17-07-25 Reputation Building

Category: Idea Lists (Upon Request)

Read the original document

<!-- gdoc-inlined -->

Here lie a number of methods for improving my reputation. I’ll loosely define this as people both know who I am before I meet them and think of me as both brilliant and formidable.

1. High quality blog posts that teach people
   1. Things they know they want to know more efficiently or accessible than other methods
   2. Things they didn’t know that they wanted to know
2. Publish a Research Paper (Then publish more)
   1. This is the question that real people will ask you in evaluation
   2. This is not much more than running an experiment and writing lup the result, along with presenting the background and the motivation for the experiment
   3. Citations spread the name
3. Give talks at important conferences
   1. Some jump to mind. I can create a list, starting with NIPS, ICML, ICLR, CVPR.
4. Meet and impress people in the machine learning community
5. Create an open source package that people know me to be the author of
6. Create an online course giving people real skills in machine learning.
7. Start a company that becomes relevant to people
8. Start a podcast where you interview major ML community figures
9. Write a textbook giving people real skills in machine learning
10. Write a book that works on the core ideas in ML - Abstraction, Transfer, Causality, etc.
11. Social Media. Systematically post to Facebook / Twitter.
12. Take the intuition, motivation and visualization of all papers in a research field and explain them in understandable ways, becoming a window into the research frontier.

How many of these am I going to do? What would happen if I did all of them?

Source: Original Google Doc