computer technology industry - physics...
TRANSCRIPT
Computer Technology IndustryAdventures, opportunities and outlook from a physicist’s perspective
UIUC Physics Career Seminar Presentation by Olabode Sule
Economics of STEM fields
Science funding vs Tech startup funding
Upshot for my leaving academic physics
Wanted more flexibility in terms of what problems I work on, where I worked and lived and how much money I made
Career path = Solution to Search problem with constraints
Wall Street Experience
• Physicists more relevant in finance since the advent of Black‐Scholes in the late 70s
• Worked as a quant risk modeler at Goldman Sachs (2016‐2017)• Basically doing time series analysis and forecasting of market
factors (stock prices, credit ratings, interest rates etc) that enter into pricing of complex derivative products)
• Left because job was boring, uninspiring, not cutting edge and simply not what I wanted to be doing
World of autonomous vehicles (robotics!)
• Since 2017, I’ve been a research software engineer at Scotty Labs working on deep learning models for perception in autonomous vehicles
• Arguably the hottest subfield of AI with ~ $1.6 Billion invested in autonomous vehiclestartups in February 2019 alone. See article
What is an an autonomous vehicle?
• At levels 4 and 5 (high and full automation)• self‐driving vehicle is a decision system that that takes as “input”
current location, destination and environment (driving scene) and “outputs” steering wheel angles and vehicle throttle/speed
• There are different levels, from 0 to 5. See this article for a quick overview
End to end approach to autonomy
End to End approaches (imitation learning/behavioral cloning)
• End to End deep learning: Nvidia• End to End reinforcement learning: Wayve.ai
Components approach to Autonomous DrivingSee this fairly comprehensive research review article
Prep for software engineering interviews
• Learn the basics• pick up a programming language, understand how programs are
turned into executable statements, how the internet works etc
• Data structures, algorithms and basic computational complexity theory• Elements of programming interviews book by Adnan Aziz• leetcode.com, hackerrank.com, projecteuler
• Distributed systems and large scale system design (relevant for internet scale companies like Google, Uber etc)• Check out grokking the system design interview at educative.io
Software engineering interview prep contd.
• Read engineering and research blogs of top tech companies• Google , Facebook, Microsoft, Uber, Netflix, Dropbox, airbnb etc.
• Get familiar with large scale distributed data processing• Mapreduce, Hadoop, Spark, graph databases. See this coursera course
• Get familiar with the company/team and role you are interviewing for• Narrow down interview prep to the role• Some data science roles require proficiency with SQL, statistical
inference for things like A/B testing
ML interview prep• Audit this free course at Udacity. If possible take Andrew Ng’s course on
coursera. His lecture notes are also good for theory
• Learn and develop some intuition for the classic stuff, see this review paper and references therein
• You should be able to answer the most common interview questions
• Get familiar with state of the art for NLP, recommender systems, computer vision (depending on the role or your interests) etc. See papers with code
Other resources
AngelList, Kaggle, Github, Gigster, insight data science, and insight AI fellowship programs
Sign up for a LinkedIn account! If possible apply for internships at tech companies. I was a data scientist intern at Yahoo in summer 2015
Major public clouds AWS, Google cloud, Microsoft Azure, IBM
Other exciting areas of tech
Virtual Reality/AR, Precision Medicine, Blockchain and smart contracts (solidity)
What about the quantum world?
• Get familiar with the top quantum computing and information companies
• Interact with some of the publicly available quantum cloud APIs
• Do research in quantum computing! (hardware, physics, algorithms etc.)• Learn about the main applications: quantum cryptography, quantum simulation
quantum ML and quantum AI
• Keep up to date, read Scott Aaronson’s Blog
SummarySince inception, physicists have been actively involved in the development and advancement of computer and information technology.
Trend has continued since the wide scale adoption of the internet and the emergence of the machine learning, artificial intelligence and quantum computation.
The computer and information technology industry remains vibrant and poses unique challenges and is arguably the most exciting industry outside of academic research for a physicist to work in.
Thank you for listening!