Technology executive Veronica Wu looks at machine learning and the process of Silicon Valley venture capital investing, a traditionally a “long shot” method, with an eye to improving win percentage. In a McKinsey question and answer session as part of their June quarterly report, Wu reveals how they use technology to build better predictive models for venture capital investing that integrate with humans. Like much of the gloss coming out of Silicon Valley regarding “machine learning,” however, the questions were light in key areas. This includes pointing to exactly where a machine scans information it wasn’t programmed to scan, “learn” knowledge or gain insight based on entirely new patterns not related to an if-then formula, and then develop an investment method that didn’t follow any patterns from past strategies? Wu’s use of computers to create models that assist in decision making are tangible nonetheless, even if it is an accomplished human more responsible for the success of the machine.
Read more on Value Walk
