Amazon, Microsoft, and IBM are under intense scrutiny to stop using gender labels like ‘man’ and ‘woman’ for their Artificial Intelligence services. This comes right after the world’s leading search engine Google announced that it was backing down from using gender classification tags.
Google decided to start instead of tagging people with their images while using neutral terms like ‘person.’ Hence, this prompted Amazon, IBM, and Microsoft to stop defining gender labels from peoples’ images using their platform.
As you may know, all the companies mentioned above have comprehensive AI tools capable of classifying people based on their images. Google made its landmark decision to stop using their AI image classification services because it’s impossible to determine someone’s gender by merely looking at their photograph (s).
However, this move was prompted by numerous researchers on AI who are putting more pressure on Amazon, IBM, and Microsoft to join the direction Google took. The artificial intelligence services under scrutiny by these researchers include IBM’s Watson, Microsoft’s Azure, and Amazon’s Rekognition.
According to Business Insider, research done by Joy Buolamwini – a computer scientist showed that AI tools from both IBM and Microsoft had a higher chance of wrongly classifying someone’s gender based on their complexion.
The computer scientist at MIT suggested that it would be best if firms reexamined the identity labels being used to avoid biasing. If the move is successful, tags such as class, race, or disability status would be dropped at an instant.
Why Did Google Drop their AI Classification Services?
Although some people felt that Google’s move was a result of political influence, there is an essence in their decision. Most people don’t understand the dangers of nonconsensual insinuation of someone’s gender based on their appearance only.
While most companies refrained from commenting on their AI’s classification weaknesses, Amazon referred to their guidelines, suggesting that the Rekognition tool predicts the gender only based on the physical appearance of the face in a particular image.
Hence, it doesn’t determine or define someone’s gender identity, thereby shouldn’t be put to the test for such identification.
All four companies benefited a lot from determining the kind of person engaged in their platform. This is because both the male and female have different motives when using the named platforms. With the knowledge of gender, such companies can work on different optimization practices on their platforms to attract different genders.
Therefore, the decision to stop such artificial intelligence services to classify the genders logged in would limit the companies’ data intake. But because they are under pressure, coming up with an alternative would be an obvious consideration.
This would be a safe and effective way for most companies to avoid falling under scrutiny when determining their clients’ gender.
Tech’s Approach to Fixing Its Gender Inequality
Gender inequality is a problem in the tech sector, where code is often written and products conceived by gender imbalanced teams. Will software that has been designed solely by male computer programmers be the best possible product it could be? Equality isn’t just about helping women, it’s about being better businesspeople.
Gender APIs get much attention recently and have been put to many kinds of research, including studies of the gender gap in many aspects of society. Being able to quickly understand the gender of individuals or the authors of scientific papers, Gender APIs help illustrate the skewed representation in these industries. For example, Seattle’s Allen Institute for Artificial Intelligence analyzed research papers from the academic search engine by using a Gender API and predicted that only by the 2137 year female authorship will hit the 45 percent level.
Why is it so hard to achieve gender parity in the tech field? Technology organizations need to highlight how technology is a force for good if they want to attract more females to the sector. Half of the females say that feeling like the work they do makes the world a better place is the most important factor when deciding their future career.