Working Up

by Dwayne Phillips

I have to write all the parts to have a whole. But the parts that comprise the whole will probably change.

I am facing a big writing project (“big” of course is subjective). I have divided the whole big project into parts. Now, write each part, check off the parts in the list, and I will have the whole.

I don’t feel like writing the first part first. I don’t feel like writing the second, third, or even fourth part first. What do do?

1. Write the part that I most want to write, first.
2. Look at what is left and repeat step 1.

Well, I suppose that will work. Correction, I know that will work. But what’s the use? I mean, I have to write all the parts to have a whole. Why not just “bear down” (or some other cliche) and write the parts in order?

Simple answer to this obvious but no-fun question: until I am finished with the whole, I really don’t know what parts comprise the whole.

Once I have been writing for a while, the list of parts will probably change. The part that I didn’t want to write, may be unnecessary. I didn’t have to struggle with it after all. It went into the trash can before I wrote it.

Lazy approach? Maybe, but who am I to call myself “lazy” when I am writing all the necessary parts that comprise a whole and I ship the whole piece to those who need it?

Write the most interesting (to me) part first. The rest will fall in place. At least, that is my experience.

by Dwayne Phillips

These chattering bots can “generate” all sorts of information. Humans must do more and better to keep their paying jobs.

I recently sat in a lecture about some worthwhile topic. The lecturer started with, “There are eight aspects to this topic that we will cover in the next ten sessions.” The lecturer listed those eight aspects on the PowerPoint screen and described each in some detail.

One of my fellow lecture attenders pulled me aside afterwards. While sitting bored in the lecture, he had asked ChatGPT, “What are the major aspects of this topic?” ChatGPT immediately spit out the same eight things our learned lecturer listed. Further prompts produced better summaries and details of the eight topics than the lecturer produced. (There are many other systems that do the same, but I will use “ChatGPT” in this short essay.)

Our lecturer could not spell ChatGPT let alone use it. Hence, his lecture was the result of long hours of hard work, i.e., someone paid the expert lecturer lots of money to do something that a novice could do in 15 minutes with ChatGPT.

And hence we arrive at higher expectations for human intelligence and human experts.

“ChatGPT could have pumped out that material. I am paying you money to do much better than that.”

I have yet to hear the above statement. That is an unfortunate indictment of ignorance on those who hire human experts. It is also a call to action for human experts everywhere. Laymen can produce accurate information on many topics. If a human expert wants to continue to be paid for expertise, its time to up our game.

Start lectures with, “You could pull much information on this topic from ChatGPT. That includes these eight main aspects. I point you to the accompanying handout for such. Now we will delve into material that ChatGPT doesn’t ‘know’ yet.”

New tools are valuable. They are also forcing experts to do better. Let’s do better.

by Dwayne Phillips

You think the US government is big and spends lots of money? It is now overshadowed by industry—especially in computing.

I was an employee of the US Federal government for 28 years. We did big things in computing that cost big dollars. I once worked in a lab where we had four supercomputers. One of them cost $6 million while the other three cost $4 million each. No one had that type of computing power.

There is news that Microsoft and OpenAI will build a data center with an AI super-duper-computer that will cost $100 Billion. That is Billion with a B. No government in the world can afford such a computer center.

At least in the field of computing, governments have fallen far behind industry. Innovation and production occurs in industry. Governments sheepishly ask for handouts.

“Can we please use Copilot for a reduced price? We have 10,000 employees. That should be worth some consideration, huh?” That doesn’t make a dent to a company that has 100 million registered users. There are no home-town discounts to a government that drug your company in court for ten years with no result.

The US government tied IBM in court for a decade. No result. The US government tied Microsoft in court for ten years. No result. The US government is starting the same with Apple. Experts predict that same no result.

Adjusting to the new world? Hardly. The US House of Representatives bans the use of Copilot among its staffers. No need for better product and productivity. Just continue to plod along. We are, after all, THE GOVERNMENT.

That doesn’t carry any weight any longer. Some haven’t realized this. Many outside of government realized it a decade ago. I am still associated with the US Federal government. We can do better. Here’s hoping for a better future.

by Dwayne Phillips

Take caution with negative statements as they often contradict themselves.

“There is no crying in baseball,” is a famous line from the movie A League of Their Own. (My wife says that is a movie about sisters. I say it is a movie about baseball.)

That is a self-contradictory statement. One person is crying tears while another is crying out loud. Both are crying while not crying in baseball.

Most negative statements can be self-contradictory:

• We cannot have cheating on this test. (Why would we have a test where cheating is beneficial. Isn’t that cheating the purpose?)
• We cannot have tests that discriminate. (If a test does not separate those who know from those who do not know, how is it a test?)
• We cannot object to what is said. (If we are saying things that everyone agrees with, why are we wasting time saying things that everyone already finds agreeable?)
• There is no talking in this hallway. (You are in the hallway telling us not to tell anything. Why are you violating your own instruction?)

I could go on with statements and then explanations of why they are self-contradictory. Solution?

Positive statements about what we desire to happen:

• Show what you know on this test.
• Once we enter the hallway, we will all walk silently.

Again, I could go on with positive statements that do not contradict themselves.

Take great care with negative instructions or admonitions. We can do better. Decide what it is we want to say and say that—not a twisted negative statement that attempts to convey our meaning but usually conveys the opposite.

by Dwayne Phillips

Keep your earth-shattering new thing quiet for a while. Once it works, shout. This is basic risk management.

Something has never been done before. That means it is difficult. I have a solution. I can do it or I think I can do it.

The best practice, in my experience of doing something or other for the first time, is to keep my mouth shut. Do the work. Make it work. Fly under the radar and keep it quiet.

Once it works, shout!

Since it has never been done before, the work carries risk—lots of risk. All prior efforts have failed. Simple logic shows that my effort will also fail, until it succeeds.

This is risk management. Do the work quietly, but do the work diligently. This isn’t easy, but it can work.

by Dwayne Phillips

Of course software running for the thousandth time works better than some people at some tasks. It has for half-a-dozen decades. Why does this continue to surprise us?

Here is a recent breath-taking story about how AI performs better than doctors at detecting a type of cancer. Of course it does. Put a digital camera on the front end and software running on a computer on the back end. It works better than a human.

The simple reason is the software doesn’t have bad days. The software isn’t tired after a restless night caused by a sick child or worry of a cancer-ridden relative. The software wasn’t in a car accident on the way to work or didn’t have to circle the parking lot for half-an-hour looking for a space.

It has been this way for 50 or 60 years or more. Basic classification algorithms existed long ago and ran on what we would consider to be archaic computers with almost no compute power and memory.

We had these discussions in the 1980s. Expert systems performed better than people. Of course they did. The were not, however, 100% correct. Hospitals stayed away from systems known to be less than 100% because they didn’t want to be sued. Other industries stayed away for the same reasons.

Today, the systems are not 100%, but we accept that now for one reason or another.

Today, we call it “AI” and run supercomputers that slip into our pants pockets. We gasp at the performance. The end of the world is nigh. Well, at least the end of someone’s career is nigh. Old news. Let’s use what we have and move on to some other problem that needs solving.

by Dwayne Phillips

Reducing the time spent typing on the keyboards reduces the number of interruptions. There are new tools that reduce both.

I can type fairly fast. At least I give myself credit for that. Like everyone else, I am interrupted while typing. I could estimate number of interruptions per hour or some rate like that.

The math shows it obvious that one way to reduce the number of interruptions is to reduce the time typing. I should learn to type faster. I really should, but that method of reducing interruptions has its limits, and I am pretty close to those limits now.

Enter a new technology called the large language model and these generative AI tools. What I find in my use is that they reduce typing. I ask questions, a 500-word essay pops out, copy, paste, and edit.

I find the biggest benefit of these tools is the reduced typing. Reduced typing time, reduced interruptions.

Yes, there are problems with these new tools. No, I don’t use them blindly. Yes, they can reduce typing and reduce interruptions. That is a benefit that comes with several noteworthy detriments. Still, it is a benefit. Let’s try to take advantage and reduce interruptions.

by Dwayne Phillips

We have yet more examples showing how remote sensing is difficult. One day, we learn this well enough to anticipate it?

There have been several unmanned craft land on the moon recently. That is a great accomplishment to send something to the moon and have it land soft enough to still function. There is some old saying about any landing you walk away from is good.

Anyways, a couple of the recent unmanned landings on the moon were soft, but, well, uh, not great. One craft landed upside down while another landing sideways, sort of.

I recall the manned landings on the moon in 1969 and the following few years. Those had experienced pilots on board who looked out the window, saw the ground, and made all the adjustments that a person makes when on the scene.

The recent not-so-food unmanned landings were piloted remotely, well, sort of. A type of “auto pilot” guided the vehicles to the surface. Remote control was not possible as the delay in transmission prohibited real-time remote control.

The landings were characterized by a lot of remote sensing. And, well, we know that remote sensing is difficult. I have written on this topic several times before. Remote sensing is STILL DIFFICULT.

Perhaps we will acknowledge that well enough one day to anticipate it. Unmanned spacecraft landing on the moon and other remote bodies depend on remote sensing to work well. That is difficult. That requires a Plan A, a Plan B, and so on.

Flying all the way to the moon is difficult. Landing well via remote sensing is really difficult. It’s that last 100 feet that is critical. Let’s keep trying and let’s do better.

by Dwayne Phillips

To be intentional is to do something that you intended to do. I guess that is better than doing something accidentally or unintentionally. Yet, it has no meaning.

To be intentional is to do something that you intended to do. It is to do something on purpose.

Eating is intentional. I eat because I want to eat; I intend to eat, and I eat on purpose. I am an intentional eater.

Working for a paycheck is intentional. I want a paycheck, so I work. I work on purpose with a goal in mind—a paycheck. I am an intentional worker.

Watching TV is intentional. I watch TV on purpose. I want to watch TV. I am an intentional TV watcher.

All these any many other actions are intentional. I guess I can find something I do that I do not do on purpose.

Hmm, give me some more time. My searching for an unintentional task is quite intentional.

Did I waste enough words intentionally?

by Dwayne Phillips

We reward the fireman, the person who extinguishes a fire. Did the fireman, however, start the fire?

I used to see this often. I worked in a place where engineers would plan projects and deliver systems per their plans. The trouble was: the engineers were terrible planners. They were good system designers and builders, but terrible planners.

The good designers and bad planners would plan projects that would start, finish, and deliver in short periods of time. That was the problem: the periods of time were much too short. The plans were overly optimistic, yet were written in stone. One week into a six-week project, everyone realized that it was a ten- or twelve-week project.

Well, promises had been made. To ask for more time would be to admit a mistake, and mistakes were not allowed. The result was the engineers worked 12-hour days seven days a week to meet the original plan. One engineer I recall cancelled a second honeymoon to meet the plan. These were heroic efforts. Heroes were rewarded like heroes.

But these were firemen arsonists. They set a house on fire (bad plan) and rushed in to put out the fire (12×7 work). They were rewarded for a terrible mistake and then a cover-up of the terrible mistake.

Some of us engineers planned well, worked, delivered, and were never rewarded. We weren’t heroes who worked all those extra, unpaid hours. We planned well, worked the plan, and delivered. We didn’t extinguish any fires because we didn’t create any fires. Only heroic firemen were rewarded as heroes.

I suppose the fireman arsonist still exists. Overly optimistic plans lead to heroic dedication to cover up a mistake. Rewards flow to heroes. Those who plan well are overlooked. I think we can do better. Let’s try.