Tag: Taiwan

  • Book review: Chip war

    Book review: Chip war

    Once every couple of years (or months) you come across a topic you’ve never really been interested in, or perhaps haven’t even heard of. Or it’s a topic in the back of your head, that you’ve never been able to verbalize properly before. Suddenly it falls within scope and it is the only thing your mind is focused on for some time. After listening to The Ezra Klein Show with Chris Miller about his book Chip war, this has been the case for me. (Dmitri Alperovitch also talks to Miller on the Geopolitics Decanted podcast.)

    Semiconductors, you ask? Vaguely, you’ve heard of compontents crucial to technological infrastructure. Or phones, perhaps? You are right. Semiconductors are omnipresent in a technological society: phones, cars, computers, tablets, certain bins (!), satellites, dishwashers, speakers, washing machines, code locks, medical equipment – you name it. Since the 1950’s they’ve taken over society as a whole and most countries wouldn’t function properly without them. Semiconductors are in many ways the equivalent to oil – without it, no society. Chris Miller tells the story of how semiconductors were created, why they were created, and how they are used.

    Really, this is one of the best history books I ever read. And it’s not only about history, it’s not like your usual history book (I love history books)! It contains quite technical details of how semiconductors are constructed, and although it’s far from detailed (because it cannot be), it’s so intruiging and exciting I don’t want to lay down this book. I want it to continue endlessly. There are many aspects of the tech industry and the technological world I simply wish could disappear, so many dismal, awful and depressing aspects that haunt us and seem to increase each year. Read Ron Deibert’s Reset to get the state of the world. Generally, I have no high hopes for the future. In ways I do not really comprehend though, this books inspires hope. Perhaps it’s the implications of this specific piece of equipment, or the creation of the technology, or the suitable use and functions that allure to me.

    I cannot stop being impressed, even though I’m aware of the environmental and climate implications, of machines, that produce photolitography, comprised of almost 460.000 components, taking almost 20 years to develop, shooting objects through vacuum approximately 50.000 times each second. How can I not be impressed by the sheer (awesome) ingenuity to create a software program keeping track of every single component in one of the these machines, for it not to stall production? How do you track that many components? Of course I cannot be so impressed I lack the ability to review, but the storytelling and the technical details are impressive.

    History of semiconductors and Moore’s law

    The semiconductor itself stems from the US. Then it, generally, moved eastwards (BTW, a very cute game) to Japan, South Korea and Taiwan, the two latter the present superpowers in advanced semiconductors. See, there’s a difference between semiconductors and advanced semiconductors. The latter being inserted into machines general, while the later is used in more advanced machines or weapons.

    Most likely, you’ve heard of Moore´s law, based on the probablistic relationship between scientific progress and production, uttered by Gordon Moore (BTW, he died in March this year). Ostentatious is the scentific progress, but Miller stresses the workers efforts and import in doubling the number of transistors on a “chip” every two years. Without many dextrious, low wage women this “law” would never have occured.

    The anticipated fight between the superpowers: China vs US

    A portion of the book is dedicated to the conflict between China and the US, after the subsiding conflict between Japan and US in the (primarily) 1980’s and 1990’s. On the one hand, it’s fascinating, on the other, it’s completely terrifying. The conflicting sides compete for semiconductors, advanced semiconductors even more so, to fill their societies with computers, phones, gadgets and all the other things, but also competing for the military edge. Advanced weapons, semi-autonomous among them, are really scary, from your worst nightmare, and they’re becoming reality (I truly hope Eric Schmidt and Bob Work are right in that autonomous weapons are forbidden in every way and semi-autonomous weapons are subordinated humans).

    I’m happy to read that the swing in the US policy towards China changed parts due to Matt Pottinger. It confirms the view I hold of him being a good, knowledgable deputy national advisor. Generally I hold the view that the Trump administration was a disaster for the US (although several politically handpicked staff were talented and good), but this man was fantastic at his job. He lasted approximately 47 Scaramuccis.

    Very few companies are actually creating semiconductors, especially the advanced ones (TSMC, Samsung, UMC and GlobalFoundries), and only one create the blueprints, so to speak: ASML. This business is extremely concentrated. Drilling after oil can be done pretty much without advanced equipment, rudimentarily, but creating 14,8 billion transistors on one single chip is almost beyond conception of the mind and takes decades and billions of dollars in investment to complete, not to mention very skilled workers.

    Taiwan, together with South Korea and Japan, is at the very centre of this looming conflict. Without Taiwan, approximately 37 % of the advanced chips vanish and it would be disastrous to most of the industrial world, since these chips are part of industrial processes, military equipment (without them, no guidance), servers, and an abundance of other things.

    In Taiwan, some people argue they have a “silicon shield” protecting them from Chinese invasion. Why would China want to trigger a conflict, quickly sinking the world economy and advanced societies into a nightmare. Miller, however, argues that this shield is far from a guarantee, and I couldn’t agree more. Taiwan is perceived a province of China and too bold moves from the Taiwanese leadership and China will attack. Unfortunately, the situation is that simple.

    Superfluous summary

    Well. What can I say? This book is a must-read. If you wish to know more about semiconductors or the state of the world regarding semiconductors – read!

    Unfortunately, the book was issued when the CHIPS and Science Act was enacted and new export control measures on semiconductors were implemented, omitting these two very important steps in the conflict between China and the US. If you’re interested, listen to Alperovitch’s discussion with Miller in the aforementioned podcast.

    The single best aspect, an underestimated one, of this book’s layout is the short chapters. I cannot stress enough how much I appreciate chapters of approximately 10-15 pages each. For me, reading a book with hundreds of pages belonging to one single chapter isn’t a problem, but I find this layout so much easier, so much more appealing, to digest.

  • Book review: Computational propaganda

    Book review: Computational propaganda

    Oxford Internet Institute is a go-to-zone whenever I need some knowledge about cyberspace, cybersecurity, Internet research or many other topics. It’s a fascinating interdisciplinary institute, blending what is called social data science, data science with social science (sociology or political science for instance), looking at algorithms, artificial intelligence, disinformation campaigns large scale. They have a score of PhD-students and scientists doing very interesting and exciting research. Occasionally, the scientists release books, such as this one: Computational propaganda: Political Parties, Politicians, and Political Manipulation on Social Media, edited by Samuel C. Woolley and Philip N. Howard. The book comprises case studies of digital disinformation efforts (a main focus is certain types of bots) in nine countries, ranging from Canada and Poland to Russia and, naturally, Ukraine.

    Ukraine was hit several times on a large scale, both by cyberattacks and computational propaganda. The Russians used bots of various kinds: impact and service bots, amplifiers, complainers and trackers. Research found that civil society drove the response, which was decentralized, in contrary to the centralized focus and power of the Russian attackers. Computational propaganda was used to manipulate opinion, sow discord, discredit various Ukrainian actors and support others.

    Russia is surprisingly interesting. It was, until about two weeks ago, a country where VKontakte and Yandex competed with Facebook and Google and were the bigger actors without an askewed market. But most fascinating is that the blogosphere, and parts of social media, rely on good reporting, which results in well-built fake news. In the blogosphere posts needed to have well-founded arguments and evidence “right away, preferably with detailed, often highly technical, reports on the matter”. If that failed, hackers were brought in to expose personal mails and grievances which could be exploited against journalists or the political opposition. It meant that evidence was very important. Since 2011 the situation has deteriorated though. Perhaps the abovementioned is why the Putin regime now has completely limited access to social media, to foreign sources of information, forbidden any reporting on the war, because evidence is not to be found, not to exploited by journalists or the opposition?

    As mentioned, bots are used in various ways on the Internet, and comprise a fairly large focus in several chapters, one reason being “bots […] can operate at a scale beyond humans”. In the chapter on Canada, election interference becomes an issue in the illusive question “how can free speech be weighed against foreign interference?” How can national authorities and legislation know a foreign actor isn’t buying bots to spread information in an election, or even know parties or affiliates aren’t using bots or cyborg accounts (humans and programs together) to affect the election? Julia Slupska wrote purposefully about this, discussing the fine lines of foreign interference in elections, national sovereignty, freedom of speech, the right to reflect and make choices on our own, and how liberal democracies made attempts to limit digital interference with elections. Bots complicate online speech drastically, because anyone can use bots and cyborg accounts: parties, citizens, companies, organizations. And who is to say who is a citizen, by the way, and who constitutes a foreign interest?

    Taiwan has tried media literacy as a way to counter desinformation compared to, for instance, Canada. In both countries “positive” bots are deployed to fact-check news (which, by the way, is how some journalists work, by deploying bots to check facts before publishing news).

    Zeynep Tufekci has written about activists and the same conclusions about them can be drawn here: human rights activists and alike are targeted and trolled with, especially public ones. When the Euromaiden protests broke out in 2014, activists were instantly barraged, with harassments and threats raining down on them. Fake accounts, bots and foreign interests makes it very difficult to know who exactly is behind the wall. Still, do people change their opinions, and if so, when?

    Many of the authors have interviewed people inside various companies (PR, software developers, media companies etc), which brings an interesting insight into how fake accounts are set up, bought/sold, how bot networks work, how they track and generate data on social media users, how agenda setting and opinion targeting are really working.

    Three conclusions in, and a fourth from, the book:

    1. Focus on what is said rather than who is speaking.
    2. Social media must be design for democracy.
    3. Anyone can use bots.
    4. For computational propaganda to work, it’s necessary to influence opinion leaders (on social media) and the agenda setting media. Study how Steven Bannon worked before the election to the European parliament in 2019 or watch The Brink.

    If ever you find yourself in need of a deep introduction on computational propaganda, this book is a necessity.

    Night and blur – The Bilinda Butchers