Inside AI - January 8th, 2021

This could counteract the lack of alternative funding for smaller resource-constrained research labs, which can't compete with larger tech giants like Google that "have disproportionate control over the direction of AI research." Refocus on common-sense understanding in AI, so comprehension takes a priority over prediction. This could help combat social problems, such as algorithmic discrimination and deep learning adversarial attacks, as well as result in more "technically robust systems." Empower marginalized researchers, so more value is given to the ways AI influences society and people who are disempowered. A lack of diversity in the field, for example, leads to biased algorithms, which has been highlighted by researchers such as former Google expert Timnit Gebru.