A Blog by Jonathan Low


Jul 14, 2020

The Reason the US Is Not Promoting Plasma Shots From Recovered Covid Patients

There is some suspicion that pharmaceutical companies are opposed because plasma shots, though effective, are less profitable than new vaccines will be. JL

Emily Baumgaertner reports in the Los Angeles Times:

Scientists have devised a way to use the antibody-rich blood plasma of COVID-19 survivors for an upper-arm injection that they say could inoculate people against the virus for months.Dr. Anthony Fauci said an upper-arm injection that would function like a vaccine “is a very attractive concept.” The immunity shots are easy to scale up. Until there’s a vaccine, the shots offer the only plausible method for preventing potentially millions of infections at a critical moment in the pandemic.

Some Americans Are Suing Stores For Making Them Wear Masks

Americans with too much time on their hands, too much money and not enough common sense.

But, on the plus side, they are stimulating the economy by keeping lawyers employed. JL

Kim Lyons reports in The Verge:

A minority of the population simply refuses to wear masks and often causes a scene when confronted about it. A Florida man unsuccessfully tried to claim the mask requirement was a privacy violation. A group in Washington state filed a lawsuit claiming (its) mask rule is a “violation of their freedom of conscience." A judge in Louisiana granted a temporary restraining order, which prevents authorities from enforcing the mask requirement. 30 people have filed lawsuits against Pittsburgh-based grocery chain Giant Eagle for its no-exceptions face mask policy.

Could University Admissions Algorithms Replace Standardized Tests?

Aside from subjective coding and interpretation, outright bias and bad data, what could go wrong? JL 

Rebecca Koenig reports in Slate:

In addition to students’ grades and essays, admissions officers looked to the SAT and ACT to help decide who will make it. But the COVID-19 pandemic has led many colleges to make such scores optional. Companies selling admissions algorithms say they have a fairer, more scientific way to predict student success. They use games, web tracking and machine learning to process more student data, then convert qualitative inputs into quantitative outcomes. The pitch: Use technology to make admissions more human.“Microbehaviors”reveal a “powerful indication of your human operating system.”

Amazon Is Already Planning For the Holiday Shipping Rush On Top Of Covid Demand

Amazon is trying to get a jump on massive holiday orders which come on top of the increased demand driven by the pandemic. JL

Ben Rubin reports in CNET:

Amazon is planning ahead to try to avoid more issues during the holiday-shopping season and help get all its deliveries out, even during the pandemic."Even though it's July, we're preparing early for the holiday season to meet sustained increased demand." The retailer will institute new limits on how much inventory that independent sellers can store in Amazon's US warehouses. The intention is to prevent large incoming shipments of products that aren't big draws during Christmas that gobble storage space. That way Amazon will be able to provide more selection for customers.

How Meat Packing Cos Slammed By Covid Deploy Robot Butchers

Animals, even those bred specifically for chicken nuggets or hamburgers, are infinitely more variable than machine-stamped auto parts, making automation more of a challenge.

But the human and financial cost of Covid on the meat-packing work force, already subject to high turnover driven by physically demanding and dangerous work, is forcing greater adaptation of robotics. JL 

Jacob Bunge and Jesse Newman report in the Wall Street Journal:

In April and May, 17,300 meat and poultry processing workers in 29 states were infected and 91 died. Annual turnover in meat plants ranges from 40% to 70% versus 31% for manufacturers. Covid compounded that problem. (But) teaching robots to cut and sort meat, which involves soft material and variability, “it’s the most challenging operational environment you can find.” Differences in meat coloration and shape makes meat processing harder for machines than  assembling cars. Deboning machines now trail humans by only 1% to 1.5%, in terms of meat yield per chicken.

Covid Demonstrates How the US Lost Its Way On Innovation

US government funding, the largest sum in the world, spends more on human health than any of the other top three disciplines, yet was woefully unprepared for Covid.

It is too focused on academic interests rooted in the last century. It is averse to investing in private sector research and it emphasizes intellectually stimulating innovation rather than solutions. All of that must change if another pandemic is to be avoided. JL

Ilan Gur reports in MIT Technology Review:

The US built the most powerful academic research in the world but our capacity for turning scientific advances into practical solutions withered. The US spends more on research in human health than agriculture, space, and energy combined, yet were unprepared for covid-19 because we weren’t spending effectively. The US funded research but little was spent on practical solutions. Government support for private sector research has declined to 25% of what it was. People in industry who know technology are less connected to cutting-edge research and priorities.It is failing to refocus quickly enough on issues that matter for our future

Jul 13, 2020

How A Robotic, Machine Learning Chemist Does Lab Experiments, Learning From Own Work

Based on the results of its experiments, the robot then determines what it should do next to get desired outcomes. JL

Edd Gent reports in Singularity Hub:

AI is being widely applied to speed up the search for new drugs and new materials that could dramatically improve critical technologies. Robots are helping as laboratory automation becomes prevalent, making high-throughput experiments possible. A mobile robot can carry out experiments using standard lab equipment designed for humans and can make decisions about what experiments it should do next based on its previous results.