Not everything that can be counted counts. The post title and first sentence of today’s missive are a somewhat famous remark often attributed to Albert Einstein, but actually of unclear origin.
I think that some people who rely heavily on data analysis to make decisions–for example, like some football coaches–are actually doing so in order to absolve themselves of responsibility. “Well, the analytics said if we did X we had a 71% chance of winning, but if we did Y our chance was only 69%.”
Not all nominal differences in data are statistically significant. In addition, the underlying relationships between variables and outcomes can change as behaviors change, even if they modify only slightly, often before we realize they have changed. Such shifts render “old” data far less meaningful.
Not sure why I was compelled to write about this today. Yes, I see the irony of someone who was a pioneer in sports analytics and a “father” of Moneyball writing about the limitations of data.
Despite the fact some people believe that “machine learning” and “deep learning” will provide the answers to almost everything, I think that even when we have computer chips planted in our brains that can spit out the history of the universe at a billion words a second, we will still fail to correctly predict everything that happens.
My wonderful wife and I had important business to attend to yesterday that prevented us from attending the Barrett-Jackson auction. I think we will go today, but that is not a certainty.
Here are just a few photos from the auction.
While they are not affordable for me at present, these C7 ZR1s are becoming more appealing. I could have a 700+ HP/700+ LB-FT of torque car without the risks inherent in after-market tuning.
We’ll see how this turns out. Below is a picture of most of the letter I received from the GI practice against whom I decided to file a complaint when I arrived late for an appointment and the receptionist refused to do anything to help me.
Am I just a typical self-entitled American? I don’t think so. A medical practice exists to help patients. Yes, I was late to an appointment (an honest mistake as I thought the appointment was an hour later than actually scheduled), but the receptionist made absolutely no effort to help me. The more I thought about it, the angrier I became. I didn’t file the complaint until six days had passed. I was hoping I might cool off, but I didn’t.
I doubt that anything substantive will result from this, but I had to express myself, nevertheless. Have any of you ever filed a formal complaint against a company? If you don’t mind sharing, I think most of us would like to read about the details. Of course, I once sued a very large American firm, but that’s another story for another day.
If you like this blog please tell your friends and share the blog URL (https://disaffectedmusings.com). Thanks.
8 thoughts on “Not Everything That Counts Can Be Counted”
I am missing Barrett-Jackson Scottsdale this year. I toyed with heading to Kissimmee for the Mecum auction, but the long drive and my schedule didn’t allow for it this year. I will plan better for next year.
Thanks for sharing, JS.
LikeLiked by 1 person
” we will still fail to correctly predict everything that happens.”
I think it’s nearly impossible to predict everything with 100% accuracy, whether by humans or computers. And for the same reasons. The prediction is only as good as the information available, as you pointed out, and that info is everchanging. A term that used to be used a lot regarding computers was GIGO. Garbage In, Garbage Out. The same is applicable to us mortals. We make decisions based on our research, most times, and there are many variations.
A quick example: About 10 years ago a good friend decided to buy a new car and decided to buy a Mercedes, based on his research and their history as dependable, safe and held their value. Ended up with an S class that he documented was in the shop for repairs for over 5 months in the first 16 months of ownership. Abnormality? According to him, not so much based on the number of cars that were also in the shop for repairs. Yes, they were warranty repairs, but after spending 60K+ on a car, he was hoping to get more use out of it. These days there is more info available, if you’re willing to spend the time researching and separating the wheat from the chaff, so perhaps his decision would be different today.
Many thanks for sharing, DDM.
I was at Apple store in Palm Desert, Calif., yesterday. Check-in employee up front wanted to know why I was there. Told him wanted to buy Apple Watch. He told me to come back in 45 minutes, they would then have someone available to help. I said no thanks, wasn’t willing to wait to spend money. Apple just lost an $800 purchase. I’m getting same phone at Costco for less. I’m not complaining, but am in amazement at service attitudes by some.
“Herb,” sadly your experience is all too commonplace. Too many people don’t want to give a real effort at work and too many employers are afraid to fire and/or discipline these workers.
Interesting that you talked about AI/Deep Learning in your post. I was just reading about AI in medicine this morning. Back in 2017 (I wish I could put bold italics here), IBM worked with a former employer who shall remain nameless to bring AI to pharma/medicine. Despite having two multi-billion dollar companies with expansive talent working on the issue, the project was an abject failure.
What folks fail to realize is AI is only as good as the “Real Intelligence” that goes into creating the environment in which AI will function. For example, AI to diagnose heart disease will only work if the humans involved are AWARE AND INTERESTED in looking into preventing heart disease in the first place. Most practicing nephrologists are not concerned with protein in the urine until it hits a level of 300 even though their own society’s guidelines suggest that increased risks begin at levels of 30.
Many thanks, Doc.
“What folks fail to realize is AI is only as good as the “Real Intelligence” that goes into creating the environment in which AI will function.” A-Frickin’-Men!
Comments are closed.