Salesforce creates modules for its Trailhead Artificial Intelligence Developer Education Platform (AI), reports ZDNet. That way, more people should be able to build algorithms, and then again in a responsible way.

The Trailhead modules, according to Kathy Baxter, architect of ethical AI practice at Salesforce, are designed for employees, customers and partners who work with Einstein and use various models. Salesforce believes that with low code developer tools, more technology and business professionals can and will build algorithms. But then they have to understand how to do it responsibly.

There has been a lot of talk lately about AI ethics, prejudices and transparency. A major concern is that individual algorithms do not have prejudices, but they do when combined with other models.

“As AI becomes more accessible and the impact begins to become clear, we are faced with new issues about how to ensure that AI’s future is responsible and fair – even when built by people without technical training or AI expertise,” said Baxter.

Responsible Creation of Artificial Intelligence

According to Baxter, the first step in building responsible AI solutions is to understand that prejudices may be in the models and training data. Trailhead therefore receives several additions aimed at AI ethics and preventing prejudices.

For now, however, it is one called Responsible Creation of Artificial Intelligence. This module contains information about ethical and human use of technology, understanding AI and recognizing prejudices in AI. Attention is also paid to the removal of exclusions from data and algorithms.

The Trailhead modules may also be able to provide information on which future functions need to be built in Einstein. Einstein already has a number of tools to prevent AI prejudices, which are built into various pilots. “We’re trying to help customers understand what’s in a model, explain what’s being done, and communicate how it works,” says Baxter.

This news article was automatically translated from Dutch to give Techzine.eu a head start. All news articles after September 1, 2019 are written in native English and NOT translated. All our background stories are written in native English as well. For more information read our launch article.