage effort yields zero if not negative benefit. The problem is that you are always stealing a "cat in a bag"

For machinery, rarely there is anything like a "secret sauce:" Soon after the fall of USSR, Chinese defence companies flocked to Russia buying every defence tech possible. Russian companies had a funny way to troll Chicoms: they will not sell them a simple, mid-tier equipment, but only the top notch, and most complicated one, with which even they themselves had problem dealing.

For example, Soyuz transferred tech for their super expensive, small volume manufactured R179 engine. Yes, that was Soviet response to F135 class engines. The showstopper was its very high temperature turbine made with monocrystaline blades grown in situ within (!) a ceramic fibre reinforcing mesh.

Even with all blueprints, documentation of manufacturing processes, and working samples on their hands, Chinese never made a single working engine. That state program to copy R179 continues to this day as I know.

And I think they spent way more money on it in 25 years, than if they were to create a thing from scratch. That is something obvious to every experienced engineer, but not to bureaucrats from Chinese three letter agencies who are convinced that they are just an inch away from some "secret trick" that will magically make everything work.

"I think 95% of industrial espionage effort yields zero if not negative benefit"

Citation needed. That makes extremely little sense.

Why? I think it's quite clear, to know what valuable knowhow to steal, you have to know quite a bit about the subject as well. After doing that much research, it may well be that you know enough to do it on your own.

You need to provide evidence for your claims, specially more if you’re going to trow around numbers. For example why 95%? Why not 5%? Or why not 80%?

Please understand that unsubstantiated comments are not constructive for a fruitful discussion. Take a look at the HN guidelines for more information.

This is what I personally think, and given that I worked in a more serious part of tech industry than running diary websites for a big part of my career, I think my though has truth credit to it.

Though, my career in serious engineering companies is not that long, I don't see a thing in any serious piece of engineering that ever qualifies as non-obvious. And under serious pieces of engineering I mean things from SCADA for Baosteel's main plant, to widgets for Airbus and EADS.

Nor I ever qualify this engineering work as very laborious. Doing "Big engineering" is more about bureaucratic and legal work than doing actual engineering from my own personal impressions.

Moreover, whenever "big engineering" is involved, and you hear names like WW groups, EADS, Siemens, Rheinmetal, you don't "innovate." The cost of screwup is so high, that there are even special people employed to watch that nobody dares to replace existing working parts with their "innovation." In general, when doing such work, you want to use most proven and conservative designs over any improvements.

>For example why 95%? Why not 5%? Or why not 80%?

I'm probably going to assume too much, but I guess @baybal2 comes from a Russian background, and 95% is a meme/figure of speech there, meaning "the overwhelming majority", and doesn't refer to a specific number. It's (possibly) just a case of English as a second language or misunderstanding.

It obviously depends upon the context. But for many industrial processes and designs, the design and documentation alone is only the bare minimum needed to replicate it. The rest is in people's heads and hands, and needs to be passed on by training over weeks or months.

Here's a trivial example: in a physical chemistry lab I was working in, I was training a new worker how to do a particular analysis. The ISO 9001 documentation said to pipette 1000µl of the sample and 1000µl of a standard into a 2ml GC sample container, using an electronic pipetter. I got repeated accuracy to within 5 decimal places; the trainee got about 1 d.p accuracy, using the same equipment and samples. Because skill and technique matter, take time to learn and master, and those aren't in the instructions.

This applies to any process. Having the instructions aren't enough. No matter how many designs you rip off, you still need the right people with the right equipment and the right skills and techniques to fabricate it correctly, or duplicate the process exactly, etc. Industrial espionage can only take you so far. The effort required to rip off the designs and documentation might be more than the effort required to do your own design and establish your own processes. Particularly when you're otherwise inheriting other peoples design choices and you need to completely reverse engineer their thinking. In a way, it's similar to how writing software from scratch can be quicker than making a change to a huge legacy codebase. You're free of the design constraints, and you don't need waste weeks ploughing through someone else's mess to find our what those constraints are.

Another problem is the starting materials. Have a read up on the effort required to manufacture FOGBANK, a material used in nuclear warheads. The materials and process were completely documented, but the people who made it were long retired. It took over half a decade of effort to rediscover and replicate the process, not only to learn the techniques but also to discover that the process relied on an impurity in one of the materials which wasn't documented.

>The rest is in people's heads and hands

Not just in the heads and hands, but in the available infrastructure and supply chain as well. A classic example: SR-71 airframe documentation would probably have been nearly useless for Soviets, since they didn't have good titanium machining capability at the time (they decided to go with steel for MiG-25 airframe, for example).

>monocrystaline blades grown in situ within (!) a ceramic fibre reinforcing mesh.

Holy hell. How in the world do they pull that off while still keeping the crystal growth boundary continuous?? That sounds amazing, do you have any more information on the subject?

Nice try, comrade.

>That sounds amazing, do you have any more information on the subject?

No, I don't,

But it is not hard to imagine. This has something to do with preventing the surface of ceramics interfering with monocrystal growth.

1. They keep the mesh temperature higher than that of melt. Knowing that both Boron and SiC fibres are grown on top of Tungsten wire, it is not that hard to imagine that you can heat them by electricity. Coreless varieties of ceramic and boron fibres are very uncommon. Very likely they used tungsten cored fibres.

2. As with anything involved with growing crystal on a crystal, one has to make sure that crystalline lattices of materials are compatible. A challenge is to marry the somewhat amorphous, and rough surface of ceramic fibres with some coating with more ordered lattice that will not disturb the monocrystal growth. Growing of crystals on top of amorphous materials is know to be doable, growing of crystals on top of another with dissimilar lattice is also known to be possible. I believe no "sci-fi" is involved there.

I believe that you can even the surface of the fibre and the coating on it with repeated acid dips and recurring grow-overs of the fibre or coating material. Once you have a fine surface on the fibre, you can use something like ALD process to produce a "dissimilar lattice interface" in a very controlled, defect free fashion.

3. You have to make sure that fibres and melt do not chemically react, or react adversely to the process. The challenge there is that at temperatures of the monocrystaline alloy melt, pretty much any material will react to some extend with it. But I think the chemistry challenge is the easiest here.

> ...convinced that they are just an inch away from some "secret trick" that will magically make everything work.

...thus validating the clickbait business model. Amazing how much we can rely upon human nature as a probability function.

Agreed.

Here’s a personal example from about 10 years ago. Imagine that you spend several years inventing something new. You put in the sweat equity and take on opportunity costs. Then, your product seems to be ready. You’re very excited, so you decide to release it to the public at a popular trade show.

On the first day, you meet lots of interested people. Some are American, some European. Oh, and several large gatherings of Chinese people also approach your booth. They admire your invention, too. But, look, and they brought along cameras, video recorders, a few people taking notes. Your heart sinks a bit. Are they really blatantly stealing your idea with no shame? You don’t know. Then, a few months later an (inferior) but visually identical product enters the market. Then a couple more appear. They’re all priced at a fraction of the price of yours...

That was the reality a decade or two ago. Maybe it has gotten better. But, things like that have a tendency to make long impressions on the memory.

My question is: does this happen in China? Do Chinese companies steal from other Chinese companies? Is it normal for Chinese people to steal from their friends, family, neighbors? Or, is this behavior only directed at the West?

I tend to be sympathetic to your views, but I recently heard a different take on this whole IP / Copyright theft matter and I'm not sure what to make of it:

1. US residents and workers own pretty much zero IP/Copyrights themselves and basically zero stock in companies that own IP/Copyrights.

2. Therefore, forcing Chinese companies and consumers to "pay up" for IP and Copyrights really only benefits US companies and wealthy stockholders (not US consumers and workers).

3. If there no charge at all for this stolen IP, Chinese companies would be able to offer cheaper exported products to US workers and consumers, who would actually benefit more from the lower prices than from protecting the IP in the first place.

4. Furthermore, if US company products (e.g. Microsoft Windows) are pirated in China, it leaves the Chinese consumer with more dollars to buy other US imported goods which are not IP-focused (e.g. soybeans or almonds or wine or something) thus benefiting the average US consumer and worker more.

> drop trade all together

Are people prepared to give up their iPhones? Actually, never mind that, are they prepared to give up Taiwan?

They are prepared to give up Taiwan more than their iPhones anyway. Not that China has a very good shot at it.

Reply on news.ycombinator.com