The Real Advantage of AI? It Saves You Time (Premium)

I’ve told this story before, but it’s perhaps instructive. Years ago, I stayed with friends in France for a week and after pulling a ThinkPad out of my bag, my friend’s wife commented that she used a Mac because of all the bugs and problems with Windows. How long ago was that, I asked? It was seven years ago, she said. Then your opinion is out-of-date, was my reply, Windows has improved a lot since then. (And I wasn’t trying to be a jerk about it, but left unsaid was that I also used a Mac and very much preferred Windows, as I still do today.)

I don’t point out this story to embarrass my friend’s wife, she’s great. Like all of us, she’s the product, in part, of her life experiences. The issue, of course, is that many of our experiences are out-of-date or even non-existent. And yet, many of us still have opinions–very strong opinions, sometimes–about topics we’re perhaps not well versed on.

AI is like that. Hell, AI is that.

I can’t tell you how many times I’ve been told, unsolicited, that someone has “never found a use for AI,” or that all AI is does is “regurgitate useless information,” “produce garbage,” or whatever. The details never matter. These opinions are voiced mostly so that person can receive calming affirmation from whomever they’re accosting. And I suspect that most of these utterances, whether they’re about AI or any other topic–Windows 11 on Arm or the new Outlook, anyone?–are indeed met with a vague or even explicit agreement. Sometimes just so they can get past the rant and move on. And sometimes, sure, maybe they do agree. I may agree.

But these opinions about AI, mine included, are as out-of-date as my friend’s wife’s opinions about Windows and PCs. The difference between now and then is that AI is evolving much faster than Windows and the PC did. And our opinions aren’t keeping up. Not that that’s ever stopped anybody. It’s a grand American tradition, especially among us middle-aged white guys, to opine on any topic regardless of experience, and to expect the audience to nod along silently and admire your genius. I know this tradition isn’t restricted to America. But we are uniquely good at it. He says, trying to avoid that mirror.

Here’s the thing.

Like virtually everyone, I struggled with understanding AI when it exploded out of the gate in the wake of Microsoft’s first Copilot (then Bing Chat) announcement in early 2023. I went through what I jokingly call the seven stages of grief in trying to understand the impact that AI would or would not have on our lives. (The key to this joke being I can rename any or all of the steps each time I mention it.) I watched in astonishment as conservative Microsoft, before then a slow-moving and predictable cloud-focused company under Satya Nadella, suddenly turned on the afterburners and got aggressive again. I have never seen a turnaround quite like it, and I’m still not sure what to make of the company we now deal with. It’s entirely different in many ways to the pre-AI Microsoft, more chaotic and less predictable.

I also have to remind myself–and maybe you–that it’s also a demonstrably richer company today as a result. Microsoft’s market cap just before its AI push at the start of 2023 was “just” $1.7 trillion, give or take. Today? It’s over $3 trillion. Not quite double, but close enough. (And Microsoft’s market cap did hit double that $1.7 trillion figure in July 2024.) You may not like what Microsoft is doing, to the products or services you use every day and rely on, but there’s your rationale. What it’s doing is working.

But maybe it’s not working for you. Sometimes it’s helpful to consider that maybe you are the problem. By which I mean, I am the problem, too.

Consider the recent story about the Linux kernel contributor quitting in a huff after a kernel check-in he wrote in the Rust programming language was rejected by the Linux core kernel maintainer. Rust code has been accepted into the Linux kernel since 2022, but the maintainer felt that the code he submitted–a driver abstraction that would allow Rust code to call a C-based memory access routine–was more of an added burden than a benefit. This event was publicized as old-school developers not wanting to deal with modern Rust code, but that’s not all that happened there. And when Linux Torvalds suggested to the offended contributor that, “maybe the problem is you,” the guy quit. And then left the Asahi Linux distribution that he had previously founded.

Put simply–and leaving aside any opinions about Torvalds, the contributor, the maintainer, or the entire Rust vs. C debate–this was just personalities and ideals clashing. And that is very much what’s happening with AI. It’s not really about the technology or its efficacy at content creation, that’s obviously real and working incredibly well. (Just ask any developer about giving up GitHub Copilot to understand how real it is.) It’s that the dividing line here is really between those stuck in the past–peak word processing was in 1997, or whatever–and those who can admit that they aren’t the apex of human intelligence and efficiency and maybe could use a little help.

I can feel you pushing back. Give it a second. The trick here, literally, is admitting that we can all use a little help. And we can all use a little help. Somewhere, somehow.

In last week’s Ask Paul, I told a story about my wife’s first meaningful use of AI for work. She hasn’t, doesn’t, and won’t use AI for writing, meaning the final act of creation and this thing that you submit to a boss or client. But she did find some ways that AI worked for her. Meaning, more specifically, that it saved her time, which gave her more time to focus on that thing she does, which is writing. And this, to me, is key. It’s the bit that AI doubters, myself included, seem to miss or ignore. We may be an expert at whatever. But that doesn’t mean AI can’t still be useful to us elsewhere in life. And that in incorporating AI elsewhere, it can perhaps free up time to spend on the things we are good at or prefer doing.

Before my wife’s specific and real example, all I could come up with were some pseudo-examples from my own experiences. Thinking about what I’m good at, and what I’m not good at, I can see the potential early on for a few time savers of my own. For example, early every year, I write an annual recap of PC sales. (Here’s the latest example.) When I write this article each year, I go into my archives, find the Excel spreadsheet I had used the previous year for the chart at the top, and make a copy of it. Then I edit that by adding the past year’s PC sales data so I can update the chart, which graphically compares PC sales each year dating back to 2006.

Every time I do this–and I mean, every single time–it’s like it’s the first time I’ve ever used Excel, or made a chart. I can never remember how to get it to use different colors for each year in the chart. In fact, I often give up. If you look at the wrap-up article for 2023 PC sales, for example, you can see that the bars for each year are all the same color. That year, I didn’t figure it out. And this is exactly the type of thing AI is good for. Copilot, or any other AI, really, could have helped me solve that problem. The thing is, I wasn’t ready a year ago to take that step. I was still operating in this mode, in which I knew better and could figure it out. Even though Excel is not a strength or an interest of any kind, and all I ever want is to get the thing done and then not ever see it again.

We all do this. It’s a trap.

One of my big efforts this year can be found in the Online Accounts (2025) series, which was thrust upon me when Google/YouTube lost its mind and locked me out of my Thurrott.com YouTube account. Faced with the unexpected need to back up what could be as much as 3 TB of data, I was reminded of the time 20 years earlier when my 30 GB data drive failed. Which is interesting because that event shaped my approach to data backup and sync, and though the details have changed thanks to improvements in technology, I have been on that path ever since. You get religion when things go wrong, I opined decades ago. And in that case, I had gotten the backup religion because I lost data.

That’s human nature. But there are disasters and there are miracles. And either one can be the inspiration for change. And sorry, guys, but AI is a miracle. It’s a million tiny miracles, and it’s up to all of us to figure out where and how we’re going to allow those miracles to help us. And, again, I can feel you pushing back. But hold on, please. Because we’ve all heard the adages about time being the one thing you can never get back. I have, in fact, written about this exact topic, this notion of constantly wishing I had more time. And in the end, that is exactly what AI can do. It can give us more time.

Instead of bemoaning that AI is being added to all the digital tools we rely on, from lowly Notepad all the way up to the video generation capabilities now available in Adobe Premiere Pro and Firefly, maybe it’s time to think about all the ways we could use that functionality. Or ignore it.

On this morning’s First Ring Daily, Brad–old and crotchety before his time, something I’ve always found amusing–complained that he had pasted some content into Microsoft Word and a “Paste with Copilot” UI popped up. He found this infuriating to some degree, but I think it’s important to remember in these situations that maybe you know how to use this tool better than roughly 95 percent of the total audience. And is it really so outrageous that Microsoft Word would pop up a UI related to creating … words? That Copilot UI could be used to rewrite the text you’re pasting in to the app. I suspect many people would find that useful. I write, as a professional writer who would never use that tool.

My lack of desire or need for writing help is just me being me, I guess. But not wanting other people–inarguably most other people–to benefit from this functionality is selfish and shortsighted. I do at least have the empathy to understand that others can benefit from things I may not need. And the key to that empathy is being honest to yourself, and about yourself. Because you can use help, too. Somewhere in your lives.

There’s a lot to this. Burnout, which is understandable in anyone middle-aged or older, is part of it. I feel it. But I also catch myself babbling excitedly to my wife about some tech story or topic, and have to sometimes remind myself that this maybe doesn’t interest her and to reign it in. This literally happened yesterday at lunch, and I suddenly caught myself, apologized, and stopped talking. My wife has enough empathy herself to realize this was weird for me, and she tried to save it by feigning some interest in whatever it was. And OK, thanks. But one takeaway from this is that here I am, 58 years old now and with grown children, 30+ years into a career that frankly could have started even earlier if my head had been screwed on straight, and I can still be perhaps overly enthusiastic on this industry that I care about so much. This feels vaguely positive to me. And I’m perhaps well known for being negative, somehow. I call it realistic, but I guess I would.

Generational distrust is difficult to overcome, too. Mary Jo and I gave a talk several years ago and were asked to focus on how those middled aged or older like ourselves should handle life in a world being flooded by younger, less experienced coworkers. After the talk, the first question was from a younger person who jokingly started off with, “so, aside from the fact that you hate young people…” He was kidding, but that wasn’t really the point. Our point was that experienced people should be proud of their expertise and should use it to their advantage. And part of doing that involves not just saying no to every potential change. No one gets everything right, but the only benefit of age, perhaps literally, is the knowledge and perspective that comes with it. There is nothing sadder than seeing someone squander that.

At least try.

I’m going to try. I mean, I am … trying. I still don’t see a use case for using AI for literal writing–I am a writer, after all–but I would like to (try and) examine ways in which it can save me time elsewhere, similar to what my wife did (and is evolving). I have resisted using GitHub Copilot too much with this year’s .NETpad project, but I have no problem Googling for answers to my many questions and waste a lot of time pouring through irrelevant results. This is going to require effort on my part.

Since I just mentioned it, I decided to use my own example above to see whether AI could have saved me some time in making that bar chart. Predictably, it could: Copilot (in Excel), ChatGPT, and Gemini all made short work of this task. Interestingly, I had to ask each to make small adjustments to the original output. For example, with ChatGPT, the year labels were written with decimal points initially (2025.0, etc.). So I asked it to fix that. (Gemini’s colors were too similar, so I had to similarly ask it for a fix.) But that was in itself instructive. It’s very natural to fall into a conversation with these AIs as you work to fine-tune the results. Each handled this interaction similarly, “remembering” the context of the conversation. In so many ways, this is nothing like using Google Search.

As I was writing this, my watch burbled on my wrist: I had a meeting at the top of the hour. Because I was writing, there was a good chance I would zone out and just keep writing past the beginning of the meeting. I know myself that well, at least. And so I picked up my phone to make a reminder. There are probably dozens of ways to do that on my phone, but this time I went in a non-traditional direction (for me). I tapped the Gemini icon on my home screen. Using its voice feature, I asked it to remind me about the meeting 5 minutes before it started. “Got it,” the AI responded, “I’ll remind you at 9:55 am that you have a meeting.” (It added this item to Google Tasks, which is what I would have done using the Google Calendar app, normally.)

Am I describing yet another way to make a reminder as a miracle? No. But this isn’t about a reminder. It’s about the dozens or even hundreds of tiny ways in which all of us can use AI, right now, to save time, in turn improving our lives. We need to get over that hurdle, stop bitching about nothing, and make it work for us. Many of us. Me among you. Which is part of the point. I’m right there with you. But it feels silly to stand in the face of progress when it’s increasingly obvious that that’s what this is.

I’m going to try. But first, I have a meeting.