Home › Nugget
4 - Diversity of thought in the drawing room - Kartik Hosanagar a Human’s guide to Machine Intelligence
Kartik speaks about how we need to be mindful of the risk of being exposed to a certain type of content or viewpoint as the algorithms are solving for engagement and are likely to show content that we are likely to gravitate towards. He speaks about some of the mechanisms he employs to ensure that he builds diversity of thought in his head as he goes about processing the world around him.
More from same Guest
Kartik speaks about the extent to which machines and algorithms have pervaded our lives. To give an example, 80% of view on Netflix is based on algorithmic recommendations and 70% of Youtube consumption is based on what it suggests. He talks about what this means for human beings to stay relevant in the future where the machines are getting exponentially smarter by the day.
Kartik speaks about how AI has moved from being expert systems (where humans input a certain set of rules that machines follow) to machine learning systems (where human expose the machine to tonnes of data with the relevant input and output parameters) and how that leads to situations where the machines often come up with actions that are beyond our comprehension. He also takes the example of US Constitution and the Code of Hammurabi to make the distinction between the two types of systems and the trade-offs therein.
Kartik speaks about how different FAANG Companies (Facebook, Amazon, Apple, Netflix and Google) are trying to inject hose pipes into our lives and capture data for their respective algorithms to get smarter over time. He speaks about how, as consumers we need to walk the tightrope between leveraging the benefits of these platforms while protecting our privacy while doing it.
Kartik speaks about how we need to be mindful of the risk of being exposed to a certain type of content or viewpoint as the algorithms are solving for engagement and are likely to show content that we are likely to gravitate towards. He speaks about some of the mechanisms he employs to ensure that he builds diversity of thought in his head as he goes about processing the world around him.
Kartik speaks about how, if we are not watchful, algorithms might end up creating outcomes that we hadn’t really pictured when we started using them. He speaks about one particular example of an instance with Amazon where using algorithms to screen resumes led to the gender bias being further amplified before Amazon noticed it and addressed it.
Kartik uses the example of music (Pandora, Last.FM and Spotify) and speaks about the different approaches to algorithm design and the implications on the kind of content we are likely to see as consumers. He also speaks about how the designers of algorithms need to have a holistic approach to developing metrics to evaluate the efficacy of the algorithms.
Kartik speaks about how we should think about using algorithms for decision making versus decisions support. He urges to think about machines as augmenting and not substituting human capability. He speaks about how we should consider the extent of consequences and social implications to think about how we leverage the power of the machines.
Kartik speaks about the impact of AI on jobs of the future. He cautions that it is not just the menial blue collar type of jobs that are at risk but a wider array of jobs where machines could replace man. He goes on to talk about the implication for us and how we should think about staying relevant in the future.
Kartik speaks about how we all could be thoughtful about equipping ourselves with some basic level of literacy around AI. Even if you are not in a technology-led company, it is likely that you will be impacted by AI in some shape or form as a leader, as a consumer or some other form.