top of page

Unsupervised crowd wisdom wheels

Investopedia Definition: Wisdom of crowds is the idea that large groups of people are collectively smarter than individual experts when it comes to problem-solving, decision making, innovating and predicting.

If crowds are smarter in a lot of contexts how can we leverage that wisdom systematically? In light of the current global situation surrounding COVID-19, the crowd is a shunned word. Humanity will fight back and recover but what will be the shape of that future? What is the place for crowds, social distance, lockdowns, sharing, etc? What happens to crowd wisdom? Is there a way to capture that?

The good news is that it exists and is already perpetuating via social platforms. Think of a YouTube recommendation engine. It exists due to crowd input. Similar recommendations & search systems proliferate search engines, commerce sites, etc. These systems are driven by Artificial Intelligence and are learning the crowd preferences all the time. Can this be leveraged for crowd wisdom?

Current thinking is to use AI in your system to learn about patterns, here we turn the concept on its head. We assume the other system is intelligent (has AI) and algorithmically plays with it to derive crowd wisdom.

Let’s take an example. People search in the billions and use multiple queries around the same concept. Let us take a couple of trending topics and experience the methodology.

Deep Learning… a big AI paradigm… is clearly trending as seen below basis search queries.

Just to contrast, remember Tablet-PCs… they are not trending.

Back to Deep Learning… All related queries and their ranking with respect to each other are programmatically pinged and connected in a network. What you get is the below “Crowd Wisdom Wheel”.

These are all the queries around the topic that have been searched billions of times. At the center are core queries and at the periphery are newer and lower frequency queries.

Basis the flow of the network key nodes can be figured (ones which connect to a lot + have multiple short paths between concepts etc) and clustered to find what are the top areas people are searching for. It is simplified for this blog into the top 5.

Deep learning is about the science, prediction, neurons, graphical processing units, and data as per crowd wisdom. Deep Learning is inspired by brain neurons, uses GPUs and needs a lot of data. Its primary function is prediction. The human search footprint provides the answer.

The whole thing was generated in an “unsupervised” way. No labeling, no training… just algorithms playing with other AI systems.

COVID-19: What does the crowd say about the crisis we face? The shape of the curve indicates the severity of the crisis.

The wheel has aspects about the virus and a lot of countries majorly affected at the center… China, USA, Italy, etc

The main areas are around the virus itself, symptoms and infection. Influenza has a relation to this. People need a vaccine. The USA is reeling currently. The world currently is looking up to some key organizations like the Centers For Disease Control & Prevention, John Hopkins (see in the wheel), etc. Today statistics of the disease spread are at the center of the radar.

Are there topics that should be at the center that isn’t? Are there topics that need to move to the periphery? Are there new trending topics at the periphery? The tool can be used to shape conversations.

Moving forward as people interact more with technology & AI systems a new paradigm of learning would emerge akin to what was described above. People can be understood, patterns can be deciphered, connections can be made without “contact” even under a “lockdown”.

Crowd wisdom is available thanks to somebody else’s AI. Unsupervised.

9 views0 comments

Recent Posts

See All
bottom of page