AsianScientist (Apr. 23, 2021) – From viral filters to suspiciously specific—though convenient—search recommendations, artificial intelligence (AI) underpins just about every action we take on social media today. Similar to other successful modern technologies like smartphones and the internet, AI has become ubiquitous—almost to the point of invisibility.
For tech behemoths like Facebook and Google, AI provides an unparalleled opportunity to generate insights from massive amounts of data at scale. However, beyond the buzz of Silicon Valley, other organizations worldwide—both large and small—continue to struggle with implementing machine learning technology.
“We knew that the opportunity for machine learning adoption outside the big tech companies was monumental,” shared Liu Feng-Yuan, CEO and co-founder of Singapore-headquartered BasisAI, which builds bespoke AI solutions for enterprises. “Many organizations want to be more data-driven and automated in their approaches, but face a whole host of challenges in getting machine learning to work for their businesses.”
Some of these issues arise from the sheer diversity of applications that AI can be applied to. For instance, according to Liu, data scientists and software engineers can have vastly different methods and mental models of the best machine learning approach to a given problem. Still, the real challenge comes into play after—putting the resulting machine learning models into practice is easier said than done.
A lack of infrastructure, especially for smaller companies, can thwart the ability to track the performance of machine learning models after deployment. Just like a ‘black box,’ companies can see the initial data and final decision, but are unable to visualize the processes in between. More ominously, harmful biases can unknowingly creep into AI systems—as is often the case when organizations take the plunge into AI without fully understanding the technology.
As AI automates life-changing decisions like hiring and even criminal sentencing, such biased ‘black boxes’ begin to raise major concerns. Because AI systems are typically trained using existing data, they can inherit existing societal prejudices.
“Decisions have an impact on the real world and usually. decisions are made by human beings with agency,” explained Liu. “Now the moment AI software makes decisions, there arise questions about how these decisions are made and whether they are made in an ethical way.”
Joining forces with seasoned Silicon Valley veterans (and brothers) Linus and Silvanus Lee, Liu sought to address all these seemingly disparate challenges through one solution: BasisAI’s proprietary machine learning platform, Bedrock. As its name implies, Bedrock gives enterprises a foundation to deploy AI for decision-making successfully—and responsibly.
In other words, Bedrock enables users to finally peer inside the black box of AI. With the power to experiment and explain the models used, organizations can reduce biases and scale AI solutions to match their specific needs. While BasisAI has unsurprisingly courted customers across financial services, government, tech and insurance spaces, Bedrock’s applications extend even to creating bespoke AI engines for animal conservation.
“We have come a really long way in just two years—from securing S$8 million in seed funding from Temasek and Sequoia at the start of 2019 to launching Bedrock in July 2020,” said Liu. “It’s time to look at expanding the business, building out the team, our product and our commitment to responsible AI for enterprises in the region and beyond.”
With BasisAI leading the charge for accessible and responsible AI, Liu hopes to realize a future where important decisions are made with facts and data rather than steered by biases and emotions.
“We need to choose a brighter future. One in which we harness the power of AI and machine learning to deliver outcomes that make the world a better place for everyone,” concluded Liu.
Copyright: Asian Scientist Magazine.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.