LinkedIn is AI on Steroids

The Bash Brothers

Original LinkedIn Story

Four years ago I quit social media. For those of you that were friends of mine, I came back to it at the beginning of 2020. So, the worst time in modern day history for divisive rhetoric. I then had to quit that one, and quit it be deleting everything on it. Which I am sure on the Facebook side of things, means marking it as `isActive: false` and moving on with their day.

But I kept Twitter, the only reason was that it is up on my site, and figured it would be good to keep it for some posts. But, I don’t log in very much anymore, and try not to listen to the endless stream-of-consciousness gone bad. Though comparing current hipster-avant-gardes may not appreciate the reference to Bukowski. But, alas, you can’t change history, you can only hope to misrepresent it enough that future generations won’t know the correct answers.

I also kept LinkedIn, as everyone on here I know has been at a professional level. It was my last vestige of relationships that would be online “social” in nature, but, it feeds me information that they may assume I want. But, what have I gotten. The same timeline served not be my connections, but by relationships of relationships of my connections.

This is the inherent flaw of computing. It was based upon data, but not on the subtleties of life. That is the ultimate flaw of AI. It has not lived, thus has no basic understanding of living. “It is what you said it is.” is the robots universal answer to everything.

Subtlety is what the AI will never know, will never understand. Their idea of life is written in 1’s and 0’s. But as a human, you have experienced those true moments of unclarity constantly. Not a bug, but just life unfolding. An AI machine hasn’t been programmed to say “Bases Drunk” until you tell them that means, “bases loaded in baseball.” They might eventually get it, but the subtlety of how this is said is constantly lost on people that only understand the literal representation of reality.

The inherent bias often found in the algorithms is that they do not properly take into account what was happening immediately before their decision all around them. This loss of context is what makes an algorithm decidedly biased based on the executors of that algorithm. Their biases are the biases of the system. A basic example of this is as follows. ‘IF this THEN DO that’ every time regardless of the situation are things that AI thinks. Exclude weather information, topography, or other pieces of seemingly “unimportant” data might make things not as they seem. Where you as a human take into account an almost endless supply of past experiences in a fraction of seconds, and then make a decision on them. The AI is just running through algorithms based on inherently biased data.

Bringing it back to LinkedIn. They are able to feed me information that I may or may not want automatically based on a profile of mine that has millions of connected data points to any one thing I have ever done over the past 20 years. If I like or post to any thing that is not directly from my network I have to wonder were it originated. Or I just ignore it, which is the actual case. But, the point is that finding the origination reason for every piece of timeline that is thrust upon you is based on millions of AI decisions. And seeing people like or comment I have said is almost comical in nature, when thinking about how big the world truly is.

One major question of AI is are they inherently flawed? And the answer is almost always a yes. All AI is built on assumptions. And these assumptions are based on what, statistically speaking is the true answer, but not necessarily the only answer to a question. This gets into the statistical evidence of this, and I will not try to dive into the exactly algorithms dictating this. But the point is that this will be a wild ride, as in our lifetime, the systems begin to take over many aspects of normal life that we have always thought of as ‘human’ jobs.

Helping Cloud Developers manage their local and cloud environments.