Think Vs Compute (Part 2)
In a previous issue (here), I wrote about the differences between think vs compute. After writing the posts and gathered some feedback from subscribers. I felt that I still have not captured the essence of it and thus continuing to ponder about it during my long walks.
So in this issue, this is a refinement further on what is thinking vs compute over here, or at least the differences between humans and machines. Hopefully by understanding the current differences I can work on how to incorporate them together, moving towards the vision of machine+human.
Team Human
All along, humans are known to be able to bring context together, which machines are lacking. For instance, humans knows sarcasm or jokes when they’ve seen/or hear one. Sarcasm and jokes, in order to bring out its ‘essence’ needs context. Without the context, it is very difficult to laugh at it. Below is a classic example, from Andrej Karpathy’s blog.
I fed the following picture to ChatGPT and asked if this picture is funny. Below is a summarized answers.
“The humor in a photograph can often depend on the context and the viewer's perspective. In the image you've provided, it appears to be a candid moment where one man is being measured for his height, while another man, who is watching, seems to be playfully imitating the stance of the man being measured.”
“However, whether or not this picture is funny would vary greatly from person to person. Some might find the playfulness and the scenario amusing, while others might not see it as particularly humorous. Humor is highly subjective, and what resonates with one person might not resonate with another.”
The machine does not even want to choose side on whether the picture is funny or not! Interesting!
However, if you post this picture to any humans (so far), most of them find it funny or at the very least break into a smile.
So humans’ strength is in the word “Context”. So my next question is, what is “Context” then? Continuing with this picture, Andrej in his post has mentioned that in order to find this picture funny, there are many points (or dots) that has to come together in order to find it funny. This is where an epiphany struck on me, we humans are very good at connecting the dots, each dots being part of the context/circumstances, dots that are not represented by data collected.
So when it comes to humans, our strength is in identifying and bringing the different pieces relevant context together and making sense of it, or per a friend of mine calls it, “Sensemaking”.
Team Machine
So in this perspective, where does machines or computers comes in? Computers compute much faster than we humans are. Combining the ease and volume of data collected together with algorithms that will take humans a much longer time to calculate, computers can make certain dots more visible, dots that before digitization may not be visible to the humans.
Combining Our Powers
Tapping onto machines strong computation power, putting data and algorithms together, the machines present more dots to the humans to draw them together. Humans presented with more dots will put on their thinking cap, and draw relevant dots together in order to paint a clearer picture of the challenges at hand, i.e. bringing the relevant pieces together to form the necessary context needed to answer the challenge at hand.
Concluding…
What does this say then? Well, jobs that requires a lot of context, such as analysis, communication, design thinking, strategic thinking, problem solving may not be replaced by machines that soon but instead can use machines to increase the job holder’s productivity by taking advantage of the fast compute.
Connecting the dots or sensemaking will be Team Human’s advantage for a while more until we are able to integrate knowledge graphs well into machine intelligence.
If you are keen to stay ahead of machine for your career, learn to tap onto machines’ strong compute power and work on how to bring more dots/context into your work, namely applications.
What are your thoughts on this? Keen to hear from you! :)
Recommended Content
I’ve written a few blog posts on building up skills that might interest you.
Consider supporting my work and content by sharing, or making a small donation! :)
The idea of 'connecting the dots' has always intrigued me. E.g. the connection change if we change the context, even the same dots are there. E.g. is there data about context that can be separated from non-contextual data? How do we as humans recognise it?