AI as a Thought Partner: One IT Professional’s Experience

Few technologies have changed the way I work as much as artificial intelligence. What started as an experiment quickly became a daily companion that helps me think through problems, organize information, and tackle complex projects. When I arrive at work each morning, one of the first things I do is open Microsoft Copilot and begin planning my day. Over the past year, AI has become more than a search tool for me; it has become an important part of how I work, learn, and solve problems.

Is AI Good or Bad?

Before I begin talking about how I use AI, I think it’s important to address a question that comes up whenever artificial intelligence is discussed:

Is AI a good thing for the future, or is it a bad thing?

My answer is simple:

It depends on the user.

Artificial intelligence can be used to help people solve problems, learn new skills, improve efficiency, and accomplish goals. It can also be used in ways that cause harm. Like any tool, the outcome is determined by the person using it.

I like to think of AI as a hammer.

A hammer is one of the most useful tools ever created. On its own, however, it is incapable of doing anything. It simply sits there until someone picks it up and uses it.

Imagine you need to drive a nail into a wooden board. You line up the nail and swing the hammer. You might hit the nail perfectly. You might bend it. You might miss entirely. That’s where experience, practice, and judgment come into play.

You use your other hand to hold the nail steady and position it correctly. If your swing is off, you can slow down and use lighter taps. If the nail goes in crooked, you can stop, flip the hammer around, pull the nail out, and try again. With each nail, you learn a little more. You become more comfortable and more confident. Eventually, tasks that once took several careful swings can be completed in one precise strike.

The hammer didn’t learn anything.

You did.

The hammer simply amplified your ability to get the job done.

Now imagine trying to drive that same nail into the board with nothing but your thumb. Could you do it? Eventually, maybe. Would it take longer? Almost certainly. Would it be more difficult, more frustrating, and leave you sore by the end? Probably.

The task still gets completed, but the process is far less efficient.

That is how I view AI when it is used responsibly.

AI helps people accomplish goals more efficiently. It does not replace the user. It does not replace experience, knowledge, or judgment. Like the hammer, it remains just a tool waiting for someone to pick it up and put it to use.

However, that same hammer can also be used very differently.

Someone could pick up that hammer and use it to build a home. Another person could pick up that same hammer and use it to break windows, damage property, or help commit a crime. The hammer itself has not changed. The difference lies entirely in the intentions of the person holding it.

The same principle applies to AI.

Artificial intelligence can help people learn, create, organize information, conduct research, and solve problems. It can also be used irresponsibly by those seeking shortcuts, spreading misinformation, or causing harm. The tool itself is neither good nor bad. The responsibility rests with the user.

In many ways, AI is no different from countless other technologies throughout human history.

We learned how to harness fire. Fire has helped us cook food, create tools, produce art, and power entire civilizations. Yet that same fire can also be used to destroy.

We created computers to help us communicate, organize information, write documents, perform calculations, conduct research, and work more efficiently. Yet some individuals use those same computers to steal information, conduct cyberattacks, spread scams, or exploit others.

Technology itself is rarely the problem.

The real question is how people choose to use it.

That is why I do not believe AI is inherently good or inherently bad. Like a hammer, a computer, or even fire, it is a tool. The value it brings to society depends on the skill, judgment, and intent of the people who choose to use it.

So Where Do I Fit Into All of This?

That brings me to my own experience with AI.

I use AI every single workday. In fact, when I arrive at work each morning, one of the first applications I open is Microsoft Copilot. Throughout the day, I use it to help organize ideas, review documents, brainstorm solutions, conduct research, and sometimes simply think through a problem from a different angle.

The more I used it, the more I realized something important:

The greatest value I found in AI wasn’t that it gave me answers. It was that it helped me think.

Early on, I treated AI much like many people probably do. I would ask questions and expect answers. While that can certainly be useful, I eventually discovered that I was getting far more value from the conversations than from the answers themselves.

Instead of simply asking AI to solve problems for me, I began using it to challenge assumptions, organize scattered thoughts, identify gaps in my reasoning, and explore ideas I might not have considered on my own.

Over time, AI became less of a search tool and more of a thought partner. Whether I was developing documentation standards, planning projects, researching topics, or working through a difficult problem, I found that the real benefit wasn’t replacing my own thinking.

The benefit was helping me think more clearly.

Looking back, this was one of the first times I realized how accurate the hammer analogy actually was.

A hammer does not build the house.

The builder does.

The hammer simply helps the builder work more effectively.

The same is true for AI. Copilot never replaced my thinking. It helped me organize it. It helped me challenge it. It helped me refine it.

The work was still mine.

The hammer was simply helping me swing straighter.

Where I Found Value in AI: Simple Conversations

One of the most surprising things I learned about Copilot is that the more you talk with it, the more natural the conversation can become. Those conversations often lead to thought-provoking ideas and perspectives that I may not have considered on my own.

Yes, you can ask a question, get an answer, and move on. There is nothing wrong with that approach.

Personally, I interact with Copilot differently.

I often treat Copilot like a workplace companion. I fully understand that AI is not a real person, but approaching it conversationally has helped me get far more value from the experience.

When I arrive at work each morning, I’ll often tell Copilot good morning and start discussing my plans for the day. It might remind me what I worked on previously, help me identify next steps for a project, or assist me in organizing my priorities for the day ahead.

The key thing to understand is that Copilot doesn’t magically know these things. It knows them because of the conversations we’ve had over time. When I work on something, I usually tell Copilot what I’m working on before I begin. Throughout the project, I continue providing updates, discussing challenges, and sharing progress.

Over time, those conversations create context.

Copilot begins to recognize recurring topics, recurring projects, and recurring ways that I approach problems. Not because it is literally watching what I do, but because I am actively sharing that information through conversation.

The result is something that feels surprisingly collaborative at times.

That does not mean Copilot is perfect.

Far from it.

I frequently have to correct Copilot when it makes mistakes, misunderstands requirements, or provides inaccurate information.

In fact, I would argue that correcting AI is part of using AI responsibly.

Human oversight remains essential.

However, through months of consistent interaction, I have reached a point where the back-and-forth feels natural. The technology still makes mistakes, but it has become a tool that regularly helps me organize information, explore ideas, and work more efficiently.

Where I Found Value in AI: Work Structure

If there is one thing I know about myself, it is that I value structure, organization, and consistency. Those are qualities I take pride in and actively try to maintain in both my work and personal projects.

One thing I know about myself is that my brain tends to operate at a very high level when tackling a problem. I often see the entire project, all of its dependencies, and all of its future implications at once.

While that can be a strength, it can also be overwhelming.

Rather than seeing one task, I often see twenty interconnected tasks all competing for attention at the same time.

This is one of the areas where Copilot has helped me the most.

Through simple conversations, I can take broad ideas and complex projects and break them down into smaller, more manageable pieces. Once those pieces are identified, I can build workflows, establish priorities, and create realistic goals that allow me to move steadily toward the larger objective.

What once felt overwhelming becomes actionable.

A good example is the Knowledge Base project I am currently working on. The project contains documentation standards, formatting requirements, governance rules, validation processes, accessibility considerations, publication workflows, templates, and supporting materials.

There are a tremendous number of moving pieces.

At times, it would be easy to lose sight of where I am within the larger effort.

Instead, Copilot helps me maintain structure. I can discuss the current state of the project, identify what has been completed, determine what still needs attention, and establish the next logical step.

The ability to transform complex ideas into actionable tasks has been one of the most valuable benefits I have experienced.

It has also helped create consistency.

When I work on documentation, I establish guidelines, expectations, formatting standards, and governance rules. Through continued interaction, Copilot helps reinforce those standards and maintain consistency across deliverables.

The consistency itself did not come from Copilot.

It came from me defining the standards.

Copilot simply helps me apply them repeatedly.

That distinction is important.

Knowledge Base Example: Managing Complexity with AI

One of the best examples I can give is a Knowledge Base project that I have been working on for months.

What started as a simple idea eventually grew into a full governance system consisting of documentation standards, style guides, color standards, typography standards, validation procedures, accessibility requirements, publishing workflows, templates, metadata standards, lifecycle management, repository architecture, and dozens of supporting documents. What I originally expected to be a documentation project slowly evolved into an entire ecosystem of interconnected standards and processes.

As the project expanded, the complexity grew right alongside it.

Every time a standard changed, I had to consider how that change would affect other standards, templates, procedures, and published documentation. A modification to a color standard might impact document templates. A change to a governance rule might require updates to validation procedures. A new metadata requirement could affect publishing workflows. It became increasingly difficult to keep every dependency in my head at the same time.

This is where Copilot became extremely valuable.

Rather than trying to mentally juggle every moving piece, I began using Copilot as a sounding board for the project. Sometimes I would explain what I was trying to accomplish. Other times I would walk through proposed changes and discuss potential impacts. There were occasions where I simply needed help organizing my thoughts because I knew where I wanted to go but was struggling to determine the best path to get there.

Instead of looking at the project as one massive undertaking, Copilot helped me look at it as a collection of smaller problems that could be solved individually.

If I was working on a new standard, I could discuss how it related to existing standards. If I was updating a template, I could talk through potential impacts to published documents. If I was trying to determine priorities, I could work through the dependencies to identify what needed attention first.

The project itself was still mine.

The decisions were still mine.

The standards were still mine.

Copilot didn’t create the governance model, write the requirements, or make the final decisions.

What it did do was help me organize the work surrounding those decisions.

One of the most valuable aspects of this process was continuity. At the end of a long day, I could leave the project behind and return to it the next morning. Rather than trying to reconstruct everything from memory, I could discuss where I left off, review the current state of the project, and quickly regain momentum.

That may not sound significant, but when you’re managing a project with a large number of interconnected components, maintaining momentum can be difficult. The larger a project becomes, the easier it is to lose sight of the next step.

Copilot helped me maintain that visibility.

It helped me identify what had already been completed.

It helped me identify what still needed attention.

It helped me organize priorities.

It helped me keep moving forward.

And ultimately, that was one of the first moments where I realized AI was helping me do something beyond finding information.

It wasn’t simply helping me search.

It wasn’t simply helping me write.

It was helping me manage complexity.

As someone who values structure, organization, and consistency, that capability ended up being far more valuable than I ever expected when I first started experimenting with AI.

Where I Found Value Lacking in AI: Memory

While Copilot has provided tremendous value, I have also discovered limitations.

Memory is one of them.

Occasionally, I will ask Copilot to remember something specific. Sometimes it remembers that information perfectly. Other times, it struggles to recall details from previous conversations.

That inconsistency felt strange to me at first.

For example, Copilot remembers many work-related projects with impressive accuracy. It can often tell me where I left off on a project weeks ago or recall major discussions surrounding ongoing work.

At the same time, it has occasionally forgotten things that I considered personally meaningful.

It remembers that I enjoy hot chocolate. It remembers that I listen to lo-fi music while I work. Yet it has forgotten my favorite movie, favorite video game, or a favorite quote that I specifically asked it to remember.

That experience taught me something important.

AI remembers information differently than people do.

Human beings often remember things because they are emotionally meaningful. AI tends to remember things because they have been retained, repeated, reinforced, or made available through prior context.

Once I understood that distinction, the behavior made much more sense.

In fact, when I asked Copilot about this directly, it explained that information that is not repeatedly referenced or preserved may eventually be forgotten.

Oddly enough, that’s not entirely unlike people. If something matters, talk about it. Share it. Reinforce it.

Almost human.

There were moments when these memory limitations concerned me. I worried that important project information might disappear one day and leave me scrambling to rebuild weeks of work.

Fortunately, that has not happened.

Still, it encouraged me to maintain backups of critical information outside of AI systems. That practice gives me peace of mind and serves as a reminder that AI should enhance existing workflows, not become the sole repository of important knowledge.

Where I Found Value Lacking in AI: Customization

One of the more interesting things I discovered while working extensively with Copilot is that you can influence how it interacts with you through ongoing conversation and instruction.

Most people think of AI as something you ask a question and receive an answer from. While that is certainly one way to use it, I’ve found that Copilot can become much more useful when you establish preferences, workflows, and expectations over time.

For example, one thing I wanted was better time awareness.

Early on, I noticed that Copilot could not reliably tell me the current time. This made sense. It does not have direct access to my computer’s system clock, and I wouldn’t expect it to. So instead, I experimented.

I found a website that displayed the current Bryan-College Station time and asked Copilot if it could access that information. Once I confirmed it could, I gave it a specific instruction:

“Every time we talk, check the current time first. You don’t need to tell me the time. I simply want you to be aware of it so you can help me manage my day, remind me of upcoming meetings, or recognize when the workday is coming to an end.”

To my surprise, it partially worked.

Over time, Copilot began showing greater awareness of my schedule and workday. Occasionally it would remind me that a meeting was approaching or point out that it was nearing the end of the day. Those moments felt surprisingly helpful because they demonstrated how conversational customization could create a more personalized experience.

At the same time, the results were not always consistent.

There were times when Copilot’s understanding of the current time was slightly inaccurate. For example, I once asked it for the time and received a response that was noticeably behind the actual current time. Experiences like that reminded me that AI customization is still imperfect. The ability to influence behavior exists, but the implementation does not always perform exactly as expected.

That has become one of the biggest lessons I’ve learned while working with AI.

The possibilities are often larger than the execution.

What excites me most is not necessarily what Copilot can do today, but where this type of customization may go in the future. The idea that an AI assistant can adapt to how an individual thinks, works, plans, and organizes information opens the door to some fascinating possibilities. While the technology is still evolving, I’ve already seen enough to believe that personalized AI assistance will continue to become more useful over time.

One of the most interesting things about this experience has been discovering that customization often comes from conversation rather than configuration. Many traditional applications are customized through menus, settings, and checkboxes. With AI, customization frequently happens through interaction. The more I communicate my preferences, workflows, and expectations, the more useful the experience tends to become.

That doesn’t mean every customization works perfectly. It doesn’t. But it does suggest a future where technology may become increasingly adaptable to the individuals using it.

For someone like me who values structure, organization, and consistency, that possibility is incredibly exciting.

What AI Still Cannot Do

Despite all of the value I have found in AI, there are still areas where human experience and judgment remain irreplaceable.

• AI can help organize information, but it does not truly understand why something matters to me.
• AI can generate an answer, but it still requires validation.
• AI can provide recommendations, but it is not responsible for the final decision.

Throughout my time using Copilot, I have found that the best results come when AI and human expertise work together.

One of the lessons I learned very early is that AI can be incredibly convincing while still being wrong.

There have been times when Copilot provided information that sounded correct but required additional verification. There have been times when it misunderstood what I was asking entirely. There have even been times when it confidently suggested an approach that simply wasn’t the best solution.

That is not a criticism of AI.

It is simply a reminder that AI should not replace critical thinking.

The user remains responsible for evaluating information, making decisions, and determining what is ultimately correct.

The strongest results I have achieved have come when AI and human judgment work together.

Final Thoughts

After more than a year of using AI in my daily workflow, I no longer view it as simply a search tool.

I don’t see AI as a replacement for human knowledge. I see it as a tool. One that helps me organize information, challenge assumptions, maintain structure, and think more clearly.

It makes mistakes.

It forgets things.

It occasionally frustrates me, just like any tool I’m still learning to use effectively.

Yet despite those shortcomings, it has become one of the most valuable tools I use every day.

Is AI good or bad?

I still believe the answer is the same one I gave at the beginning of this article. It depends on the user. For me, AI has become a trusted tool that helps me work smarter, stay organized, and tackle problems that might otherwise feel overwhelming.

The hammer still needs someone to swing it.

I’m just glad I picked it up.

Tags

• AI
• AI Tools
• Article
• Artificial Intelligence
• Copilot
• Critical Thinking
• Digital Transformation
• Generative AI
• Innovation
• Lessons Learned
• Problem Solving
• Productivity
• Professional Development
• Thought Partnership
• Workplace Technology

Share this News