March’s WE ARE AI is in full swing! Not only have we launched a new contest to crown a new rising star in AI, but we’re also rolling on with our series of Q&As with the leading voices, disruptors, and decisionmakers in AI.
This time, we spoke with the team at Layr, an insurtech startup building better insurance. Layr uses artificial intelligence to recommend and match companies with the insurance policies and coverage they need and accurately predict carrier pricing. Business owners complete an application completely online and instantly see a tailored quote that is paid monthly on a credit card.
Layr was recently named to Atlanta Inno’s 20 Startups to Watch in 2020 list and the Technology Association of Georgia’s Top 10 Innovative Technology Companies 2020 list. Head of Engineering Brice Hartmann and Machine Learning Engineer Matt Munns teamed up to answer our questions.
AI is often talked about as changing the world – for better or worse. What are your views on that?
AI is a tool and like other tools must be wielded thoughtfully. Advancements in AI and ML have been woven into our lives in many positive ways, and there is enormous potential for this positive impact to continue. However, the increased interaction and dependence on AI and AI-powered applications creates a responsibility for researchers and practitioners to address moral and ethical issues that arise.
What are the most meaningful ways AI can have an impact?
AI and AI-powered applications have an unprecedented opportunity to scale and the benefit of this efficiency can be passed to consumers. We already see this through applications like Siri, which offer every iPhone user a personalized voice-activated assistant. Similarly, many eCommerce websites apply personalized recommendations to every user, helping us more easily find products that we’re looking for. More critically, AI has been applied to medical problems. Machine learning models can identify pneumonia from x-ray images quicker and more accurately than doctors.
AI is poised to expand on these positive impacts, through improvements in existing applications as well as expansions to new industries. Expansions of AI into finance, healthcare, manufacturing and insurance offer the possibility of increased efficiency and greater simplicity for end-users in industries previously fraught with time-intensive and error-prone work.
What are the most important issues we can’t ignore?
As humans become increasingly reliant on AI, ethical concerns must be taken seriously. Fairness and bias are among the most pressing ethical concerns as AI expands to new industries.
For example, fairness in applications involving automated resume screening in HR applications could perpetuate systemic bias present in training data. Algorithms might screen out applicants only on the basis that their backgrounds are not similar to what has been previously identified as “successful,” regardless of merit.
Also, as AI expands, we must ensure that the goals of algorithms align with human goals. For example, an application designed to keep users on a certain media site must not ignore the human cost of this metric, such as the dissemination of misleading information or the potential for addiction.
What do we have wrong about AI, and how can we address these myths?
AI isn’t magic – every output or prediction made by AI has a logical reason that is attributable to the inputs and the designed goal of the algorithm. Keeping this in mind will help address a lot of the issues mentioned above and will improve outcomes. We can do this by raising the bar for technological literacy, as well as being honest about drawbacks and limitations of existing or newly developed methods in AI.
How does Layr help push AI forward?
Layr is expanding the application of AI and machine learning to the insurance industry. The potential impact of AI isn’t realized unless it is applied to a real-world problem and Layr is making this impact by blending machine learning and insurance industry knowledge to improve customer experience. AI at Layr leads to a simpler customer experience by replicating insurance expertise in an instant, avoiding time-consuming and confusing elements of the insurance buying process.