Working Up

by Dwayne Phillips

I am pretty good at standing on the shoulders of others. Can I learn to stand on my own shoulders?

I am pretty good at standing on the shoulders of others’.

1 Someone else works, works, and works to arrive at something. Whew. They are exhausted.

2 I read it. Aha! How about extending it this or that way? Alright!

How can I do 1 and 2 with my own work? I work, work, and work and have something. Whew. I’m exhausted.

I want to wake the next morning renewed and recreated. I want to see my own work with clear eyes and a clear mind. I want to extend what I did yesterday.

I struggle with that.

Can I stand on my own shoulders?

by Dwayne Phillips

There is much good in the act of sitting down with another person and explaining to them what I think I am doing.

It isn’t always easy to explain to another person what I am about to do. It is, however, a good use of time. That act usually shows me what I am doing and if it is a good idea or not.

Years ago, the person sat across a small table from me. He was to describe a new program that was about to commence. He looked at me. Before he opened his mouth, it hit him. This was a bad idea that was not going to work.

It was a simple moment. All the great ideas, all the great conversation held in a closed group, when pushed out into another arena with other persons melted and evaporated. Nothing remained. Nothing.

A worthwhile exercise.

I guess there is much that could be written about this. Conversations among a cloistered group are suspect. The light of day exposes cracks in concepts. The metaphors are endless.

Find someone who hasn’t heard my idea. Find someone who is smarter than me. Find someone who has no stake in the outcome. Sit across a small table. Look them in the eye. Explain myself. Perhaps I will understand myself.

by Dwayne Phillips

The pandemic has its own phrase for the use of bad or junk science and poor thought and communication.

It’s in the “news” almost everyday now. I feel it came about from the pandemic, but there are other excuses we can use. It is that phrase in the title,

We know more now.

Well, of course we know more now. We are a day older and we probably learned something yesterday that we didn’t know the day before yesterday.

We told the world definitively a few months ago that, “This is the way it is. This is what we are doing.”

Recently, “We know more now, so forget what we told you a few months ago and abide by what we are telling you today.”

A few examples that illustrate the folly:

• The Kansas City Chiefs are the best team in football. We know more now, so…
• I told a hundred persons that I would be faithful to my wife till death do us part. I know more now so…
• I told the judge I would tell the truth, the whole truth, and nothing but the truth. I know more now, so…
• We convicted so-and-so of murder before we had any DNA science. We know more now, so…

The reason we go with “we know more now, so…” is that we didn’t think it through a few months ago. We didn’t choose our words. We didn’t communicate what we knew and what we didn’t know and what new evidence might change our recommendations.

In other words, we were lazy and a bit dishonest on top of laziness.

Professional professors used to say things like, “Given the evidence to date, here are our conclusions and recommendations. Our knowledge is limited in the following areas. If we learn such-and-such, that will change our recommendations. We will search in those areas and advise others to do the same and listen for changes in evidence. We have x% confidence in our current conclusions, so abide by our recommendations understanding that percentage.

Gosh. That’s a lot of work. Let’s just tell folks to wash their hands all day (NOT).

by Dwayne Phillips

We tend to reinvent the past as we move into the future. Remember JCL?

I took an class in operating systems in 1980 (yes, I am that old). At least that was the name of the class. In reality, it was a class in what IBM called Job Control language or JCL.

You setup a computer—the IBM 360—with a few lines of something that the professor called a programming language. As students, it looked like strange symbols and other things we could not describe.

There were 80 columns on a punch card or terminal line. Depending on whether you put a comma, colon, or semi-colon in column 77, the machine appeared differently and performed different operations. Now of course, that depended on what symbol you put in columns 76 and 71. That all changed depending on what…

Are you kidding me? That was torture.

Fast forward a decade or few and we have Infrastructure as Code. Write ASCII characters that describe the computer you want to use. Cool. Great idea. Think of all we can do with this concept.

Then, however, you look at examples of Infrastructure as Code. They aren’t as bad as JCL was, but, well, really? We could have done better. Huh?

Perhaps I expect too much. Perhaps I never understood the purpose of JCL and don’t understand the purpose of Infrastructure as Code. My expectations are all askew.

by Dwayne Phillips

Once again, someone creates all sorts of fancy explanations for what was simply a bad idea poorly conducted.

There seems to be a lot of “bias” in the machine learning area of artificial intelligence research and practice. Or can we explain the problems without using such fancy terms like “bias?”

• What were the requirements?
• What were the tests?
• Did anyone review any of this?
• Did anyone ask any questions?
• Was anyone allowed to ask any questions?

It is easier for stupidity to slip through when we skip these steps. Afterwards, we create all sorts of fancy reasons for the stupidity (bias, over sampling, under sampling, data integrity, governance, provenance, etc.).

There are persons who know how to approach technical problems. There are sound and proven techniques for approaching technical problems. Many see these persons and techniques as boring or old fashioned or something. That is unfortunate.

If you are managing such advanced efforts, please, step back and use some of these old, proven techniques and maybe even bring in some of these old, proven persons as well. At least for a day.

by Dwayne Phillips

The flat, horizontal surface is a natural. Home architects know their value. The designers of automobile interiors, however, seem to be living on another planet.

In automobiles, the dash (the area above the displays) is curved and sloped. Flat surfaces don’t exist.

So where do I set something that I want to stay put for a moment? The phone slides off. The tablet and laptop slides off. Pencils slide off. Books slide off. (Need I go on with the list?)

Coffee cup? Let’s not even mention that one. There are the cup holders, but they are in the most inconvenient place possible and only work if your cup is the same size as the hole they gave you.

Please, please give me a flat surface with a lip around the edges so that spills stay put and pencils don’t roll off.

Does this require some sort of genius?

In the age of work from anywhere but the office building, this seems like a natural. Does Weather Tech have to make a mat that sits on the curbed dash, levels it out into a flat, horizontal surface with a lip to contain spills and rolling objects?

Hey Weather Tech, here’s a million dollar idea!!!

by Dwayne Phillips

When many volunteers arrive, the expense shoots up. Are we prepared for this?

Volunteers are great! All those fine people arriving and willing to work for no pay.

Well, not exactly. A person’s time is worth money. A person’s time is worth other things. When a person volunteers, they are spending their resources on my cause instead of something else.

One hundred volunteers arrive. That is $100,000 an hour expense whether I pay them money or not. Consider the responsibility:

Am I managing this effort well enough for a $100,000 per hour expense?

Years ago I wrote a report titled, “Dedication Requires Management.” The point of the report was when there are dedicated persons who do the best they can all the time, I need to match that dedication with good management.

When I have 100 volunteers appear, I need to match that effort with good management. If I cannot provide that high-quality management, I have too many volunteers. I must politely and respectfully send 90 of them home until I am ready for them. I will probably need them one day, but only when my management expertise is up to the task.

Many volunteers bring a high expense. High expense means high waste unless I am able to manage well.

by Dwayne Phillips

Research is expensive. Someone pays the bills. Sorry about the lesson in economics.

The future of companies is often in the research they perform in the back rooms. Not glamorous, but often necessary. And those smart folks in the back rooms cost lots of money. The company has to be making enough profits to pay those bills. Sorry about the lesson in economics.

DeepMind is a place full of brilliant folks. It also costs a lot of money with little immediate return. Sorry about the lesson in economics.

Google recently fired an AI researchers and wants more positive papers. These things are awful. Sorry about the lesson in economics.

AI researchers are good folks trying hard to do good things. They aren’t exempt from the First Rule:

There is no free lunch.

Sorry about the lesson in economics.

by Dwayne Phillips

We often use a “figure of speech” to describe something. Well, that’s nice, but I know that such-and-such cannot literally do something-or-other. Can we please have some literal descriptions?

I have recently been reading constantly about a very popular software app. I won’t mention which one because that would take us off the topic of figurative and literal descriptions.

All the descriptions of this very popular software app use figures of speech. An example figure of speech is, “This software manages that thing.”

Well, no it doesn’t. People manage things. Sometimes people write software to help them manage things. The software does some specific tasks that help the people do their job.

Those specific tasks can be described literally. Please speak literally instead of figuratively.

For example, this piece of software appears to manage things and actually helps me manage things when it:

1. reads parameters that I entered in a text file
2. reads status from a text file
3. uses if…then statements that use the parameters and the status
4. calls control routines based on the if…then statements

Why don’t we have more literal descriptions? I believe the answer is that it is more difficult to understand something well enough in context to describe it literally.

It’s simply harder. Fatigue always wins.

Oh, thank you for the effort required to describe something literally. Now I understand.

by Dwayne Phillips

Are we spending time in gold mines or coal mines?

Of course we want to be in gold mines. Untold riches. Right? Consider a few things.

Gold Mine resource: Now and then we find a nugget of gold that is valuable almost beyond measure. Most of the time, most of the days, there is nothing but dirt.

Coal Mine resource: Every shovel brings energy.

Now, where do we spend our time?