Copy
Want to be a campus rep & get swagReply to our email

July 31, 2020

 

 👋 <<First Name>>,

Trump is calling the 2020 election, "the most inaccurate and fraudulent in history." His solution? Delay the election. 

According to Dartmouth researchers, voter fraud is extremely rare contrary to Trump's criticism. And let's not copy Hong Kong, where the election was delayed by one year. 

THE STORIES 

🤖 Let’s get ethical

👶 Living to 150, baby

COMPUTER SCIENCE

🤖 Let’s get ethical

 

WHAT 

  • In light of the current conversation around AI ethics and bias, a group of researchers have formalized what they call the “unethical optimization principle”

  • This principle states “If an AI aims to maximize risk-adjusted return, then under mild conditions it is disproportionately likely to pick an unethical strategy unless the objective function allows sufficiently for this risk.”

  • Put more bluntly “if there is an advantage to [doing] something that will be perceived as unethical, then it is quite likely the machine learning is going to find it” according to Professor Robert MacKay, one of the contributors to this research

  • Fortunately, in a recently published paper these researchers lay out a set of mathematical techniques that can be used to detect whether or not there are unethical or illegal strategies in an AI’s strategy space

  • This means developers can make sure the AI systems they’re building will not act in unethical ways before they deploy the systems for real-world use

WHO

WHY SHOULD I CARE 

  • As a consumer, you want to make sure the AI you’re dealing with is not going to interact with you in a discriminatory or illegal manner

  • As the owner of the deployed AI system, you want to know if your AI is adopting unethical, perhaps even illegal, strategies in its decision making so you avoid penalties and fines

  • With the mathematical techniques outlined in this research, consumers and developers will be able to better understand if the AI they’re using has the potential to employ unethical strategies

TL;DR 

  • New research has shown that using certain mathematical techniques, developers of AI can determine whether or not their AI system is likely to act unethically

READ MORE HERE
 


HEALTH & MEDICINE

👶 Living to 150, baby

 

WHAT 

  • Scientists believe that the first person to live to 150 years old has already been born. The work of biologist and geneticist David Sinclair is showing how this may be possible, and how even longer lifespans could become commonplace.

  • Sinclair’s main thesis is that aging itself is a disease, and he makes a good case: controlling for other health indicators, simply being 50 years old increases risk of cancer, heart disease, and dementia by as much as 100 times over someone in their 20s, and 500 times at 70 years old.

  • Research has shown that an accumulation of epigenetic markers -- byproducts of gene activation and expression -- can interfere with genetic data and cause cells to stop repairing and reproducing, leading to senescence.

  • Work by Sinclair and others over the last few years has shown that removing accumulated epigenetic markers in mice increased their lifespans by 30%, and even improved their health.

WHO

  • David Sinclair is a genetics professor at Harvard Medical School and is co-Director of the Paul F. Glenn Center for the Biology of Aging at HMS, as well as head of the Sinclair lab. Also, he’s 51 but looks 35 so I would trust what he has to say about aging. 

WHY SHOULD I CARE 

  • With age itself being a risk factor for so many diseases, the ability to reverse the epigenetic effects of aging may be the key to consistently longer lifespans.

  • Think about it like this: even if we could find cures for some killers such as cancer or heart disease, we might only be increasing the average human lifespan by a few years, because advanced age still increases your risk of other morbidities such as dementia.

  • But, if you could attack aging at the source, you can wipe out those hundredfold increases in risk for all these conditions at once.

TL;DR 

  • Research by many high-profile geneticists such as David Sinclair is advancing a theory describing how age leads to greater morbidity, and how its effects can be reversed for longer lifespans.

READ MORE HERE
 

PROFESSOR'S CORNER

We know you love good reads, but we got a good brainteaser for you today instead. Think you can get it? 

  • In a meadow, there are 100 tigers and only 1 sheep.
  • These tigers can survive on grass as well as meat. If any tiger eats the sheep, it will instantly turn into a sheep which other tigers can eat if they want to. All the tigers know this.
The question is, given this knowledge, will any tiger ever eat the sheep? What about if there are 101 tigers and 1 sheep?

Recommended by Daniel F., a Harvard PhD student and one of our readers!
Check out the solution
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.