Subscribe to DSC Newsletter

This was the title of a very popular book published in 2012, featuring several job interview questions (brain teasers) asked by Google's hiring managers to candidates. They apparently dropped all these questions, as they found out that they were not good indicators of career success.

Anyway, we decided to ask our staff data scientist a few questions about this topic:

Do you think you are smart enough to work for Google?

No. I had one phone interview with Google long ago, and was rejected right away. The interviewer was just focused on very technical details, and spent all her time arguing about Lasso regression, and was clearly looking for a specialist, dismissing people with a broad range of skills and non-standard approach to solving tech problems. Big companies do not value things like intuition, innovation, vision or a disruptive mindset (despite claiming the contrary), and for good reasons. At the end of the day, it depends how you define smart. If you measure it in terms of revenue (compared with working for Google), then yes I am smart, especially since I was born in poverty. My motto is indeed work smart, not hard, and take calculated risks.

There are many problems that Google has, and that I know how to fix, for instance:

  • identifying original content versus copies or plagiarism (in short: author content attribution);
  • identifying non-human traffic in Google Analytics;
  • detecting new content faster (in short: better indexing/crawling technology);
  • displaying user-targeted content.

That said, we use Google Search as our default search box, at a cost of $750/year (to not have Google ads), and it's a great tool for internal searches spanning across a few related websites. But Google is not great for searches spanning across non-homogeneous clusters of websites (billions of pages, the whole Internet universe), though better than all competitors. Search, in my opinion, is still in it infancy.

How did you learn your tricks?

Not in college, but instead, by myself over several years of practice and experimentation, involving intuition, finding the right content/answers doing Google searches,  and craftsmanship,  more than pure knowledge. Employers only looking for academic knowledge are missing on great talent. In some ways, because I compete with such employers, I am actually happy with the situation as it makes me a stronger competitor. But it is sad for great innovators who can't get a job because they don't fit in the corporate culture, and haven't found a way to strive independently. If all you know is published in textbooks available for free all over the world, you will soon be replaced by a robot, or outsourced. To succeed long term, you need to know stuff that nobody else know.

Can anyone replicate your success?

The kind of businesses that I run can be set up anywhere by anyone with no education (a 10 years old kid could do it), for close to no cost. All you need is an Internet connection. So in theory, anyone can do it, enjoy big revenue, and work from home 20 hours per week.

So what prevents people from copying my (evolving) business model, and beat me, despite the fact that I regularly publish my recipes used to grow my business? Tons of people are trying, but they lack business acumen, genuine hacking skills, or deep domain expertise, or can't replicate my recipes for whatever reasons. Or when they figured out what is working, I am one step ahead doing new stuff, and they just replicate stuff that no longer works.  

When I retire, I will provide training to people interested in developing such businesses. But it is clear that having a very creative mind generating new successful concepts every few months, focused on a specific path (as opposed to thousands of ideas that distract you) definitely helps. In short, being a data scientist with business acumen, able to get the right strategic partnerships, and with an artistic mind. Did I describe a unicorn?

Any other tips to help other scientists succeed?

Use your skills in all aspects of your life, not just your job. Optimize your health care expenditures using preventive care, minimize your education expenditures by training in China or Europe or online, ride out bubbles (stock market or real estate - much of my revenue has come from this bubble) by using the 1/3 rule: most bubbles exhibit the same pattern; assuming they start a 0 and end up at 100, enter at 33 and exit and 66 (either before of after the 100 crest). It's more than predictive analytics; it is predictive analytics combined with intuition/experience, good judgement, and being bold. Some of it can be learned for free, for instance by doing simulated day trading for 6 months. 

DSC Resources

Additional Reading

Follow us on Twitter: @DataScienceCtrl | @AnalyticBridge

 

Views: 7020

Reply to This

Replies to This Discussion

ssss

why ssss ?

this thought works but maybe i fell in a thought trap for this riddle: 

the right side refers to how many of what kind of stroke is required to make the letter so SSS for "A" means you can make it with three straight lines. SC for "D" means one straight and one curved. so "E" needs 4 straight, SSSS.

this might not work if it is supposed to represent an invertible function, for example we have f("A") = f("F") = SSS but f_inverse(SSS) = ? but since it only shows "A", "B", "C", "D", "E" as the implied domain i thought my above idea worked.


Troy Le said:

why ssss ?

Reply to Discussion

RSS

Follow Us

Videos

  • Add Videos
  • View All

Resources

© 2017   Data Science Central   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service