Social AI, Environment, Telecommuting, and WomenEscrito por A.Freire L.Gonzalez E.Oliveras H.Ramalhinho el 29/04/2020 a las 10:21:38
By Helena Ramalhinho, Ana Freire, Libertad Gonzalez and Ester Oliveras (Top bottom - left right)
Why, you may wonder, a title like this with four such different topics? Because after this terrible pandemic, there are at least four topics that we should pay much more attention to in order to create a better society. These four are, in our opinion, fundamental to get something good out of this crisis.
One of the main consequences of this general pandemic has been the movement to the digital world: telecommuting, online grocery shopping, video conferences with our relatives and close friends… even our leisure has become digital. Most of us have significantly increased the use of, not only technology, but also Artificial Intelligence (AI). This mysterious tool has seeped in through our lives more than ever: when being suggested the next most appropriate routine in our home workout, when choosing a product recommended by our favorite online shop, and even when changing our background during a call with our friends.
Artificial Intelligence is key in helping to fight this pandemic. AI is being used for forecasting how coronavirus will spread around the globe, for spotting signs of COVID-19 in medical images, or for finding specific drugs for an effective treatment.
This is the AI that we want and need: social AI. We should drive this powerful tool to shedding light on new challenges addressed to improve human and planetary wellbeing. It’s the time for governments and the AI industry to regulate the boundaries of Artificial Intelligence, emphasizing the following aspects:
- The effects on people’s lives: Algorithmic fairness and transparency are highly required in order to avoid systems with a negative impact on people’s lives, especially nowadays when several approaches have arisen for monitoring people’s activity in order to control the pandemic.
- Bias towards minorities: AI might suffer from biased datasets that have been historically dominated by male and white patients. For instance, some dermatological diseases cannot be identified in dark-skinned people if systems are trained on datasets mainly composed of white skin patients. Many diseases, including COVID-19, might have different symptoms and effects depending on previous diseases, gender, age, or ethnic group. Collecting representative datasets for training these systems is key in helping personalized medicine. Also, building more diverse teams in the AI industry can positively impact the design process of AI tools, leading to less discrimination. This implies adapting the hiring processes in the AI industry to include candidates from underrepresented groups.
- The AI industry effect on climate change: The unprecedented use of AI applications, demands an increase in the computational resources used, turned into huge physical infrastructures consuming a lot of energy. One study from the University of Massachusetts Amherst estimated that the cost of training BERT (a language representation model proposed by Google for improving search) on GPU is roughly equivalent to a trans-American flight.
These and more claims have been collected in the 2019 report from the AI Now Institute (New York University).
There is a certain interdependence between the climate emergency, the sanitary crisis, and the ensuing economic crisis.
Right now, our biosphere is one of the few planetary stakeholders benefiting from the COVID-19 crisis. The levels of CO2 emissions into the atmosphere have dropped due to the sharp decrease in mobility, and air quality has experienced a radical improvement. Since the COVID-19 mortality rate intensifies in more contaminated cities, the harsh confinement that it brings about will help reduce the mortality due to both COVID-19 and pollution (an estimated 7 million people every year).
On the downside, the sanitary crisis is causing delays or even outright cancellations in climate emergency actions. The Glasgow Climate Change Conference, for instance, has been postponed to 2021. Important budget allocations, such as the EU Green Deal, have been halted, with the growing possibility that its original targets might be watered down due to disagreements between state members on how to fund the current sanitary crisis. The Trump administration has decided to freeze any decrease in greenhouse emissions so as to favor the short-term viability of oil companies, and of the manufacturing sector at large. The looming economic crisis is forcing the private sector to focus on their survival rather than accelerating investments to reduce carbon footprints. And further delays on taking decisive actions on climate will bring upon society further sanitary crises.
The economic recovery should be aligned with the best environmental practices: a boost on circular economy, on products with an optimized lifecycle, on sustainable and clean energy, and on zero-emissions based mobility solutions.
The labor market
A third important issue regarding the longer-term impacts of the current epidemic has to do with the functioning of the labor market. The need to stay home for non-essential workers has forced many firms to allow their employees to work from home and under a more flexible schedule. This change may prove to be more than temporary. As firms and workers learn about the new ways to work long-distance, this may lead to a long-term increase in the prevalence of flexible work arrangements. Since women have in general stronger preferences for flexible schedules, a potential unintended consequence of the current crisis may be to make the labor market more women-friendly.
And since we’re talking about women… The Forbes article “What Do Countries with The Best Coronavirus Responses Have in Common? Women Leaders” explains very well the positive impact of having a woman leader on some countries that had and are having a very good response to the COVID-19 crisis… is this a coincidence? The ones that believe so, must get better informed on the way women lead and manage a department, company and country! Several studies confirm that companies that have more women are doing better. Another article from Forbes says “Why Women-Led Companies Are Better For Employees”, or this MIT Sloan article that mentions “The good news is that a growing body of research shows that increasing the number of women leaders can be key to your company’s future success.”, just to cite two examples. Most of the personnel that are on the front-line fighting the COVID-19 are women, however most of the decisions are made by men: “Women comprise the majority of frontline healthcare workers globally, meaning that female representation is vital in tackling the coronavirus crisis. 70% of the world's healthcare staff are made up of women, but only 25% of global leaders are female.”
In the research and university sector, where we work, we need to have a deep rethinking about the role of women. We only represent a small percentage of the full professors and the women are frequently underrepresented in committees, awards, leading positions, etc. Relevant studies mention that at the university, women are frequently looked down or not respected. Their professionality is questioned frequently and they are assigned volunteer and hardworking tasks but without any recognition or with strategic impact. Many women can tell a lot (too many!!!) of examples where our work was put in question. But, are we sure we want to continue in this line when we observe all over the positive leadership of women, in politics, in health, in research, etc.?
Citing Secretary-General António Guterres: “Gender equality and women’s rights are essential to getting through this pandemic together, to recovering faster, and to building a better future for everyone.”
"We once heard from a woman scientist that she was sharply criticized as “aggressive” when she brought up a flaw in a male colleague’s analysis; after that she felt she needed to just “bring in baked goods and be agreeable.”” "...displays of confidence and directness decrease women’s influence but increase men’s”
A successful recovery from the forthcoming economic and social crisis should take into account these four areas in an interlinked approach: technology, in the widest sense: artificial intelligence and more flexible approaches to work, together with a heightened awareness of the natural limits of our planet. A classical patriarchal view of technology, economics or leadership is no longer valid. A female vision must be incorporated in order to create a sustainable future.
La inteligencia artificial, clave entre las 50 compañías más innovadoras
La IA se impondrá en 2020 en los programas formativos empresariales