Working Up

by Dwayne Phillips

Sometimes, the better thing to do is let people do what they think is best or at least what they think is better on any given day.

“We are a fill-in-the-blank organization. We use best fill-in-the-blank practices as described by fill-in-the-second-blank who described fill-in-the-blank in the seminal blog post on fill-in-the-blank,” said an earnest leader of a team when questioned about their practices.

There are many worthwhile practices. Agile, SCRUM, waterfall, prototyping, documenting, not documenting, etc.

It is worthwhile for an entire organization of thousands of people to use the same practices on thousands of days in thousands of different situations.

It is worthwhile to have a shared and solid foundation. Let’s all start at the beginning with our feet solidly planted on something solid. Life is tough enough to attempt it while juggling 15 balls when off balance.

And let’s add one more worthwhile practice to this list of worthwhile things: let’s let competent persons do what is best, or at least better, on any given day in any given situation.

Well, yeah, but, you see… Of course that last worthwhile practice is a bit scary. What if the competent persons are wrong on this day in this situation? Things would be messed up and we would have a mess and we would have to fix the mess just to get back to zero. What a mess.

I hired these persons to decide what is best or at least what is better. Did I hire the wrong persons? Wait, don’t pin this mess on me. Wait, I was put in a position of hiring. Yes, this mess flows up to me.

And now I have to decide what to do with this mess.

This mess is today’s situation. Let the competent person do what is best, or at least what is better, given this mess. And that includes me. I can do what is best, or at least what is better, given this mess.

If you wanted guarantees when you became a manager, sorry to disappoint you. There are none. We are working with people, not levels of liquid in bottles in manufacturing. There are no lower-control and upper-control levels with people. We do. Some days better, some days worse.

Let’s observe, think, and do what is best, or at least what is better.

by Dwayne Phillips

Once again, thinking at one layer higher than usual may bring insights that prove effective.

There is competence: someone is able to do something well. They know the topic, they know the skills, and they apply them all. The adjective “well” is used often and truthfully.

Then there is meta-competence: this has something to do with being competent with and about competence.

I struggled to find this concept for several decades. I worked in government. Employees were often judged for their abilities. The simple idea being that persons judged to be more competent would be promoted to positions requiring more competence. The simple idea often fell broken on the rocks of reality.

The problem I often found was, “Who are the judges?” I frequently met judges who were not competent. They couldn’t write a sentence, but judged others on their ability to write. They had not read a book or paper on technology in ten years, but judged others on their technical abilities.

We lacked people who were competent about competence, i.e., we had little meta-competence.

At this point, I am supposed to provide the solution to this meta-competence problem. Well, that is the problem. Someone has to judge competence with competence to decide competence, or something like that. A high-level manager needs to hire a hiring expert so that competent people are hired. How does the high-level manager know whom to hire? Who is the hiring expert, i.e., the person with meta-competence. How is the high-level manager supposed to be competent at meta-competence?

At some point, someone has to do the first task and do it well.

Advice? Do it quickly. Judge the results quickly. If the first step was a mistake, change the first step quickly. Time doesn’t always heal all wounds.

by Dwayne Phillips

Take care when confusing lab projects with real products. Both are good, but they are not the same.

In all fields of endeavor, we have lab projects and real products. Both can be good. They, however, are not the same. Sometimes product managers and marketers confuse these at their peril.

Students in graduate school have lab projects. They experiment and write their results. Advanced degrees are awarded. Great stuff.

Sometimes, engineers et al have lab projects at work. They experiment and write their results. In wise organizations, the results carefully and painstakingly become real products sold to consumers. In not-so-wise organizations, the results are hastily declared to be products and sold to consumers who then wince in pain and write bad reviews on Reddit et al.

Experiments don’t produce products that are ready for market. Sure, this happens once a century and it works. The rest of the time, it doesn’t work. Real products take time to refine through engineering, testing, redesign, testing, rework, testing, test marketing, etc. Those things can drive people nuts with the time required. Those things certainly raise the cost and resulting price of the final products. Those things also reduce the chance of horrible reviews on Reddit.

Results of lab projects can be quite informative. Information about what is possible is not a consumer-ready product. We can do better than this.

by Dwayne Phillips

Games are simulations of real life. Some games simulate real life better than others. The key is to connect the simulations to real life in a meaningful manner.

Games are simulations. The Monopoly game is supposed to simulate business, commerce, and acquiring wealth or something like that. Tick-Tack-Toe simulates spatial logic and the wisdom and folly of wasting time playing a game that no one can win or something like that. There seems to be a million video games out there that simulate this and that or something like that.

As a simple example, I watch my grandson playing something on his phone. I have no idea what game he is playing or what it is supposed to simulate, but I see his reaction to the game. As always, the reaction is more important than the event. My grandson reacts with joy, angst, anger, frustration, etc.

Whatever game he is playing is simulating real life quite well. Real life is full of joy, angst, anger, frustration, etc.

The key is connecting these reactions to the game to reactions in real life. Having to wait longer than expected and the reaction of frustration is like being frustrated while playing a video game on a phone. The reaction of frustration is the same regardless the source event. Learn from the simulation.

Not sure what is causing troubles with you and your team at work? Play a game, i.e., use a simulation for an hour. Observe events and reactions among the team. Learn. Apply the lessons from the simulation to the real-world work.

We can do better, and simulations can help us if we make meaningful connections to real life.

by Dwayne Phillips

I share an idea for a manual writing instrument that I had several decades ago. Today’s technology might be ready for it.

Several decades ago, I had this idea for a pencil, pen, or white board marker. I shared the idea with a few smart folks. Nothing became of it as the technology wasn’t available. Today, perhaps, someone can make this.

Spellcheckers in word processors are pretty good. The software that guesses what word I want to type on message apps is pretty good. This generative AI stuff is pretty good. The software seems to know what word I want and how to spell it correctly.

Okay, I want this to help me when I am writing with pen, pencil, and especially when writing on the white board in a meeting attended by important folks.

The pen monitors the movements as I write. Those movements translate to the word I am attempting to write. Let’s see, is the letter T double at the end or just single? How many times does the letter S appear in assessments? I can’t remmember.

I want the pen to beep when I have probably misspelled a word. Then the pen’s display or speech tells me how to spell the word correctly.

There it is folks. Someone go out and make this manual writing device. White board writers everywhere need it.

by Dwayne Phillips

Here is one way to begin what is likely to be an unpleasant conversation. It lessens the unpleasantness and moves toward a more productive conversation.

Let’s begin with, “John, I am upset.”

(Substitute the other person’s name for John. That was the most alias of names I could find.)

There, said the ugly part first. I am upset. That describes my state of mind and the feelings burning inside me. I don’t think clearly when I have those burning feelings. Perhaps John will understand some of the stupid things that are about to come out of my mouth.

Excuse? Yes, declaring upset is sort of setting the stage for stupidity and requesting forgiveness beforehand.

Wise? Perhaps. Let’s put this out right now. This isn’t going to be a fun-and-games chit chat. I am about to say some unpleasant things and will say them in an unpleasant manner. I am saying them to you because I respect your ability to handle unpleasantness and work through it with me.

Perhaps giving John a copy of this little post to read will ease the conversation. Perhaps reading it aloud to John will do the same. It may calm my upset and help me to speak using better words and tone.

Regardless, let’s get into this. I am upset, and I’m about to hit you with the reasons.

by Dwayne Phillips

Loving and caring people want to help others. So we put one more thought into the writing and one more statement into the speaking. Delete those.

Here are tips on editing writing that don’t need much thought, but hold true about 98.6% of the time:

• Delete the last word of a phrase
• Delete the last phrase of a sentence
• Delete the last sentence of a paragraph
• Delete the last paragraph of a page
• Delete the last page of a chapter
• Delete the last thing of a bigger thing

Get the idea?

Loving and caring people want to help others. We want to be sure the other person understands, so we add just one more little bit to what we have already conveyed. That extra little bit will help the other person an extra little bit and that will make all the difference.

Sigh. We are our own worst enemies. Brevity often leads to clarity. Delete that extra little bit that we added to make sure the other person understands. The other person understands already. That extra little bit is often more than a little bit, and it doesn’t help.

by Dwayne Phillips

Large Language Models appear to have all sorts of problems. I wonder why companies that build such don’t hire adults to help build them.

I recently learned that large language models (LLMs) have “sleeper agents” in them. Given some inputs, the LLM starts doing crazy things that it shouldn’t do.

Gosh.

LLMs are pieces of simple software that trigger large matrix multiplications and pick words based on the results of multiply, add, subtract, divide, etc. LLMs don’t “act” strange.

What is happening? Here is the simple answer. The well-meaning folks that build the LLMs are not testing them sufficiently before releasing them as products. The input space, all those possible combinations of words, is huge. The builders don’t exhaust the input space and correct mistakes. “Well,” the logic goes, “We tested this thing using a million or so test cases. It’s pretty good. Let’s release it.”

Some business manager looks at the numbers, decides that enough is enough, and releases the product to recoup some cost $$$. Well, that makes business sense.

Engineering? Are you kidding? Systems engineering? Are you kidding? Testing? Are you kidding?

Do these companies hire any old engineers who understand how to build, test, and release products when they are ready? Where are the adults? Where are the people who are old enough to have seen bad things happen?

There is much value in having people who have experienced tragedy on your development team. Tragedy is no fun. It is, however, real life. It does, however, sear scars into the heart and mind. Have such people on your team. Pull them aside and talk with them. Learn what could possibly go wrong and the results. We can do better.

by Dwayne Phillips

We have yet another problem with machine learning. This one, however, has an easy solution.

Several years ago I wrote of a fundamental problem with machine learning. I guess we have yet another problem being called “poisoned” machine learning.

With poisoning, someone knows that I am about to “learn the machine” something using a large dataset. That poisoner knows where the dataset is and they put data into my dataset. Their data will mislead my “learned machine” into making mistakes like the poisoner wants it to make.

Alas, garbage in and garbage out, again.

Solutions? Why in the world is my dataset out there where others can poison it? Don’t I know anything about data? Can’t I keep my mouth shut about what I am doing? What is wrong with me?

Solutions? Don’t I know what data I am using to “learn my machine?” Why am I using bad data? What is wrong with me?

I guess this is all more complicated than I understand. Otherwise, the solutions wouldn’t seem so obvious to me. Perhaps I am just too old to “get” all this.

by Dwayne Phillips

In our personal lives, we often take all the risk without asking others to share it. The basic, but rarely asked, question is, “What does each person have to lose?”

I am sitting here sipping coffee pondering what to do. Someone is asking me to wait, and wait, and wait. Perhaps they will show up, perhaps later. If they don’t show up…sigh, I have wasted the morning. That is what I have to lose, the entire morning.

What does the other person have to lose? Nothing. They are late because they are busy doing something else that is productive and rewarding for them.

I have my morning to lose; they have nothing to lose. I am risking my morning; they are risking nothing.

The stakes are low. Let’s go on.

Now let’s multiply the stakes. Instead of risking a morning, I am risking several months salary. The other party is still risking nothing. Uh, we are not sharing risk. I am bearing all the risk.

Let’s negotiate this a bit. Let’s make the risk carried by each party explicit and discuss ways to even the risk a bit. If the other party is not willing to share risk, the conversation ends. Nice knowing you, goodbye.

Sometimes carrying all the risk myself is worth it to me. I really want to have that meeting or that job. Sometimes carrying all the risk myself is not worth it to me. I am able to walk away.

Suggestions: (1) If I walk away, do so on the friendliest terms I can find. I value the other person. (2) Openly discuss risk and risk sharing.

The risk-sharing discussion is between two respectable and respected parties. Both parties deserve consideration from the other. Sharing risk is a respectable and respected thing to do. Speak with respect in respectable terms, but speak.

There is little reason to carry all the risk.