spinofflive
josie (1)

ScienceSeptember 15, 2020

The artificial intelligence trying to level Twitter’s toxic playing field

josie (1)

Tech start-up Areto Labs noticed online abuse was stopping women from going into politics – so it did something about it. CEO Lana Cuthbertson and creative technologist Jacqueline Comer talk to The Spinoff about their abuse-fighting bot.

“Imagine you have a job interview and every day, for a month, you have to walk down a dark alley, knowing the worst people in the world are in that alley and they will yell and scream at you,” proposes Aucklander Jacqueline Comer, a creative technologist.

“If you knew that, you wouldn’t apply for the job. And, unfortunately, that’s what women in politics have to put up with.”

Most people in the public eye cop some online criticism, but women get some of the most violent. Men receive their share of abuse, of course, with one study finding the leading subjects of online abuse for men are sexual orientation, religion, and physical or intellectual disability.

However, the gendered nature of much abusive language means the worst of it is generally directed at women.

To shine a light in the alley way, @ParityBOT_NZ was created. It’s an artificial intelligence designed to combat abuse by recognising abusive tweets and posting positive ones to create a parity of sentiment.

Comer and Areto Labs’ co-founder and CEO, Lana Cuthbertson, were intrigued by the idea of running the Canadian-born bot in New Zealand because of the country’s perceived acceptance of women in politics. We have a woman as our prime minister and also as the leader of the opposition – as a theoretically woman-friendly country, is Twitter safer for women here?

To find out, Areto Labs and Comer set the bot up to monitor mentions of women running in our 2020 election. Tweets are scored according to their toxicity, and those that meet a certain threshold, which is currently set at 0.9 or 90% likely to be toxic, are considered abusive by the bot and triggers it to tweet something nice.

The bot is trained on a data set largely from the New York Times comment section. “We’re using the one category of toxicity, but there are other labels like racist, identity attack – there are about 15 of them that are available in the system that we could use to analyse this data,” explains Cuthbertson.

Lana Cuthbertson, CEO of Areto Labs

Areto Labs recognises that Māori, Pacific people, women of colour and gender-diverse women are particularly targeted, and this was an uncomfortable part of training the bot. “One of the things we had to work on in setting up the bot for New Zealand was trying to understand the colloquialisms here,” says Comer. “I had to ask some family members, ‘what are some racist and derogatory terms for different New Zealanders?’ and they had to actually tell me.”

The AI uses natural language processing that looks beyond a tweet’s vocabulary and into how the sentence as a whole is constructed. “There’s sentiment analysis involved,” says Cuthbertson. She knows it’s not perfect. “It’s pretty likely to score a tweet as toxic if it’s got a fairly harsh swear word in it, but sometimes people use swear words in enthusiastic or positive ways.”

“The way people speak here is a bit more passive aggressive, so how do you create a machine learning model that can start to understand that?” adds Comer.

It’s been running for almost a month, and has found about 2% of tweets sent to women running for parliament are toxic enough to trigger the bot. In three weeks, they’ve found 737 tweets that are more than 90% toxic.

“The distribution of toxicity is not even across the candidates,” says Cuthbertson. “Higher-profile women will receive a lot of toxicity, so Jacinda Ardern and Judith Collins are getting lots more than some of the maybe lesser-known candidates, who aren’t receiving any, and then there’s a distribution in between.”

On the higher end of this distribution is Green MP Golriz Ghahraman, who’s received barrages of messages via social media and directly to her emails since she first ran for parliament in 2017.

Ghahraman recently received her first tweet from Parity Bot, she tells The Spinoff. “It actually made me worry, because I was like, ‘oh God, what’s happened, have I met this threshold of abuse?’”

She says she’s better able to tune out online abuse these days, but it still takes its toll. “It’s latent, but you’re still expending energy tuning it out. It’s not nothing.”

She says Parity Bot is one approach to addressing the issue, but she’s also interested in seeing more humans take to the keyboards and combat the trolls. “I think that is needed, not necessarily for me, but for making the whole environment safer. Because people come to my feed to get more information, they want to see Green Party policies.”

Ghahraman says there are quiet days and there are spikes. After the Christchurch attacks, she noticed online abuse of anyone even adjacent to the Muslim community soared. “It went up, exponentially,” she says. “Forums that formed around that still exist, and I think they flare up again. There was a whole backlash group that started to form after there was a perceived sympathy for these communities.”

Marama Davidson and Golriz Ghahraman in Aotea Square (Photo: Todd Henry)

She says the abuse has made her question whether there’s a better way of achieving change than parliamentary politics. “I was getting quite serious death threats, it was like, ‘oh, I’m putting my family through this’. I was actually facing a physical threat.

“It’s much harder for me sometimes to cut through on some of the issues I actually want to talk about, because I have to talk about these things, or I don’t have the energy that day to deal with the repercussions. It’s exactly the effect that it’s meant to have, which is silencing.”

After the Christchurch attacks, a campaign Ghahraman had started the year before to reduce hate speech became difficult to progress because of the abuse directed at her in particular, so she gave it up. “Andrew Little picked that piece of work up, and we decided I wouldn’t talk about it any more, because it was so threatening when I did.

“I particularly can’t talk about it because I’m from one of the communities that hate speech impacts on.”

Cuthbertson said there’s no acceptable level of toxicity. “One toxic tweet is one too many,” she says. “It can be a really traumatising experience, and if you’re a woman putting your name forward in politics, it’s likely, unfortunately.”

Currently, the positive tweets sent by the bot are written by humans and vetted by the Areto Labs team. “There might be a world relatively soon where we’re able to automate that, says Cuthbertson. “But we’d rather craft those to the appropriate context.

“It still feels like it needs human strategy and consideration in that way.”

For her part, Ghahraman believes her work is worth these drawbacks. “I think it’s worth it in the end, and that’s why I’m still here.

“I think one of the things that makes it worthwhile for me is all of the other women, and in particular the women of colour, that I’ve ended up connecting with and who create these spaces for each other.”

To any woman hesitant about entering the dark alley that leads to a career in politics, Ghahraman says there is a space on the other side. “We are here for you.”

Keep going!
Microscopic view of the virus which causes Coronavirus disease. (Image: NIAID-RML)
Microscopic view of the virus which causes Coronavirus disease. (Image: NIAID-RML)

ScienceSeptember 11, 2020

Five things we know about Covid-19 (and five we don’t)

Microscopic view of the virus which causes Coronavirus disease. (Image: NIAID-RML)
Microscopic view of the virus which causes Coronavirus disease. (Image: NIAID-RML)

Futurist and evolutionary biologist Robert Hickson (Science Media Centre) on the known and as-yet-unknown aspects of the virus.

What we know

1. We know where the virus ultimately came from

We know that the virus originally came from bats, and most probably a species of horseshoe bat in South East Asia. However, the spike protein in SARS-CoV-2, which allows the virus to attach to cells and infect them, is more different than those from horseshoe bat sequences. While the sequence of this protein is more similar to one in a coronavirus from pangolins, researchers now think this is because the bat sequence has recombined, rather than pangolin being an intermediate host. However, this has not yet been verified.

Some epidemiologists think that a less pathogenic version of the virus may have been circulating in people from Vietnam, Laos and/or Myanmar before it appeared in Wuhan, China. This too hasn’t been proven yet.

Thousands of genomes of the virus have now been sequenced. Analysis of these has helped to monitor how the virus is evolving (not much), and how the pandemic has spread around the world and within communities. These are illustrated in interactive graphics on the NextStrain website.

2. We know a lot about how the virus spreads

SARS-CoV-2 appears to be slightly more infectious than the flu, but less easily spread than the two other recent coronaviruses that came from animals (SARS and MERS).

It is mainly spread by droplets and aerosols

The virus infects the respiratory system, so it is spread mainly by coughing, sneezing and talking. Singing can also spread infections. The virus can survive on surfaces for a few days, but the frequency of this route of transmission is unclear.

Closed spaces, crowded spaces and close contact favour infection

The virus is known to be most easily spread indoors where people are in close contact. Most cases of transmission appear related to infections within households, and in healthcare facilities where there are poor infection control measures. Most clusters of transmission outside of households are associated with “3Cs”. closed spaces with poor ventilation, crowded places, and close-contact settings. Examples include churches, bars and cruise ships.

A crowd, but distanced. (Photo: Getty Images)

Symptoms take about a week to develop, but people can spread the virus before then

Symptoms typically appear about one week after infection and last generally for another two weeks. The virus appears to be spread mostly from 2-3 days before symptoms are noticed and up to one week after symptoms appear. A review of many studies estimated that 48% to 62% of transmission may occur before people show symptoms.

Some people may not show symptoms, but can still spread the virus

Some people do not show symptoms after infection (or don’t notice mild symptoms), but can still pass on the virus. It is unclear how many people fall into this “asymptomatic” category since it varies by setting, and depends on having good testing. Estimates vary from 6% to 81% (on a cruise ship), with the WHO estimating a general asymptomatic level of 16%, the Centres for Disease Control and Prevention estimating 40% in the US, and another study considering 40-45% may show no symptoms. More systematic testing makes the asymptomatic infection estimates more accurate.

3. We know a lot about how the virus affects the body, but are still learning too

The virus is different from the flu and SARS due to the broad range of symptoms and conditions that it can cause.

An international review found that the most common symptoms in those who go to hospital with Covid-19 are fever (70%-90%), dry cough (60%-86%), shortness of breath (53%-80%), fatigue (38%), muscle pains (15%-44%), nausea or diarrhea (15%-39%), and general weakness (25%). Loss of smell may be the only symptom in about 3% of infected individuals.

While the virus commonly causes respiratory complications, it can also lead to a range of other injuries and diseases. It is less certain whether the virus directly causes all of these impacts, or whether they may be due to an overstimulated immune response. An unbalanced immune response is associated with more severe complications.

About 5% of patients with severe symptoms can require intensive care, and more than 75% of hospitalised patients have needed oxygen.

Those without symptoms may also still show tissue damage.

4. We know who is most at risk

Those most at risk of developing severe conditions, or dying, are men, older people, those who are obese or overweight, those who are socially deprived, and those from non-white ethnic groups (in Western countries).

Patients older than 80 can be at least 20 times more likely to die from Covid-19 than those in their 50s. The elderly may be at more risk due to a variety of factors, such as weaker immune responses, but these factors are still being studied. The socially deprived are usually less likely to get medical help earlier and may have lower-quality health care.

However, young people and those without pre-existing health conditions can sometimes become severely ill and can die from Covid-19.

5. We know some effective treatments and interventions 

Regular handwashing and avoiding close contact with others are known to reduce risks. Wearing well-fitted masks also helps reduce the chance of spreading infection, though by how much is less certain. Avoiding large gatherings and confined spaces, particularly where there is singing or shouting, is also known to be effective.

Lying patients on their stomachs, providing extra oxygen, and using ventilators are effective treatments for those who have difficulty breathing due to Covid-19. Hospitals are improving their care as they treat more people, so patient outcomes are improving.

The antiviral drug remdesivir has shown promise in treating patients with lung infections if it is used in combination with other drugs. One study has shown that it shortens recovery time. However, so far it hasn’t been shown to reduce deaths.

The steroid dexamethasone, and some other corticosteroids, have been shown to reduce deaths amongst those with severe disease, such as those on ventilators. The effects of these steroids on people with less severe conditions is still being investigated.

What we still don’t know

1. Why do people respond differently?

While we know some factors that influence the risk of becoming very sick, we often do not know why that is. For example, why do older men appear to be at more risk than older women? It is also unknown why some people don’t show any symptoms, although this is seen for other coronavirus infections too. Some infected people, called “long haulers”, take months to recover, and it is not known why. It is likely that a combination of factors, such as genetics, behaviours, and the individuality of people’s immune systems, play a role.

Health workers take part in ICU training for Covid-19 at Hutt Hospital (Photo: RNZ / Dom Thomas)

2. Do people develop immunity, and how long could it last?

People do develop antibodies and cellular immune responses after being infected. However, we don’t yet know if this prevents further infections, or reduces the severity of the infection. Studies have shown that people can get re-infected with other coronaviruses within a year, and cases of re-infection of SARS-CoV-2 are emerging, so there may not be any long-lasting immunity to Covid-19.

3. How well will a vaccine work?

There are over 300 possible vaccines being tested. Some vaccines have begun phase three clinical trials, which will indicate how effective they may be. Results are likely to start becoming available before the end of the year. Vaccines may not provide long term immunity. They could moderate symptoms rather than prevent infection.

However, combined with other treatments even an imperfect vaccine can help slow or eliminate the spread and reduce the impacts of the virus.

Producing and distributing vaccines is complicated and takes time, so even if/after vaccines are approved most people may have to wait months (if not longer) before they can be vaccinated.

4. Will Covid-19 become more, or less, dangerous?

One research paper suggested that the virus was mutating to become more infectious. However, this hasn’t been proven. Neither is there evidence to show that the virus is causing more severe disease as it spreads, or is becoming less infectious.

Viruses do mutate, but the mutation rate for SARS-CoV-2 – about two mutations a month, or every second or third transmission – is similar to other RNA viruses like influenza.

5. What will the next pandemic be?

There are thousands of viruses in animals with the potential to infect people, but we are not studying them systematically to identify the main risks.

But we do know what needs to be done to reduce the likelihood of further pandemics.

  • Reduce habitat destruction that increases the risk of contact between humans and wild animals,
  • Eliminate markets for wild animals,
  • Develop epidemic and pandemic plans (and test them),
  • Ensure there are good stockpiles of personal protective equipment,
  • Improve surveillance for new diseases, and
  • Respond quickly to outbreaks.
  • International scientific and political collaboration and coordination are also essential. The Covid-19 pandemic has shown that good science and good governance are essential to prevent and fight pandemics.
But wait there's more!