Pages

Saturday, March 9, 2019

Review: Bad Blood: Secrets and Lies in a Silicon Valley Startup

Bad Blood: Secrets and Lies in a Silicon Valley Startup Bad Blood: Secrets and Lies in a Silicon Valley Startup by John Carreyrou
My rating: 5 of 5 stars

In 2014, a young and ambitious Oxford university under-grad Jim Watson had a bold idea. Inspired by the story of the success of Deepmind, and its acquisition by Google (now Alphabet), Jim dreamed of using AI to uncover the emotional state of a person using just a sentence long enough to be a tweet. Given the recent advancements, brilliant young Jim decided that the time was ripe to jump on the AI bandwagon and develop this world-changing product. So Jim dropped out of Oxford, and started the journey towards turning this audacious idea into a real product. Jim was incredibly inspired by Andrew Ng, one of the world's foremost experts of Machine learning and deep learning. So inspired, that he started wearing only blue shirts like his hero Andrew Ng, and decided to learn Mandarin as well.

So our charismatic, visionary and glib hero assembles a team of highly qualified people and starts developing the product. But no matter how sexy you make it sound or how brilliantly you talk about it, technology that IS magical cannot be developed magically through the flourish of a wand. It takes an inordinate amount of time, effort and resources, and some luck. Language is inherently complex and dynamic. It keeps changing. Developing an AI that understands human language is the one of the most difficult problems in the world and claiming to assess the emotional state of a person from a sentence long enough to be a tweet is a really tall order. But only idiots who are rooted to the ground and don't dare to dream big will be shaken by that. Visionaries like Watson who posses the ability to project a reality-distortion field can always find a way around any problem.

Given that the product was far from complete, Watson instead hired a small army of human translators, who apparently were only assisting to refine the algorithm, but in reality were doing the actual translation. Under a cloak of secrecy, through open and veiled threats, and a philosophy of "deceive-inveigle-obfuscate", Watson managed not only to keep the truth hidden, but convinced a host of luminaries from different walks of life that the technology existed, worked and was just a few months away from release. And in this fashion, Jim Watson managed to raise hopes, raise funds, raise his status and managed to create a company that was valued at close to ten billion dollars. Until the bubble burst.

The aforementioned Jim Watson is merely a figment of my imagination. If you replace the "tweet-long-sentence" with blood, change the product from emotional-analysis to blood testing, replace Andrew Ng with the late Steve Jobs, and Watson with Holmes, you basically have the plot of Bad Blood.

While reading the book, I kept repeating "I can't believe this really happened" and "How on earth is this even possible" to myself. How can you hoodwink a group of some of the smartest, most astute people in the world, be valued at $ 9 billion dollars and yet have a joke of a product? That too, in an industry which is regulated and which affects the lives of people. We are not talking about some dopey social media application or some gaming app. This product was designed to test BLOOD. Consider the ramifications of the testing going wrong. A false positive would potentially lead to more tests and/or treatment for a condition the patient does not have, in a country notorious for its extremely high health care costs. A false negative, could lead to a misdiagnosed or undiagnosed patient, assuming the caregiver doesn't order the test from another credible source. And who is to be blamed if such a patient dies because of this erroneous diagnosis?

The one thing that struck me the most was the extent to which the esteemed backers of Holmes defended her and could see nothing wrong with her. I fail to understand why. What kind of emotions would cloud your judgement to that extent, especially when the stakes are so high? I am sure in the ancient times, having such sway over people would have been deemed sorcery, and Holmes could have been branded a witch. Good for her, that we don't live in those times. Theranos was a veritable house of cards. Its product was vaporware. The ultimate lesson is that the proof is in the pudding. Any product, big or small, has to work consistently, all the time. In the words of Abraham Lincoln, "You can fool all the people some of the time, and some of the people all the time, but you cannot fool all the people all the time." Kudos to John Carreyrou for publishing the story. Bad Blood would be etched in my memory for a very long time.

View all my reviews

No comments: