Some Blog Site

One of the memorable quotes I like is “If you are doing something for the right reason, you will probably be doing the right thing.”

That might not be an exact quote, but I think I got the concept right. The context of the quote was an experienced engineer (Lyle D. Feisel, in Lyle’s Laws) advising a younger engineer on how to know if he was doing the right thing. The setting was engineering, of course, but I think it applies to much of life.

For example: if one child is wondering whether he should tattle on another child, ask him to apply this reasoning. Why does he want to tell on him – to get him in trouble or to keep him safe? This might not work, depending on the age of the child, since kids can be devious. But if someone can be honest with himself when answering the question, it works well.

There are some people who may not be satisfied with that rule of thumb, because there is not an objective standard for “the right reason”. And in some areas that may be true. In my line of engineering, I can ask myself “Would I be able to answer the customer if they questioned this decision?” or “Would I be able to defend this action if the federal regulatory agency inquired about it?”

In my case, I have years of working with the same customer, so I can reasonably ask that question to myself. Not everyone might have that same luxury, so their role-playing questions might not be able to provide guidance.

Even outside of engineering, this rule of thumb should work well. Not everything is going to be cut and dried, of course. But if you’re wondering how to decide on what the right thing is to do, then chances are good that you’re on the right path. If you’re asking yourself how to get out of doing something, then this line of reasoning probably is not for you.

I decided to have winter/snow tires for the minivan this year. They’re like insurance – you probably won’t need them, but if you do need them you must already have them installed.

To make life easier, I wanted dedicated rims for the tires. I have no problem keeping a spare set of tires in the barn when they are not being used, and I didn’t want to pay twice a year to have winter tires swapped and unswapped off the one set of rims for the vehicle. I’d rather pay a little extra up front for another set of rims and then it’s all set.

Then I started looking at purchasing rims.

(more…)

Based on the popularity existence of last year’s article predicting bowl games in haiku form, I present to you this year’s all-haiku bowl game predictions. Still America’s only all-haiku college football bowl game predictions.

These are listed in order of date (earliest first). Some picks are whom I think will win, and some picks are whom I want to win. I’ll leave it to you, the reader, to decide which is which.
(more…)

At first, people just coughed into the air.

Then someone realized that was spreading disease to people in the area, so it became polite to cough into one’s hand.

Then someone realized that was spreading disease to people who touched things that hand touched, so it was taught to children to cough into their elbows.

Now I’ve realized that’s fine for children but bad for adults. So now when this generation of kids grows up and instinctively coughs into the elbows, it’ll still cause problems.

Why it’s a problem is that adults use their elbows more than kids. Have you ever seen an adult go very long without crossing his arms? And where are that adult’s hands when his arms are crossed? Fully in contact with his elbow, that’s where.

Picture it – someone sneezes or coughs into his own elbow. Then, just seconds later – before any germs have a chance to die – that someone puts his hand right into the thick of the most germ-intense place around. Are we as a society saying that’s an improvement?

Okay, maybe it is a little better, but it doesn’t quite solve the problem. There are two ways to solve the problem. One is to train people not to cross their arms. The other is to pick a better spot into which one may cough or sneeze.

I’d like to go with the second option. Some possibilities: the inside of one’s shirt (as my brother is wont to do – just pull your collar up and lean your head down slightly and all the germs are contained. This is not a possibility if you are wearing a necktie.), higher up on one’s arm (gives the germs a landing place but doesn’t contain them well), or a handkerchief (that no one carries anymore, but I suppose a hat would work).

Any better ideas?

I know it’s not in the best interest of the companies that send out emails, but I don’t like how unsubscribes are handled.

I think Europe has the right idea of the right-to-be-forgotten. Right now in the US, all of my unsubscribe options take me to a screen that has my email address and my settings, so I can choose how or if I want emails from them.

The problem is that still leaves my email address in their databases. I don’t want these email lists to save any of my information – I want an option to delete my information.

My wife got some scam emails recently that claimed to have hacked her email and password and instructed her to send them some bitcoins in order for them to not send out incriminating stuff to all her contacts.

We could tell that it was a scam, because we use different passwords for everything, so we knew the password they sent us was not her email password. And because of the unique passwords, we were able to figure out what happened.

The hackers got into some website where my wife had signed into once or twice, and they stole everyone’s emails that they used to sign up for the site along with the passwords for that site. Then they sent phishing emails to everyone from there, assuming that some of the people use the same password for everything and thus wouldn’t know their email really wasn’t hacked.

If people had the option to delete their information from the site, then the damage from these types of attacks could be reduced.

The action when I click “Remove me from this list” should literally remove my information from the records, not just add a note that says I don’t want emails.

On the plus side, I’ve noticed that some of the “Manage your email” pages let me not only opt out of various types of emailings, but they also let me change my email address. I think that is, while not perfect, certainly an acceptable way to remove my real address from their system. Sure, let me update my email address to someone@example.com…

Ok, last point – it should not take 10 days for your system to update my email preferences. If your back-end database really takes 10 days for settings to propagate through it, you need to fire whoever set that up. Unless it runs off magnetic tape reels that need to be manually switched – if it’s that old then you get a pass. But any email list management software worth its salt should be able to make changes instantly. Those claims of 10 days are nothing more than feet-dragging policy.

I have to wonder at Home Depot’s logic for shipping things.

I had an order I placed online, and it was under their $45 threshold for free shipping, so I added a couple of small items that I was planning on getting eventually anyway. So the total was just over $45, and it was 5 items.

Guess what they did – they put the 5 items into 4 different shipments.

I would have expected that free shipping is made profitable by running things together – bundle a bunch of packages together for economies of scale and reducing overhead.

But not for this order. They made it about as inefficient as possible, at least from my view.

And no, it’s not like they ended up arriving together. The delivery guy came to our house 4 separate times for these 4 packages for the single order I placed. It’s slightly annoying to me, it’s probably more annoying to the delivery guy, because he drove to our house 3 days in a row to deliver a fairly small package each time.

I’d have been happier if everything arrived at once. In fact, I’ve noticed that Amazon gives me that option sometimes – deliver in as few packages as possible. That’s an option I like.

My guess is that Home Depot has focused on using the efficiencies on their side, rather than the customer’s side. Maybe they shipped each items from the closest warehouse that stocks that item, and in my case that happened to be 4 different warehouse that worked out to 4 different delivery dates to me.

It is socially irresponsible for filmmakers, authors, and others to continue to produce content that portrays AI taking over the world and threatening mankind.

Before going into that, I need to make sure we’re all on the same page regarding how AI learns things. It is not at all fancy or complicated. I think “recognize a pattern based on” is a more appropriate term for what AI does than “learn”, since learning involves intelligence and AI is just a lot of data points and a big memory. But I’m going to keep using the term “learn” because it’s shorter and easier and is the accepted terminology. Also, I think the concept of the singularity is wrong – it can never happen because of things like “life” and “soul”. Anyways…

If you want AI to learn about something, just give it a bunch of that thing and tell it those are that thing. If you tell it the wrong things, on purpose or on accident, the AI will learn the wrong thing.

The current popular example of this is sheep. Lookup “AI learns sheep wrong” and you’ll get a bunch of articles about the image processing AI being trained to recognize sheep. It was fed a bunch of photos of sheep, and it learned to recognize sheep. The problem is that all the photos were of sheep in fields, so the AI associated “sheep” with grassy meadows. If you gave it a photo of a grassy field with no sheep, it would classify it as a sheep. And if you gave it a photo of someone holding a sheep, it would classify it as a dog. It doesn’t really know what a sheep is, it can’t look at its inputs and realize if something is amiss. It just finds a common pattern and applies that to the term from the inputs.

Now apply that to the future of AI. If, as the fear goes, AI grows exponentially and takes over the world, how is it going to train itself? In other words, how will a self-aware AI know what a self-aware AI is supposed to do? It would gather instances or descriptions of AI and learn from it.

So now what are all the examples of AI running the world? Science-fiction books and movies and maybe a few TV shows. So now imagine an AI takes over the world in real life. It is processing all sorts of things including those books and movies, and it comes to understand that an AI ruling the world must necesarily be bad for people – it must fight them or enslave them or terminate them or whatever. Because those are all the examples it has, so that’s what it will learn.

Dear science fiction writers: please help the future of the world by producing stuff that shows an AI running the world and being helpful and kind to people. In case that day comes, we want some good examples for it to learn from.