Working Up

by Dwayne Phillips

Let’s curb our enthusiasm about these chatbots and large language models multiplying productivity and costing jobs.

I just saw another article about how ChatGPT and the like are being used in yet another field of endeavor with giant gains in productivity. We won’t need half the people we have in the office now. Just cut their jobs and use technology to get ‘er done.

Let’s take a deep breath—something we should do much more often, even in “normal times.”

Yes, these new chattering things are fun. I type a simple question and “poof,” there are five paragraphs I can copy and paste. Viola’. Time for a nap. I’m exhausted.

This is new and exciting—NOT. Well, yes, the user experience of type, read, copy, paste, done is new, but the Internet is not new. Wikipedia is not new. The Chicago Manual of Style online edition is not new. The dictionary and thesaurus online is not new. The online Scrabble allowed words thing (I have no idea what you call that) is not new.

For 20 years I have been able to search, read, copy, paste, done. Hey wait, that’s just like… Instead of search I type. You mean I could have been doing this for the past… Wait, no one told me…

Sorry, yes, some folks have been doing this for 20 years. We have an expression where I work, “Let me Google that for you” or LMGTFY.

Chattering (my shorthand for using these new and fabulous tools) is new and exciting. Everybody has discovered the value of all previous human knowledge. Well, some of us discovered that a while back and have been using it for quite a while now.

The current exuberance will fade a bit. My grandchildren (now teenagers) will consider chattering to be normal and most of today’s teens will forget how to do it by the next school year. Teachers will still be debating how to use things (yes, those in bureaucracies are that far behind) and students will stare at the ceiling and yawn while the “adults” are preoccupied with yesterday’s news.

What’s the point? Chattering will multiply productivity for some. We have folks who want to be more productive and get ahead. Most, however, will settle back into their long-formed habits. Their bosses will also settle back as well. No reason to do too much during the work day. That is the way things have been in this world ruled by the 80/20 rule or whatever we call it.

The exceptional folks are exceptions. Nothing to see here. Move on.

by Dwayne Phillips

Recent advances in chatting or Q&A software has provided the ability to write simple computer programs. Hooray!

There is a trend in the workplace called “low-code/no-code.” A person at work writes a ten-line computer program that is helpful in that it will do something in a minute that would take the person several hours to do. The person has automated a task that is simply tedious and error prone.

What is new about this low-code/new-code is that a person doesn’t need a computer science degree or such to learn how to write a little program that automates a tedious task. Other programmers have done all the work and created libraries of software. Write one statement that calls a jillion lines of code and there you have it. Useful stuff.

This low-code/no-code works. The only surprise is that few people are doing it.

I call this “hobby” programming. I used to write real computer programs that were complicated and took thousands of lines of code. That was years ago. My career took me down the path of supervising people who wrote real computer programs. I wrote books about how to do that—lead the programmers and manage the work of engineering.

No longer. I still write computer programs, but they are ten lines long and automate tedious, error-prone tasks. Just hobby programming.

And now we have these AI chatting software things like chat.openai.com, bard.google.com, phind.com, etc. Describe the little low-code/no-code, hobby program and the software “writes” the program for me. Hmmm, this is pretty good. And yes, it actually works.

But, but, but… sorry. It actually works. And what is nice is that I can test the software that the software writes.

But, but, but… sorry. This won’t replace real programmers who write real, complex programs. And there is some skill required to use this software that writes hobby programs. The better I specify what I want, the better the result. Hmmm, I have to talk to the software in precise, concrete, and specific terms. That English composition class I took my first semester of college pays off again.

We have a new productivity tool. Decades of work bear fruition. We stand on the shoulders of those who worked before us. Let’s not waste our time standing on one another’s toes.

by Dwayne Phillips

The old ways of doing AI are still better than some of the new ways in some cases. The answer, of course, is to combine the best of all to do something better.

Feed the following into a chatbot: “The cow jumped” The chatbot will finish with “over the moon.”

That is from the 16th century English nursery rhyme “Hey Diddle Diddle.”

In Large Language Models (LLMs), the machines “learn” from vast amounts of text acquired from the Internet. In one respect, the “memory bank” of the computer is filled with all that stuff and the LLM mimics what is there. This machine learning works pretty well.

Now to reality: cows don’t jump and they certainly don’t jump over the moon. The expression “over the moon” could be replaced by “for joy” as in “The cow jumped for joy.” Still, cows don’t jump (very high).

Now back to the old days of AI which are now called Symbolic AI. There was no scraping the world and mimicry. People wrote “rules” about things. Some of the cow rules were:

• have four legs
• females provide milk for people who can digest lactose
• males have horns
• are mammals
• have hide can be tanned and used for clothing humans
• walk slowly
• etc.

Note that “can jump” is not one of the rules. Hence, symbolic AI techniques would never conclude that the cow jumped over the moon. Hence, symbolic AI was much better at avoiding that mistake.

Symbolic AI, given its better performance in regards to cows and jumping, also had problems not experienced by LLMs and machine learning. The attributes of cows above is pretty short. It can be longer, much longer. How long does it need to be? We never discovered that. We never really wrote a good rule set describing cows.

Alas, symbolic AI works to prevent us stating that cows can jump over the moon. LLM works to do many wonderful things other than cows and jumping over moon. The solution, of course, is to use the good in both and avoid the bad in both. Simple idea, difficult to implement.

Let’s try to do better.

by Dwayne Phillips

This is a basic practice. It is unfortunate that many have either forgotten or never knew this practice concerning documents.

I find a document on the disk drive farm or network or whatever it is we call these things today. There are a jillion files out there, but I find the document I want to use.

Well, I think I found the document I want to use. There is no date or version number on the document. There is no author or authoring organization on the document. Hmmm, no one wrote this ever, but it seems to exist.

I wonder if this is the only document with the same title and filename on the bunch of disks I can access. There are all these different disks and all these different folders and folders in folders and so on. Some searches show me that there are several dozen files with the exact same file name scatter hither and yon.

Am I looking at “the right one?” Is there a “right one?” Gosh. This is confusing. Computers were supposed to make this simple(r).

Here is a fundamental practice:

• Have one file with one file name in one place.
• No duplicated copies anywhere else.
• All documents have authors, dates, and version numbers.
• If a document needs changing, the right people meet at the right time in the right place and change the document. The authors, dates, and version numbers change in a controlled manner.
• Anyone who needs the document can find it in the one place it is stored.

Boring? This depends on your point of view. Most people consider this document control stuff to be boring. Then again, most people who need the content of the document right now quickly find and use it, right now. They aren’t happy because they expect things to be this way.

Is this all passe’? Sorry, no it is not. I daily see people wasting hours trying to find the right copy of the right document in the right place at the right time.

We can do better.

by Dwayne Phillips

In which I consider what is the ultimate purchase for the home, family, and home business: the table.

Consider the humble flat horizontal surface. Narrow the consideration to the kitchen table with four chairs. A quick search shows plenty of choices from $100 on up. Let’s settle for $150.

A family eats two to three meals on this table six to seven days a week for 40 or 50 years. A quick napkin estimation shows about 30,000 uses. That’s about half a penny per use. Pretty darn inexpensive for a valuable experience.

Friends drink coffee, wine, or something at the table another three or four times a week. Children do their home work three days a week, nine months a year for twelve (or more) years each. A work-from-home professional sits at the table when all these other uses are not in progress.

Laughter, sorrow, love, angst, staring into another’s eyes, grasping hands in desperation.

All of this over a table.

The per use cost is now less than a tenth of a penny each. Is there a better value anywhere in the home or in a person’s life?

In 1983, my wife and I married. Her parents bought us a table and chairs. It sure seemed expensive to me at the time. Silly me. It was one of the best values in my life and it continues to be so.

by Dwayne Phillips

Aha! I was sort of right. Now companies have teams of folks hunting down stupid in their systems to keep the world safe from stupid.

Way back in 2015, I wrote a blog post about being a stupid hunter. This is not a hunter who is stupid, but a hunter whose job is to seek out and eliminate the stupidity in systems and organizations.

Well, this week I read of OpenAI (fill in the name of any AI-focused company) with its “red team” that hunts the answers its ChatGPT provides. When it finds “toxicity, prejudice and linguistic biases in the model” they blow their horns and bring in the rabid dogs or something terrifying to rid the toxicity, prejudice and linguistic biases from the model. What’s a “model” anyway, and is the model afraid of rabid dogs or something terrifying?

Let’s say it simply: they seek out and eliminate the stupidity in their system.

Aha! I knew it. Well, I can claim I knew it. Then again, their is so much stupid in systems and organizations that stupid hunters will always have jobs called “red team” and some such polite thing. No one, aside from me, wants to wear a shirt that proclaims STUPID HUNTER.

by Dwayne Phillips

The passing of time is usually insufficient for a person to move from task to another. Sorry. Hopes and wishes are not plans.

I have seen this time and again. A person is hired for one job with the qualifications needed for that job. A few years later, that person is moved into another job. Other qualifications for the other job?

Two questions:

1. What did we hire this person to do in the first job?
2. What has changed in this person to qualify them for the second job?

Well, this person has matured five years and … or this person has watched me do the second job for five years and …

Sorry. Insufficient answers. This person is bright, teachable, willing to learn, hard working (don’t ya’ just love when that last one is tossed in?), has never been specifically trained for the second job, and is therefore unqualified for it and… What usually happens is lots of innocent bystanders are harmed (some irrevocably) while the person learns the second job on the job.

Hopes and wishes are not plans. We sure hope and wish this person has picked up enough to do the second job. That would sure make our lives easier if our hopes and wishes would come true. And, sometimes hopes and wishes come true. It is unfortunate that is about 1% of the time.

Sorry. We know better. Let’s do better. If we are “grooming” a person for that second job, let’s do it right.

by Dwayne Phillips

It is important to remember whose words we are discussing. If you mentioned something, we are discussing your words. If I mentioned something, we are discussing my words. There are ways to back away from our words.

I read the words from the PowerPoint. They were wrong. Simply wrong. 1 + 2 = 4. Wrong. When I questioned the words, arguments began. Finally, a third person said, “Wait. Those are your words on the PowerPoint. No one made you put them there. You put them there. We are merely discussing them.”

In that instance, the words originated with that other person. In other cases, the words originated with me. No one made me put the words on the PowerPoint. I put them there. They are my words. If they are wrong, I made the mistake.

It is possible to step away from my words. It is possible for you to step away from your words.

“You are correct. My words are incorrect. I was mistaken. Let’s find the correct words and move forward.”

It really is that simple. Own our words. Use them or toss them and find better ones.

by Dwayne Phillips

Someone is paid to lead the people and manage the work. They are paid to answer, “Yes, but how are they going to do that?”

Work can be difficult to accomplish at times. And there are times when you say, “This isn’t good enough. It must be better.”

A well-meaning person replies, “Yes, but how are they going to do that?”

The answer is simple, “So-and-so is paid to lead the people and manage the work. It is their job to figure it out and do it.”

Yes but, but, but…

Sorry. There are no “buts.” Figuring out how to accomplish difficult work is someone’s job. They are paid to do that job. If no one needed to figure out how to accomplish the difficult work, it wouldn’t be difficult and we wouldn’t pay someone to do it.

Sometimes this simple little exercise sounds trite. Sometimes it sounds mean. Sorry. This is the nature of human endeavors. Leaders of people and managers of work are called upon to do these things. Let’s get to it.

by Dwayne Phillips

There are things I don’t want to know. Knowing those things means I have to do some work that I just don’t want to do.

Life is full of headaches. There are more than enough headaches for each day. I don’t need any more headaches.

Then some well-meaning person walks in the door with, “I was just checking on the ABC regulations regarding ABC things, and we are supposed to be tracking something and filing weekly and monthly reports on something regarding ABC things.”

Huh? That’s six more headaches. I don’t need six more headaches. Can’t we ignore this?

“No,” replies the well-meaning person. “I’m lucky to have caught this early enough so that we don’t get into real trouble on ABC with the ABC regulations and regulators and investigators and …”

Lucky? Lucky to have caught this? No. That is not lucky. That is another big headache.

Now that I have been dutifully informed, I know a thing I don’t want to know. I know that I have another headache and another thing that must be done today. The day—and all the resources of the day—are the same. The list of things to do this day is longer.

There are things I don’t want to know.