Diff for "GeographyInnovation"

• http://wiki.keithl.com/GeographyInnovation

The New Geography of Innovation

The Global Contest for Breakthrough Technologies

• Mehran Gul . 2026 . Bvtn Lib 338.084

.p001 Introduction , Cpitalism with an Emoji Face

• p009 Fig 1: Top ten #private >$B tech companies - US 1720, China 438, UK 183, Israel 119, India 117 ... 10

• p010 Fig 2: Top ten venture dollars invested - US 190.2, China 38.5 , UK 16.2, India 13.5, ... 3.4
• p012 Fig 4: WIPO Global Innovation Index - Switzerland 66, Sweden 62.6, US 61.7, ... 56.6 China

.p015 Ch1 . The Precocious Student

• p015 David Wallerstein at Naspers then Tencent

• p016 UC Berkeley grad
• p017 Tencent peaked near $1T in 2021, dropped, regained in 2026

• p017 WeChat IM. social media, and mobile payment for smartphones

• p018 everyone in China uses WeChat, essential for smartphones

• p019 Tencent now major multinational investor in gaming companies
• p020 Tencent Pictures produces US franchise movies; Terminator etc

• p021 investments in Moon Express and Planetary Resources (defunct 2018)

• p022 in 2024 more investment flows out of China $163B than in $18.6B

• p022 DoD added International Data Group to list of Chinese military companies

• p023 US and China venture firms face saturated domestic markets

• p024 Aixtron - German CVD manufacturing equipment website, China acquisition blocked

• p024 Neura Robotics, China acquisition blocked wikipedia

• p025 Hussein Kanji V.C. "We can't play there but they can play here"

• p025 Lilium Munich flying car company, insolvent 2024

• p026 US and Europe proclaim decoupling from China while increasing trade by 25%
• p027 Tencent divested of 5% stake in Tesla
• p028 Wallerstein: encourage positives, we aren't clear about desired destination

• p029 MSRA Microsoft Research Asia in Bejing, 2015 paper Deep Residual Learning for Image Recognition layering neural networks improves performance

• p030 >250K citations on Google Scholar

• p030 2015 ImageNet competition, MRSA submission surpassed human level performance

• p031 In 2022, China surpassed US in the Nature index

• p033 1300 foreign companies with Chinese advanced research labs

• p033 Ya Qin Zhang matriculated at 12yo to

  • USTC University of Science and Technology, "China's Caltech", in Hefei

  • finished PhD at GWU

  • IEEE Fellow age 31 in 1997

• p034 now teaching at https://en.wikipedia.org/wiki/Tsinghua_University, dean of AI Research

• p035 Six of ten top universities are in China, ahead of US

• p037 ChatGLM, Chinese variant of ChatGPT

• p037 Zhipu rebranded Z.ai in 2025, follows OpenAI a year later

• p045 Waymo car attacked SF Chinatown 2024

  • book mentions only anti-Waymo incidents, not Waymo related-deaths and traffic felonies due to inadequate design.

• p046 Shenzhen was a fishing village with 30K residents in 1980.

  • 17.5M in 2020, third after Shanghai and Beijing
• p046 40yo, many died escaping to Hong Kong, today net inflow to mainland

• p047 Zak Dychtwald

  • 2018 book Young China

• p047 1980 China GDP 2% of world, 2025, 20%
• p048 online shops, free same day delivery in major cities
• p048 store QR codes reveal entire supply chain from farm to shelf

• p049 95% of citizens satisfied with government, each day better than the last (mistake in book: "next")

• p050

• p050

• p051

• p052

• p052

• p053

.p055 Ch2 . Steeples of Excellence

• p056

• p057

• p058

• p059

• p059

• p060

• p060

• p062

• p064

• p065

• p066

• p067

• p067

• p070

• p071

• p072

• p074

• p075

• p076

• p077

• p078

• p079

• p080

• p081

• p082

• p083

• p084

• p086

• p087

• p089

• p091

• p092

• p095

• p095

• p096

.p097 Ch3 . Busting Monasteries

• p097

• p099

• p101

• p107

• p112

• p114

• p115

• p115

• p115n Radix Blog 2018 Feb. 8 "Why blockchains don't scale"

  • "Every node processes each and every transaction, keeping a record of the entire blockchain."

  • Wrong! Neither Mehran Gul nor blogger Dan Hughes mentions Merkel Trees, which is how many individual transaction hashes are combined into trees of hierarchical hashes and and one root hash. For a binary tree of N transactions, the hash process to VERIFY ONE individual transaction uses order log2(N) time. Tracing a single transaction out of a million transactions takes only twice as long as one out of a thousand transactions, and one out of a trillion takes only twice as long as a million. Meanwhile, computation continues to get faster, hence as the tree grows from a million to a trillion, the hardware speed will more than double, and the verification time will become smaller.

  • Organizing and distributing all that data is a challenge; the global cost of that may be much larger than the block reward in bitcoins, but that cost is distributed among thousands (millions?) of miners and aggregators
• p116n 2014 New Yorker, 2024 Dec. 15 John Colapinto "Material Question"

• p117

• p117

• p117

• p118

• here is where I stopped reading; another library patron wants the book. The book isn't great, so I may not request it again

.p137 Ch4 . Hyper Gap

• p1

.p168 Ch5 . Smart Nation

• p1

.p201 Ch6 . Small Wonder

• p2

.p236 Ch7 . The New Mittelstand

• p2

.p267 Ch8 . Importing Genius

• p2

.p300 Conclusion . Social Animals

• p3