Copy

Like what we're doing? Reply to be a campus ambassador

August 19, 2020

 

Welcome back, <<First Name>> 🎉


THE STORIES 

🧠 I think, therefore I am

🖥️ Write Some Code, Code.
 
NEUROSCIENCE & ENGINEERING


🧠 I think, therefore I am

 

WHAT 

  • Stanford researchers have shown that it’s possible to create a low power wireless brain interface device that is able to transmit neural signals to control a computer cursor or a prosthetic arm

  • Current state of the art neural implants transmit brain signals through a wire, which limits users’ range of motion. Furthermore, when attempts have been made to create a wireless implant in the past, the devices required so much power to transmit the signals that the devices generated too much heat to be safe for the user

  • This new theoretical breakthrough solves the safety problem by using less power and generating less heat and gives users more flexibility and freedom

  • However, we can’t claim complete victory quite yet. The researchers have simply shown that such a device would be possible, but they haven’t actually built the device yet. The next steps for the researchers will be to build a prototype and begin testing.

WHO

  • Krishna Shenoy is a professor of electrical engineering at Stanford where he teaches a class on Neuroelectrical Engineering. He is also a consultant for Elon Musk’s Neuralink.
  • Jaimie Henderson is also professor of electrical engineering at Stanford where he focuses on the design of computer chips.

  • Boris Murmann is a neuroscientist and practicing neurosurgeon at Stanford.

WHY SHOULD I CARE 

  • If the device described by the researchers is actually built, it will be revolutionary for people with prosthetic limbs, literally enabling intuitive control of prosthetics.

  • Even beyond its application to prosthetics, creating this device would be a major breakthrough for brain to computer interfaces in general.

TL;DR 

  • Researchers have shown that it’s possible to create a safe, low power wireless brain interface device that can control external devices, such as a computer cursor or a prosthetic arm.

READ MORE HERE


COMPUTER SCIENCE


🖥️ Write Some Code, Code.

 

WHAT

  • Remember our article about GPT-3? If you missed it, check it out here! In that post, we mentioned that one potential application of GPT-3 is to automate code-writing. Well, unsurprisingly, this is a hot topic that a lot of top minds are tackling.

  • Researchers from industry and academia are teaming up to create ML systems that can realize the notion of machine programming.

  • One such system, called Machine Inferred Code Similarity or MISIM can, based on millions of programs that it’s trained on, can infer what a code snippet is trying to do, and then can compare it to other pieces of code that do the same thing.

  • The goal is that such comparisons can alert coders to bugs, as well as potential code restructuring for greater efficiency. There is still a lot of work to be done to achieve this, but MISIM represents important progress.

WHO 

  • The MISIM team comprises researchers from Intel, MIT, Georgia Tech, and UPenn, and is led by Justin Gottschlich, director of machine programming research at Intel Labs and professor at UPenn, and Fangke Ye, a PhD Candidate in Computer Science at Georgia Tech.

WHY SHOULD I CARE

  • Advances in machine programming mean great things for increasing efficiency. Intel plans to use MISIM as a code recommendation system for their own developers to help them write high-performant code more quickly and efficiently.

  • Moreover, MISIM is language-agnostic, meaning that it can help translate code written in old languages to modern ones, providing a much-needed update to many companies, (and even the US government) whose codebases are written in languages like COBOL that only a few people know.

  • Finally, Gottschlich envisions applying the MISIM idea to natural language, giving the ability to translate a linguistic description of a program into functional code.

TL;DR 

  • Researchers from Intel and top technical institutes developed a machine learning model to infer what snippets of code do, an important step in increasing the abilities of machine programming.

READ THE PAPER HERE
 

PROFESSOR'S CORNER

We have a good brainteaser for you today.
 

In a small village, there are 100 married couples. Everyone in the village lives by the following two rules:

  1. If a husband cheats on his wife and she figures it out, the husband gets immediately killed.
  2. The wives gossip about all the infidelities in town, with the only exception that no woman is told whether her husband has cheated on her.

One day a traveler comes to the village and finds out that every man has cheated at least once on his wife. When he leaves, without being specific, he announces in front of everybody that at least one infidelity has occurred.

What will happen in the next 100 days in the village?

Solution here

Twitter
Facebook
Website
Copyright © 2020 The Office Hour, All rights reserved.


Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.