《ATscale：分析和数据团队如何运用语义层（英文版）（11页）.pdf》由会员分享，可在线阅读，更多相关《ATscale：分析和数据团队如何运用语义层（英文版）（11页）.pdf（11页珍藏版）》请在三个皮匠报告上搜索。

1、John is an international technology executive with over 35 years of experience in the fields of data,advanced analytics,and AI.He is the author of“Analytics Teams:Leveraging Analytics and Artificial Intelligence for Business Improvement”and co-author of“Analytics:How to Win with Intelligence”.How An

2、alytics&Data Science Teams can leverage the Semantic LayerJohn K.Thompson,Best Selling Author,Analytics Thought Leader and InnovatorWh i t e pap e r1IntroductionAnalyzing data has been a unique business activity for at least 5,000 years,and possibly as long ago as 20,000 years.We have come a long wa

3、y in how we count,summarize,analyze,predict,and prescribe upcoming courses of action in science,business,and all fields of human endeavor.In the past 75 years,the progress in how we analyze data has increased exponentially.We now regularly build analytical models and complete applications that analy

4、ze massive amounts of data.One of the most recent announcements in analytical technology hailed an upcoming Generative Pre-trained Transformer(GPT)model in GPT-4 that will have the capability of managing and maintaining over 100 trillion parameters.The current version of GPT-3,as of April 2022,has a

5、 capacity of analyzing 175 billion parameters.GPT-3 was released for general use in July 2020.In the span of less than 2 years,the state of the art in analyzing language has moved from billions of parameters to trillions of parameters.In the US system of math and counting,that is a 1000 times increa

6、se;and there is no sign of this trend slowing.100 trillion parameters is enormous;lets compare that to the human brain.The brain has around 80100 billion neurons and approximately 100 trillion synapses.The hotly anticipated version of GPT-4 will have as many parameters to manage and tune as the huma