The research was published by the Royal Statistical Society but conducted by Glass AI, a startup which uses artificial intelligence to analyse every UK website.

In a blog post, the company explained its unique approach:

“Previous related studies created for economists, policy-makers, or business analysts have tended to underuse or even ignore the web as a data source, typically only looking in any detail at a limited number of sectors of the economy, examining a small slice of geography or conducting manual (and expensive) surveys.

Worse, given a small data set, data scientists have no choice but to extrapolate and rely on small sample statistics.“

Across the entire ‘.uk’ domain, Glass AI read the genders of 2.3 million people and the positions they held in 150,000 organisations spanning 108 industry sectors.

Gender gaps were found across industries with men far more likely to be in leadership positions.

82 percent of all CEOs, 92 percent of chairpersons, and 73 percent of directors are male.

Meanwhile, support roles are dominated by women. 95 percent of receptionists, legal secretaries, and care assistants are female.

Both of the core genders are relatively equal when it comes to participating in the workforce, but the gap presents itself in the roles they attain.

Glass AI’s findings matched those of the Office for National Statistics which shows promise for the use of artificial intelligence to conduct research far quicker and with fewer resources that matches or exceeds traditional means.

deepgeniusai.com/">AI & Big Data Expo events with upcoming shows in Silicon Valley, London, and Amsterdam to learn more. Co-located with the IoT Tech Expo, , & .

The post AI-conducted study highlights ‘massive gender bias’ in the UK appeared first on AI News.

Gender gaps in STEM careers have been a problem for some time, but it’s not often the end product matters what gender it was developed by. AI is about to be everywhere, and it matters that it’s representative of those it serves.

In a report published this week, the WEF wrote:

“The equal contribution of women and men in this process of deep economic and societal transformation is critical.

More than ever, societies cannot afford to lose out on the skills, ideas and perspectives of half of humanity to realize the promise of a more prosperous and humancentric future that well-governed innovation and technology can bring.”

Shockingly, the WEF report found less than one-fourth of roles in the industry are being filled by women. To put that in perspective, the AI gender gap is around three times larger than other industry talent pools.

“It is absolutely crucial that those people who create AI are representative of the population as a whole,” said Kay Firth-Butterfield, WEF’s head of artificial intelligence and machine learning.

Bias in coding has the potential for AI to perform better for certain groups of society than others, potentially giving them an advantage. This bias is rarely intentional but has already found its way into AI developments.

A recent test of Amazon’s facial recognition technology by the ACLU (American Civil Liberties Union) found it erroneously labelled those with darker skin colours as criminals more often.

Similarly, a 2010 study by researchers at NIST and the University of Texas in Dallas found that algorithms designed and tested in East Asia are better at recognising East Asians, while those designed in Western countries are more accurate at detecting Caucasians.

More recently, Google released a predictive text feature within Gmail where the algorithm made biased assumptions referring to a nurse with female pronouns.

It’s clear, addressing the gender gap is more pressing than ever.

You can find the full report here.

 AI & >.

The post AI is at risk of bias due to serious gender gap problem appeared first on AI News.

Amnesty International and Access Now prepared the ‘Toronto Declaration’ (PDF) that’s also been signed by Human Rights Watch and the Wikimedia Foundation. As an open declaration; other companies, governments, and organisations are being called on to add their endorsement.

In a post, Access Now wrote:

“As machine learning systems advance in capability and increase in use, we must examine the positive and negative implications of these technologies.

We acknowledge the potential for these technologies to be used for good and to promote human rights, but also the potential to intentionally or inadvertently discriminate against individuals or groups of people.

We must keep our focus on how these technologies will affect individual human beings and human rights. In a world of machine learning systems, who will bear accountability for harming human rights?”

Ethics have become a major talking point in the AI industry. However, much of the conversation so far has focused on drawing red lines when it comes to surveillance and military applications.

There’s a big debate over AIs potential impact to jobs. Some believe automation will cause a work shortage, while others argue that most will simply be enhanced by AI.

If jobs are being replaced, ideas like a universal income will have to be re-examined. If jobs are being enhanced, ensuring AI is indiscriminate will be even more important.

AI has already shown discrimination

Technologies developed and used in the West are typically developed by white males.

Research has been performed into the gender and race gap of executives in Silicon Valley. This data at least provides some indication of the representation problem:

What this means is that, unintentionally, products often perform better for this particular group. Today, that could just mean something relatively trivial like Siri recognising an American male voice with greater accuracy (even as a British male, Silicon Valley-developed products often struggle with my accent!)

A 2010 study by researchers at NIST and the University of Texas in Dallas found that algorithms designed and tested in East Asia are better at recognising East Asians, while those designed in western countries are more accurate at detecting Caucasians.

However, if jobs are becoming more reliant on AI, they need to work as well for everyone who uses them. Failing to do so will put certain groups at a greater advantage than others.

“From policing, to welfare systems, online discourse, and healthcare – to name a few examples – systems employing machine learning technologies can vastly and rapidly change or reinforce power structures or inequalities on an unprecedented scale and with significant harm to human rights,” wrote Access Now.

Policing is one area of particular concern. An investigative report by ProPublica revealed that computer-generated ‘risk assessment scores’ used to determine eligibility for parole are almost twice as likely to label black defendants as potential repeat offenders, despite evidence to the contrary.

Similarly, a 2012 study (paywall) by the IEEE  found that police surveillance cameras using facial recognition to identify suspected criminals are five to 10 percent less accurate when identifying African Americans – which could lead to more innocent black people being arrested.

Machine learning models for AI are often trained on public data and therefore we must be careful about what sources are used. Microsoft’s attempt to create a chatbot which learns from the public, Tay, infamously ended up becoming a rather unsavoury character spouting racist and sexist remarks.

The declaration signed today is a great start to keep these issues in mind as AI technologies are being developed, but it will require tackling inequalities across the whole of society to make developments truly representative of those it serves.

What are your thoughts on the AI discrimination issue?

The post Editorial: Stopping AI’s discrimination will be difficult, but vital appeared first on AI News.