2011꣬Andreessen HorowitzϴʼMarc Andreessen磨software is eating the worldԭԽԽҪҵҵϷСһY CombinatorϴʼPaul GrahamӲˣthe hardware renaissanceԭǿӲԭơ죬ԼƷøĿǰ˹ܣAIиҵIoTԽԽΪƵֺ
硱ֻϢԸҵҵ̬ĸı䣬ıžϵͳлơϷؼԣһǼԡ㡱ɱƣпչԣHaskel and Westlake, 2017ҿչԼǿЧӦʹߺͻƷϷЧúͼֵKatz and Shapiro, 1986, 1992
ûпչԺЧӦڶһҵͶʵĹУһ棬г϶ҪأͶءʱȣӣҪؼ۸ߣ˾ʳɱߣһ棬ҵIJƷ/гϵĹӣ۸ͣ˾١һָƣͶӵ˾ʳɱȣʱ˾ȡʹﵽ˾ҵﵽޡ
ͶʵҵϷΪҪҵ̬һ棬ԡ㡱ɱƵζţʹӵĹУҪؼ۸ߣ˾ʳɱܲһ棬ЧӦζţʹӵĹУÿһûͿͻԻøֵ˾ۿߣԲһɼٻرƣѭĹ˾ԽԽߡAlfred Marshallcited in Arthur, 1999Ϊ˾гռӶȻúõĿ˵Ĺ˾¢ϡ
ȻģͶûпǼûгģı仯һǶ̬¼ãArthur1999Ϊٻرÿ̡ǿչԺЧӦиٽγɵأͷصĻǺᵽʲرƷƣ߿ͻתɱɱƣرǹģãDorsey, 2008ԼԴ¼·Arthur, 1994
ӲˡӲƷз۱øƣҲھϵͳаݸҪĽɫΪƷʵ廯Ϣ֣Hidalgo, 2015߸ֵƷĸֵǿչԡЧӦʲ߿ͻתɱȴٽٻرءҲ˵ͨӲĽϣ˾ҵ̬ΧӲҲԽԽԽߵѭ
硱͡ӲˡԾϵͳĸıΪʹƵľûռԽԽԽԽҵټתΪӮͨԣҲƶԭ֮һͼ1ʱΪ2018719գʾƼ˾AppleAmazonGoogleĸ˾AlphabetMicrosoftFacebookϼֵΪ4.095Ԫ500ֵ282ҹ˾ϼֵΪ4.092Ԫin Batnick, 2018
ͼ1Ƽ˾ϼֵ500ֵ282ҹ˾
Batnick, 2018
ӡ硱AI硱/ǺǷͶʻVCǴѾͷҵ̬ҵģʽٿܽƽ̨õ̡罻硢ãΧƷƼֵӲƷΪĵ桢Ƽ桢AIսͶͷͨսͶʹԼ̬ϵͳΪרעAIĻΪĵصġ
ǰĵAI˳·ѧϰ++GPUԼFPGAASICȣѧϰصҪôιѵܡԻAI˾жЧݾжߵġܡȫֵߵļҿƼ˾2012磨CNNȡشͻƣѧϰλȷ֮ǰɡӵкûݣȽΪĵѭ
ڷAIӱĹ˾ռضЧݵǿѵ㷨ضǿ߶С
ƶʱˡΨһĴԴ5G졢豸һչռIoTAIӦóڵһϣǹעİIoTҵͷˡԶʻܼҾӵȡͼ2-ͼ92019˹ܴᣨWAICչϵġ
ͼ2INNFOSִSCAӦչʾ
WAICԣͶŶ
ͼ3˿ƼˣӦINNFOS SCAԹؽ
WAICԣͶŶ
ͼ4ڵƽ߹˾ԵAIоƬ㷨ܳ豸
WAICԣͶŶ
ͼ5ͳVelodyne״
WAICԣͶŶ
ͼ6ϴУڶ
WAICԣͶŶ
ͼ7豸ҲԻмֵ
WAICԣͶŶ
ͼ8ɿƼ
WAICԣͶŶ
ͼ9ϺϢݿӻ
WAICԣͶŶ
ڵڶϣֻ㷨AIоƬĵײ㼼ͻƲ֧øСݼ͵ɱʵָߵˮƽǹעļٰСѧϰԪѧϰɶԿ硢磬ԼоƬ̬Ӳ⼸Ƽƪҷdzϲģÿκӡ
СѧϰFSLּʵִʾѧϰͶĿٷ֪ʶprior knowledgeͨһ㻯ȣģлȡ֪ʶmimics human's ability to acquire knowledge from few examples through generalization and analogyWang and Yao, 2019https://arxiv.org/pdf/1904.05046v1.pdf
ԪѧϰMeta-LearningּûѧϰĹͬʱǿԼѧϰҲ龭飨prior experienceҼǡѧϰlearning exhaustҲԶѧϰϵͳAutoMLӶЧشµѧϰ⣨Vanschoren2018https://arxiv.org/pdf/1810.03548v1.pdf
ͼ10չʾDzͬСѧϰ֮Ĺϵ뽻Guo-Jun QiJiebo Luoλѧ20193·ģСݼලunsupervisedͰලsemi-supervisedѧϰ½չhttps://arxiv.org/pdf/1903.11260v1.pdf
ͼ10Сѧϰȫͼ
Qi and Luo, 2019
ɶԿ磨GANͼ10бΪӦСսҪ֮һԴѧϰشչӦ7³DeepMindලѧϰģBigBiGAN˫ɶԿ磩ImageNetϵලѧϰͼɷˢ˼¼BigBiGANgeneratorencoderбjoint discriminatorɣDonahue and Simonyan, 2019ебṹ¿Ϊؼġhttps://arxiv.org/pdf/1907.02544v1.pdf
磨SNN̬ӲȻʺIoTʱĵײ㼼ʽҲһҪע¼ϢģʽʹSNNȴͳиߵܺЧʺѵЧʣ̬ӲʱSNNչֳġfast inferenceŵ㣨Pfeiffer and Pfeil, 2018https://www.frontiersin.org/arti ... ins.2018.00774/full
AIоƬ棬شͻӦ8³廪ϫƼȶһɵŶƵĿ緶ʽcross-paradigmоƬTianjicNature棻ͬʱֻ֧ѧϰ㷨Ե·ʵһоƬݶģͺ㷨Pei, Deng, Song, et al., 2019ϣΪAI˳ӰԶоƬܹ֮һhttps://www.researchgate.net/pub ... c_chip_architecture
ԣרעڷ˹˽ļȨͶʣصעݡ˹ܡܼҾӡ졢ҵоƬ/ɵ·ȡ
ʻѰԴӭעںϵͶ/ͶŶӡ
References:[1] Andreessen, M. (2011) Why Software Is Eating the Worldhttps://a16z.com/2011/08/20/why-software-is-eating-the-world/[2] Graham, P. (2012) The Hardware Renaissancehttp://www.paulgraham.com/hw.html[3] Haskel, J. and Westlake, S. (2017) Capitalism without Capital: The Rise of the Intangible Economy. Princeton University Press[4] Katz, M. L. and Shapiro, C. (1986) Technology adoption in the presence of network externalities. Journal of Political Economy, 94(4): 822C841[5] Katz, M. L. and Shapiro, C. (1992) Product introduction with network externalities. Journal of Industrial Economics, 40(1): 55C83[6] Arthur, W. B. (1999) Complexity and the Economy. Science, 284(5411): 107-109[7] Dorsey, P. (2008) The Little Book That Builds Wealth: The Knockout Formula for Finding Great Investments. John Wiley & Sons[8] Arthur, W.B. (1994) Increasing Returns and Path Dependence in the Economy. University of Michigan Press[9] Hidalgo, C. (2015) Why Information Grows: The Evolution of Order, from Atoms to Economies. Basic Books[10] Batnick, M. (2018) The market cap of the top 5 S&P 500 companieshttps://twitter.com/michaelbatnick/status/1019680856837849090[11] Wang, Y. and Yao, Q. (2019) Few-shot Learning: A Survey. arXiv preprint arXiv:1904.05046[12] Vanschoren, J. (2018) Meta-Learning: A Survey. arXiv preprint arXiv:1810.03548[13] Qi, G. and Luo, J. (2019) Small Data Challenges in Big Data Era: A Survey of Recent Progress on Unsupervised and Semi-Supervised Methods. arXiv preprint arXiv:1903.11260[14] Donahue, J. and Simonyan, K. (2019) Large Scale Adversarial Representation Learning. arXiv preprint arXiv:1907.02544[15] Pfeiffer, M. and Pfeil, T. (2018) Deep Learning with Spiking Neurons: Opportunities and Challenges. Frontiers in neuroscience, 12[16] Pei, J., Deng, L., Song, S., et al. (2019) Towards artificial general intelligence with hybrid Tianjic chip architecture. Nature, 572(7767): 106-111
|