Noise , Judgment , AI & Big Data
“Wherever there is a judgment there is noise. Some judgments are biased ; they are systematically off target. Other judgments are noisy, people who are expected to agree end up at very different point around the target. Many organizations, unfortunately , are afflicted by both bias and noise” (D. Kahneman).
In Kahneman’s book ‘Noise’ error in judgments or decisions are categorized under those two notions. They suggest a technique called decision hygiene to reduce noise and prevent errors. The approach is created for humans to have more reliable decisions at life. Its basically a model to decrease error in human decision making process.
Recently, AI became very popular topic, although its there for decades. A question which is arised is there noise and bias also in AI algorithms? If answer is yes, will AI need to create a decision hygiene as well? Or will life with AI be much more fairer than current situation?
There is a basic rule differs a robot than human which is consciousness. According to AI experts; AI or a robot has intelligence but does not has consciousness. For example, a chess robot cannot understand whether it wins or loses at the end of the game. It just calculates next move until the end. Since the world started to discuss AI ethics and also its capabilities , those aspects also worth to be considered.
Thomas Siebel, in his book about Digital Transformation mentions four main pillars to achieve digital transformation: IoT , Cloud Computing, AI and Big Data. And he believes that current digital transformation activities are still replication of existing processes. We are not sure about what will come next as part of current era. According to Siebel, the first Industrial Revolution , allowed humans to master mechanical power, in the last one we harnessed electronic power. In the era of digital transformation , we will master mental power. My understanding from this statement is that as if 3 other pillars are established for helping AI to operate more effectively.
The types of problems tackled by AI traditionally included natural language processing and translation, image and pattern recognition (for example, fraud detection, predicting failure, or predicting risk of chronic disease etc.) . AI expert Stuart Russell puts this type of activities under narrow AI category. Even if you ask popular AI application chat GPT which type of AI doe it use; the answer will be Narrow AI. “Narrow AI is created to solve one given problem, for example, a chatbot. Artificial General Intelligence (AGI) is a theoretical application of generalized Artificial Intelligence in any domain, solving any problem that requires AI. Though still unfulfilled, AGI inches ever closer. General AI can respond to different environments and situations and adapt its processes accordingly(levity.ai)”
Thomas Siebel states that Big data lays the foundation for the broad adoption and application of AI. Three main traits that characterizes big data: volume(size) , velocity(speed) and variety(shape). Greater data volume and velocity can improve AI algorithms and better AI performance. IoT devices are important contributors to increase velocity. The shape of data can be images, video, telemetry, human voice, handwritten communication, short messages, network graphs, emails, text messages , tweets , comments on web pages, , calls into a call center, feedbacks on company’s website etc. 70 to 90 percent of the data in the world is unstructured data.
It seems that to reach the level of human judgment in banking sector , besides narrow AI the AGI also needs more focus. And also, data variety that is being used by banks are still questionable due to limited contribution of cloud computing and unstructured data. Traditional banking use cases in supervised machine learning algorithms might be in advanced level especially in fraud and underwriting field. However, models for convincing customer to cross sell or upsell needs wider variety of data. Especially, customer emotions and engagement in social networks are influential according to recent researches.
References
- Noise A Flaw in Human Judgment , Kahneman, Sibony, Sunstein.2021.
- Artificial Intelligence: A Modern Approach. Russell. Norvig. 2021
- Digital Transformation: Survive and Thrive in an Era of Mass Extinction.2019
- https://levity.ai/blog/general-ai-vs-narrow-ai
