[Kevin 00:07]
So Zoe and I recorded a whole load of mini-interviews at the MVP Summit, and we thought rather than build those into a few larger episodes, we'd release those to you in short little snippets like this. So enjoy a set of our interviews live from Seattle with various people from Microsoft and MVPs.
[Zoe 00:30]
Hi, this is Zoe and I'm at MVP Summit in Seattle. I've been joined today by Marc. Would you like to introduce yourself?
[Kevin 00:35]
Sure, I'm Marc Anderson and I was just walking by peacefully and was snagged, but I'm happy to sit down with you.
[Zoe 00:42]
Fantastic. So, we'd love to get your honest view of copilot and that could be any of the Co-Pilots. So, first of all, what does copilot mean to you?
[Kevin 00:54]
It's certainly the drumbeat of Microsoft these days. Somebody was joking, it's the copilot company now. I think it's a promising set of technologies. I'm waiting to see how it all plays out. I think it has the potential for revolutionizing the way people work together and on their own work. But we have to sort of see how it plays plays because I think there's a lot of a lot of change that people are going to have to go through to really be able to take advantage of these technologies. We're not all techies.
[Zoe 01:31]
Yeah. We're.
[Kevin 01:31]
People with all kinds of viewpoints and skills and whatever so you know what one assumes from the technology trenches is not always what happens in the real world?
[Zoe 01:43]
Yeah I think I think that's a really important point because one of the things we've seen is that that choice to use CoPilot. It's highly personal, it's discretionary, you don't have to use it and you can just work how you always have. And the way that we all work, the habits, the work, the practices, how we manage our productivity, it's personal to each of us. So what that means is that each person has to put the work in to build that CoPilot muscle and we know that change enablement is hard. We know it's usually the thing that's cut out of projects and seen as optional And for me, it's not optional if we're going to actually realize the potential of copilot So what are your experiences with copilot that you feel have made a difference to you?
[Kevin 02:29]
To me so far. It's actually sort of been a toy. I mean, it's hey Let me try asking it this and and I think I would say that it gives me answers that I don't find Satisfying often enough that it's not something I turn to yet I do use github copilot. I write a lot of PowerShell. So I use github copilot It's on and it makes suggestions to me and I would say that, you know, two out of two or three out of ten times It's actually right on the money. Mm-hmm and the rest of the time it's It's not hallucinating. It's just telling me things that aren't quite right.
[Zoe 03:05]
Yeah.
[Kevin 03:06]
And so again, that's part of my you know I'm a natural skeptic But in practice, I've also seen that it's it it it requires me to Do extra thinking if I ask it a question as opposed to solving my problem.
[Zoe 03:22]
He's giving you extra cognitive load rather than yeah.
[Kevin 03:26]
Making me almost work harder because it's I think it's harder to just it's just like searching on the web You can find things that match your search, but are they right? Yeah, and you have to figure that out in the context of anything you're doing.
[Zoe 03:41]
Yeah.
[Kevin 03:41]
I.
[Zoe 03:42]
Think with GitHub copilot as well I've seen a lot of people talk about how this might help junior developers But that example that you gave where it's only right two or three times out of ten That's worrying for junior developers because they might not have the experience to actually know that it's not right.
[Kevin 03:57]
I think where it's been it's interesting because I think it's most useful in areas where you have expertise Mm-hmm where you can discern whether or not it's right if I'd use copilot to find out about the iron smelting process I don't know anything about the iron smelting process. So is slag really something that comes out of that process or is it just telling me something that makes no sense? So the concept that, you know, just like, to me, it's just like search. You can always find something, but that doesn't mean it's the right thing to find. And Copilot so far is the same thing. It's absolutely going to improve. I mean, I think that that's something we can all agree on. sort of a baby product.
[Zoe 04:41]
Yeah, really early.
[Kevin 04:42]
You know it has a couple, it's starting to smile at us a little bit, coos, so you know we think it's pretty cute but we'll have to see how it grows.
[Zoe 04:52]
Yeah, perfect. Is there anything that you're worried about?
[Kevin 04:56]
Oh yes.
[Zoe 04:57]
Many many.
[Kevin 04:58]
Things. You know it's interesting, we've seen some statistics this week and I'm always very careful, I don't remember which ones I'm allowed to know and which ones I'm not allowed to or at least say out loud. For example, there's a high percentage of code that's being committed to GitHub that is AI generated. That just opens a whole can of worms to me. How inefficient is that? Does that inefficiency matter? Is that code the best code that could do that thing? So that makes me think about things. And we've already sort of touched on some of my other concerns. Is what Copilot is saying, is what it's giving you something that you actually should take? Or is it actually gonna cause you bigger problems? You know, Alaska Air had asked copilot, should we have closed that door more tightly? And it said, no. Well, that's not so good.
[Zoe 05:54]
You know.
[Kevin 05:55]
So I think like I've been saying, you know, it's let's see how it plays out. But I actually wrote a blog post last week or the week before this is one of many sort of knowledge management or technology advances that Microsoft has come along and said this is going to solve everything and then they pivot and use something else to solve everything so there's also that sort of longevity and you know is this the flavor of the month? I don't think so but I also like to wait until I can see version three of things until I really believe that that they're the thing that they're supposed to.
[Zoe 06:37]
Yeah. No, that makes sense I mean, it's only what a year or two since Metaverse was the flavor of the year.
[Kevin 06:43]
No, but yeah that How do we live without?
[Zoe 06:49]
So last question, what are you most excited about or what are you hoping to see next I?
[Kevin 06:55]
Mean, I'm I'm hoping to see some real applications of people solving specific problems. I mean we were hearing We're hearing some interesting What-if scenarios, you know, we're and and this stuff demos really well Look, you can ask it this question that we already know we've asked before and we get an answer that seems pretty good but you know, how how reliable is it how can it be built into work processes and and you know, My knowledge management routes go way back So one of the one of the things that we always struggled with with knowledge management is what's the what's the ROI? And you can't really you can't prove it very early.
[Zoe 07:38]
Yeah, it's hard to quantify, isn't it? Yeah.
[Kevin 07:40]
And you can't do that with copilot. Yeah.
[Zoe 07:42]
Really easily.
[Kevin 07:42]
Either. Yes, it told me some things But did it actually make that thing I was trying to do happen faster and better or did it just take me on a little? Diversion, so I think we'll have to see how those anecdotes come out. We'll have to see what kind of stories people are telling, not just white papers, not the formalized ROI stuff. I mean, I'm willing to hear good anecdotes and stories of people saying, this solved my problem. And we'll see how that plays. But then I also want to see how, unfortunately, there's a lot of development thinking at Microsoft, as opposed to, let's just give a person a tool that they can use and make it work well. And so people are going to have to build stuff on top of this. And how is, how is that process? How does that feel? How quick can you quickly, can you generate those things and have them be valuable? I don't have a handle on it.
[Zoe 08:41]
Yeah. I think, I think one of the things I often say is that it's really important to remember just how early we are on this journey, because things like that, like, you know, how can we extend, how can we build on top of it? What is that true value and how, how are people reinventing and re-imagining how work. It will take time for that to play out. We don't have all those answers yet.
[Kevin 09:02]
Yeah, and as with anything, people don't like change.
[Zoe 09:05]
Change.
[Kevin 09:05]
Is bad, unless you're doing it to somebody else. And so, you know, we've got to let people take this journey and start, you know, we had Cortana. I mean, people could have asked that thing its questions, and it obviously didn't really work. So is this one going to work? Is it going to stick?
[Zoe 09:23]
Yeah, I heard a phrase yesterday, Cortana is dead. Long live Cortana. Exactly.
[Kevin 09:29]
We named it something else. copilot sounds so much.
[Zoe 09:32]
Better. Brilliant. Well, thank you for joining. That was excellent. Thanks for.
[Kevin 09:36]
Having me.