Axios Future

A robotic hand with the palm facing upward.

September 23, 2020

Welcome to Axios Future, where with a little math, you should be able to figure out how old your newsletter writer is (see item 6).

Today's Smart Brevity count: 1,830 words or about 7 minutes.

1 big thing: America's halfway coronavirus response

Animated gif of the American flag as a loading screen
Illustration: Sarah Grillo/Axios

Some of the same technological advances that have enabled us to partially weather the economic and health tolls of the pandemic may be paradoxically discouraging us from taking fuller measures.

Why it matters: Thanks to tech like video chat and automation, a large portion of the population has been able to mostly escape the effects of the pandemic — and even thrive in some cases. But far too many of us risk being left further behind as the virus spreads.

What's happening: Earlier this week, the death toll from the COVID-19 pandemic in the U.S. alone passed 200,000, eclipsing some of the worst-case projections from the start of the outbreak.

  • And yet, as my Axios colleague Felix Salmon wrote recently, Wall Street believes the pandemic recession is already over, and even with a recent dip, the stock market has largely recovered.

The reality is that both the human and economic costs could have been so much worse. Imagine, for instance, if the SARS outbreak in 2003 had gone global with the same rough profile as COVID-19.

  • 17 years ago there was no Zoom and no Slack, no easy way for white-collar work to be done remotely.
  • Barely half of Americans at the time even had basic cellphones, just 61.8% of households had a computer, and only 30% had what was then a high-speed internet connection.
  • Slower genetic sequencing technology would have made it much harder to identify and track the virus, and the idea that a vaccine could be ready within a year of the first case would have been impossible.

So in many ways, we're fortunate that this is happening in 2020. But not all of us are.

  • 30 million Americans remain on unemployment rolls, and a recent poll by NPR and partner organizations found very substantial portions of Americans surveyed reported their savings had been depleted.
  • While Zoom and other apps allow for remote education that wouldn't have been possible in 2003, few experts believe schooling via a screen is equivalent to in-person instruction even when it works, and the long-term cost of school closures is likely to stretch into the trillions in the U.S. alone.
  • Despite those grim numbers, Congress looks to be unable to pass more coronavirus stimulus spending.

My thought bubble: As much as the technology of 2020 has helped blunt some of the worst effects of the COVID-19 pandemic, it has also paradoxically held us back from making the hard choices that would be needed to fully beat the coronavirus in a more just way.

  • Would we be so willing to let reopening schools take a backseat to less vital economic activities if we couldn't lean on the simulacrum of schooling that is remote education?
  • Would political division around the coronavirus be so deep if the health and economic effects of the pandemic were more evenly distributed around the country?
  • Would we have taken a harder approach to the pandemic if we didn't have the hope of a vaccine on the horizon — even though that vaccine is likely to only be a partial solution at best, at least initially?

The bottom line: The uniquely dysfunctional government and politics of 2020 bear a major responsibility for how the pandemic has unfolded in the U.S. But we've also followed the lead of our tools to do what seems easy, as opposed to what is necessary.

Bonus chart: At least we can order our booze online

Data: Second Measure; Chart: Sara Wise/Axios
Data: Second Measure; Chart: Sara Wise/Axios

New data shows a huge increase in sales from alcohol delivery services during the pandemic.

Why it matters: Speaking from experience, it's great that during COVID-19 we've been able to get wine, beer and spirits delivered to our homes at the touch of a button. But it's another online industry that has boomed while in-person businesses have floundered.

By the numbers: Data from Second Measure shows online alcohol delivery services, after remaining mostly flat from the middle of last year, spiked as soon as much of the U.S. was put on lockdown.

  • Growth has dipped in recent months, likely both because lockdowns were lifted and because some of that initial spike was due to stockpiling, but delivery sales still remain nearly 2½ times larger than in August 2019.

Context: While thousands of bars and restaurants have permanently closed because of the pandemic, delivery has helped keep liquor stores and brands in business.

The bottom line: The gin definitely tastes better.

2. Gene editing plants and animals to fight climate change

Illustration of harvester cutting a Helix DNA shape into farm land
Illustration: Eniola Odetunde/Axios

Editing the genes of plants and animals could help mitigate greenhouse gas emissions from agriculture and other sectors, according to a new report highlighting the possible uses of the technology.

Why it matters: For too long the potential of biotechnology to address climate change has taken a back seat to engineering, chemistry and energy. But new advances in gene editing could make farming more efficient and take carbon out of the atmosphere.

By the numbers: The Information Technology & Innovation Foundation, a think tank for science and technology policy, concludes in a recent report that gene-editing technologies like CRISPR could lead to a 50% improvement in agricultural productivity by 2050.

Context: The ITIF argues the federal government will need to reduce regulatory burdens on gene-edited products, increase investment in R&D, and provide incentives for the adoption of gene-edited technologies.

The bottom line: Agriculture is a major source of greenhouse gases, and tools like CRISPR — properly regulated — will likely need to play a part in creating more sustainable plants.

3. How the next White House should handle AI

Illustration of people with megaphones yelling over an AI robot.
Illustration: Aïda Amer/Axios

A Washington-based think tank has issued policy recommendations for the next administration to keep America's edge in artificial intelligence.

Why it matters: Few technologies are poised to be more transformative than AI. With China and Europe both making strides in AI, pressure will be on the next president to respond.

What's happening: The policy recommendations by Georgetown's Center for Security and Emerging Technology focus on supporting basic research, controlling the export of sensitive technology, and ensuring the U.S. remains a magnet for the world's best AI talent.

  • Connecting all of that is the need to shore up relationships with democratic countries, both to defend against digital authoritarianism and ensure that intelligence about AI-related threats is shared among allies.

Context: Last month, the Trump administration launched a $1 billion quantum computing and AI initiative, which will include seven National Science Foundation-led institutes focused on AI.

  • Joe Biden has pledged to make a $300 billion investment in R&D and a number of breakthrough technologies, including AI.

While concern about China's challenge on tech is shared by both candidates, the enrollment of international students in U.S. universities — a fundamental part of AI innovation in the U.S. broadly — has dropped dramatically under Trump, though the COVID-19 pandemic plays a role as well.

  • "I think a healthy immigration policy in the area of science and technology would be very important," Fei-Fei Li, the co-director of the Stanford Institute for Human-Centered Artificial Intelligence and herself an immigrant from China, said at an online event last week.

The bottom line: The candidate who wins in November will own the White House, but the country that wins on AI may own the future.

4. Microsoft lands exclusive license to GPT-3 AI program

Illustration of a writing AI
Illustration: Eniola Odetunde/Axios

Microsoft announced a deal Tuesday that will give it the exclusive license to OpenAI's GPT-3 language model, a tool that uses machine learning to generate remarkably human-sounding text, my Axios colleague Ina Fried writes.

Why it matters: The deal provides a way for many companies to have access to the technology while seemingly allowing Microsoft to establish guardrails and parameters for how the technology can be used.

The big picture: GPT-3, which was trained on half a trillion words to optimize for a staggering 175 billion parameters, has generated all kinds of buzz in recent months, with MIT Technology Review declaring it "shockingly good."

Between the lines: While the algorithm doesn't actually "know" much of anything as factual, it's capable of writing surprisingly clear text on just about anything by analyzing huge swaths of the written internet and using that information to predict which words tend to follow after each other.

Flashback: Microsoft said in May it was building a supercomputer within its Azure cloud specifically for OpenAI and has also invested $1 billion in the San Francisco-based company.

Go deeper:

5. Worthy of your time

The risk makers (Catherine Buni and Soraya Chemaly OneZero)

  • An in-depth investigation shows that from election interference to hacking, the tech industry has failed to appreciate the risks its products create.

Putting common sense back in the driver's seat (Janet C. Daniels Harvard GSAS)

  • The "trolley problem" is a basic thought experiment in ethics and a major question around self-driving cars. There's just one problem: It doesn't actually apply to driving.

The road ahead: Charting the coronavirus pandemic over the next 12 months — and beyond (Andrew Joseph — STAT)

  • As we reach what is almost certainly not yet the halfway mark of the pandemic, an expert article that games out where we'll be a year from now.

AI planners in Minecraft could help machines design better cities (Will Douglas Heaven — MIT Tech Review)

  • Good news: That game your teen is spending all their time on has real-world benefits. For AI, at least.

6. 1 good thing: The new old age is younger

Photo of an older German man competing in the long jump
Guenther Herzberg competing at the Seniors All-Around German Championship, and jumping farther than you whipper-snappers ever will. Photo: Matthias Hangst/Getty Images for Spiegel Magazine

A new study finds that the physical and cognitive ability of older people has improved meaningfully over the past 30 years.

Why it matters: With the population of the elderly set to boom over the next several decades, their health and well-being are more important than ever. New research shows that rapid decline isn't a given.

What's happening: The study, performed by researchers at the University of Jyväskylä in Finland, compared the physical and mental performances of a group of subjects between 75 and 80 years old to a similarly aged group back in the 1990s.

  • The researchers found that muscle strength, walking speed, reactions, verbal fluency and memory were all better now among the elderly subjects they studied.
  • The essential conclusion was that older people today look, act and think younger than their counterparts from nearly 30 years ago.

How it works: The major difference between the two cohorts seemed to be the environments in which they grew up and grew older.

  • More activity over a lifetime meant that today's seniors were stronger and faster, while generally higher levels of education translated into superior cognitive performance.
  • The results suggest increasing life expectancy also means more years of higher-quality life — essentially an extended middle age.

Yes, but: Those extra high-quality years in the middle add up, but they also mean that the last years of life are now more likely to occur in very old age, increasing the need for expensive care.

"Among the aging population, two simultaneous changes are happening: continuation of healthy years to higher ages and an increased number of very old people who need external care."
— Taina Rantanen, University of Jyväskylä

The bottom line: As someone just on the other side of numerical midlife, I'll take it.