As developers unlock new AI tools, the risk for perpetuating harmful biases becomes increasingly high — especially on the heels of a year like 2020, which reimagined many of our social and cultural norms upon which AI algorithms have long been trained.

A handful of foundational models are emerging that rely upon a magnitude of training data that makes them inherently powerful, but it’s not without risk of harmful biases — and we need to collectively acknowledge that fact.

Recognition in itself is easy. Understanding is much harder, as is mitigation against future risks. Which is to say that we must first take steps to ensure that we understand the roots of these biases in an effort to better understand the risks involved with developing AI models.

Circa 2021

A crime is a deliberate act that can cause physical or psychological harm, as well as property damage or loss, and can lead to punishment by a state or other authority according to the severity of the crime. The number and forms of criminal activities are increasing at an alarming rate, forcing agencies to develop efficient methods to take preventive measures. In the current scenario of rapidly increasing crime, traditional crime-solving techniques are unable to deliver results, being slow paced and less efficient. Thus, if we can come up with ways to predict crime, in detail, before it occurs, or come up with a “machine” that can assist police officers, it would lift the burden of police and help in preventing crimes. To achieve this, we suggest including machine learning (ML) and computer vision algorithms and techniques.

Learning how to learn is something most humans do well, by leveraging previous experiences to inform the learning processes for new tasks. Endowing AI systems with such abilities however remains challenging, as it requires the machine learners to learn update rules, which typically have been manually tuned for each task.

The field of meta-learning studies how to enable machine learners to learn how to learn, and is a critical research area for improving the efficiency of AI agents. One of the approaches is for learners to learn an update rule by applying it on previous steps and then evaluating the corresponding performance.

To fully unlock the potential of meta-learning, it is necessary to overcome both the meta-optimization problem and myopic meta objectives. To tackle these issues, a research team from DeepMind has proposed an algorithm designed to enable meta-learners to teach themselves.