Saturday, February 6, 2021

Sophia, Part 2 -- February 6, 2021

Sophia, Part 1

Sophia, in a nutshell:

Sophia's software -- her computerized AI -- is not static. The software is dynamic and will change, it will be updated, as Sophia matures. To the extent possible, the software will be updated by ever-evolving software. Human engineers will step in only if there is a technological glitch. Think of HAL in 2001: A Space Odyssey

She will incorporate the entire Encyclopedia Brittanica the day she is "born." She will have the most sophisticated search engine. She will be able translate any language in real time and reply back in any language.

She will have immediate access to every piece of art. She will have the entire library of music at her disposal. She will have immediate access to every quarterly report of every company in the world. She will have access to all current research in any scientific journal. It's possible that experts in every discipline will be allowed to feed her information even before it's published in peer-reviewed journals.

So, with all that, you can imagine of what she will be capable.
There is one thing, above all else, that will be most important, and it's very simple to articulate. Actually two things, the second naturally follows the first.

Then this, from part 1: 

There is one thing, above all else, that will be most important, and it's very simple to articulate. 
Actually two things, the second naturally follows the first. I'll talk about that later but this is to give you an idea what this is all about. 
But to give a hint: think of two human chess players. One has access to Sophia; the other does not. What is the most important thing Sophia can provide her chess partner?  It's not a perfect analogy but it's the best I can do with a not-ready-for-prime-time reply. 
At a more sophisticated level, let's say Sophia is asked to participate in developing a new vaccine. What is the one thing that she can provide that the rest of her human team cannot provide? 

Now to answer that question: this is what I think Sophia/AI gives us. True AI should not be trapped by group think. 
 
As important as that is, I think the most important thing about Sophia/AI is this: Sophia/AI will ask the question that humans are not asking. One won't answer a question if it's not asked, but most importantly, the "right" question must be asked. 

This takes me back to that famous press conference with Donald Rumsfeld: 


AI should help us with the "unknown unknowns."

To take an example.

A reader and I had a sidebar conversation regarding Covid-19, the vaccine, and variants. 

Along the way, the reader wrote:
At first I looked at it as just a stronger version of the flu… And it essentially still is, however there is a random element to this bug that hits one person hard and hits another easy so until you catch it you don’t know how your body will react. 
Obviously those that have medical issues or in an older age bracket are more susceptible but it can hit a healthy 40-year-old just as easily and take them out so it is this randomness that I think is the most perplexing. 
But what I am most concerned about is the ease that this virus can mutate, seemingly more easily than other viruses can mutate, and it is just a matter of time before we get a version that is much more deadly. So that is the race we are in as I see it.
My reply:
Yes, the question not being asked/explored enough: why is there no talk of chickenpox / measles / mumps / rubella variants when it comes to our vaccines and yet variants within the SARS family seems to be the norm?
Sophia/AI will help us with this. Sophia will help us with the "unknown unknowns." Before we can solve a problem, the right question must be asked, and the better framed the question, the better.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.