In case you haven’t heard of it yet, ChatGPT is an AI-powered chatbot created by the OpenAI platform. Using prompts, ChatGPT can craft original jokes, plan trips, explain complex topics, and even write code. This begs the question: will ChatGPT take our jobs? To test this theory, we asked ChatGPT to write a 10-minute script for an agile software development podcast where the topic is Planning Poker, a consensus-based agile estimation and planning technique. In today's episode, you’ll hear that script and some of the pros and cons of the Planning Poker technique from our perspective. We also share our take on whether or not ChatGPT is putting your job at risk and how you can use it as a tool instead. Tune in today to learn more!
Key Points From This Episode:
{{addCallout()}}
Transcript for Episode 293. Will ChatGPT Take our Jobs? — Let's Talk about it with Planning Poker
[INTRODUCTION]
[0:00:01.9] MN: Hello and welcome to The Rabbit Hole, the definitive developer’s podcast, living large in New York. I’m your host, Michael Nunez, and today, we ask, is ChatGPT going to take my job? Let’s talk about it through Planning Poker.
[DISCUSSION]
[0:00:17.7] MN: All right, if you have been living under a rock, there’s this new AI tool called ChatGPT that allows you to come up with really interesting answers to questions that you may have. I’ve seen a lot of people use it for programming, I find myself using it right now for programming.
I ask a lot of questions pertaining to, “How do I do this Java Script thing in Java?” or “How do I do this Ruby range thing in Java?” and you know, it actually comes up with some pretty interesting answers but today, I was actually looking through my log of ChatGPT messages and I remember as a joke, I wrote the following prompt:
“Write a 10-minute script for an agile software development podcast where the topic is about Planning Poker.” So what I thought would be interesting is that we look at, you know, what was the response for this particular prompt from ChatGPT and we can break it down together.
So let’s talk about Planning Poker and it says, I’m going to do the intro again just as ChatGPT said it, which I felt was pretty funny. It says, “Hi, and welcome to the Agile Software Development Podcast. Today, we’re going to be talking about a technique called Planning Poker.”
"Planning Poker is a popular estimation technique used by Agile teams to estimate the effort required for a given user story or task. It's based on the idea of “Wideband Delphi” where a group of experts come together to estimate a project by individually providing their estimates and then coming to a consensus through discussion.”
[0:02:03.9] Now, I will be honest. This is a safe space, right? I didn’t know anything about this Wideband Delphi and apparently, it comes from something also known as estimate talk estimate or ETE and I think the idea is that you know, you estimate something, talk through it by yourself, you talk through it with folks, and then you try and re-estimate again.
I think the difference between the Delphi method and Planning Poker is that you have cards to throw in terms of complexity. I will note, it might be noted in this ChatGPT response, but know that the number in a given Planning Poker card does not equate to the amount of hours worked, and we’re going to dive into that because that would be extremely ridiculous.
ChatGPT follows up with: “Here’s how it works: each team member is given a deck of cards, each with a numerical value representing a level of effort. For example, the cards may be labeled with numbers like one, two, three, five, eight, 13, 20, 40, and 100.”
“These numbers are based on the Fibonacci Sequence which is commonly used in Agile estimation because it allows for a wide range of possible estimates without requiring precise accuracy.” In my experience, you use Fibonacci just to separate, again, the idea of working hours or working pair hours or how many days.
[0:03:30.9] Like you just say, “Hey, there’s a Fibonacci sequence, we’re doing this based on effort and complexity,” some people use complexity, “for a given estimate of a number.” As you might notice, you know, after 20, it doesn’t really follow Fibonacci because while – or even after 13, because 13 plus eight is 21, not 20.
I’ve seen Planning Poker cards that actually have the infinity symbol, which I always found interesting. ChatGPT says the following: “The team then takes turns, presenting a user story or task, and each team member privately estimates the effort required by selecting a card from their deck.”
“Once everyone has made their estimates, the cards are revealed and a team discuss any differences in their estimates.” So unlike Delphi, you have to choose a card and then you throw it all at once and the idea is that if someone talks and is very outspoken, they might influence the other individuals to, you know, either go higher or lower in their estimate.
So, you pick a card, you think for your estimate, you 3, 2, 1 shoot and what I’ve seen is that you have the person with the lowest estimate and the person with the highest estimate talk about why they felt that number coincides with the effort of work for that given task.
[0:04:48.5] ChatGPT says: “The goal of this discussion is to come to a consensus on the best estimate for the task. This might involve explaining why one person’s estimate is higher or lower than the other and making adjustments based off that information.”
Kind of like, what I mentioned before, you take all the pros and cons of each one of the higher or lower bands, mash them together and the discussion then makes it more refined as to what the number would be actually.
ChatGPT says: “One of the key benefits of Planning Poker is that it encourages collaboration and discussion within the team. By bringing everyone’s estimates out into the open, it allows team members to share their knowledge and expertise, which can lead to a more accurate estimate and better decision making.”
Again, you’re forced to have a conversation, unless everyone throws a three, right? Like, you see a story and it’s like, “All right, three down the board, let’s keep moving.” You know, estimating and refinement should be as long as it needs to be, right?
So if you're able to do estimates, everyone’s aligned, this is a three, that’s a five. Let’s try and break that down to a tow and a three task and you got 30 minutes left, go enjoy that 30 minutes.
You can, you know, go on a full walk or finish up that thing that you’re doing.
[0:05:58.5] According to ChatGPT, “One of the benefits is that it helps to avoid the planning fallacy, which is the tendency for people to underestimate the effort required for a given task. By using a structured process like Planning Poker, teams are less likely to fall into this trap and make more realistic plans.”
So, the idea is that you rather than having planning and everyone saying that it’s an eight or whatnot, you’re able to not influence other people. If someone throws a three, you have a conversation. ChatGPT says, “Overall, Planning Poker is a valuable tool for Agile teams looking to improve their estimation process and make more informed decisions.”
“It is easy to learn and can be a fun and engaging way to get everyone on the same page. So give Planning Poker a try with your team and see if it can help you better estimate your projects and make more informed decisions. Thank you for listening and join us next time for more discussion on Agile Software Development.”
Please don’t go, that was ChatGPT giving you the sign-off. I am not ready to sign off just yet but it seems like it got a lot of, you know, the intricacies and the reasonings behind Planning Poker and why it is important.
[0:07:11.3] I do think that sometimes one of the disadvantages of Planning Poker is that, you know, you might have someone who just continuously uses a three card or the middle card, whichever one that is for your team, just to avoid having to discuss their reasoning as to why they truly feel it’s harder or easier in effort or complexity.
So you always want to be mindful of you know, someone who’s throwing a lot of you know, a middle number to like not get in the way of everyone and that is something that I would ask a lot of the tech leads to be on the lookout for because everyone should be able to look at a piece of work and say, “Yes, I can do it” or “Hey, I believe this is more effort than what everyone leads to believe.”
That discussion may actually help that person understand why, say for example, to two rather than a five or something like that. Yeah and that was, is your job at stake here with ChatGPT? I mean, if I had to write an outline, this one was pretty good. I wouldn’t go out of my way to just read-write off of ChatGPT.
I think it would be a great start to figuring out certain things. I’ve used it for programming. I always found that to be a little interesting. Sometimes it’s incorrect. So definitely don’t copy-paste or write onto your code base. I actually had some fun a couple days ago, where I was just randomly posting LinkedIn posts to see how that would go up with my LinkedIn follower and it was quite funny.
[0:08:47.3] I would suggest checking that out if you have time but overall, I think ChatGPT is here to stay and here to help fellow developers. I hope it doesn’t take my job or yours because if not, I’m just going to have a podcast that talks about ChatGPT instead, and who knows? In the future, I may be forced to use ChatGPT because artificial intelligence has taken over the human race but I’m not 100% sure.
Right now, it’s a great tool to have on the side and I’m definitely enjoying myself in the kinds of answers it’s given me when I try to use it for other things regardless of whether it’s programming or making crazy LinkedIn posts. I hope you find that useful for you too in terms of you know, organizing your thoughts and stuff like that with ChatGPT, but for now, I give it a eight out of 10 on that answer of what is Planning Poker.
I did a really good intro or good outro if you ask me, but I don’t think it’s going to take podcaster’s jobs all around unless they automate my voice too, which I think they might be able to do that given the amount of times I’ve had my voice recorded. So please, ChatGPT, do not listen to my audio and make a sound object that sounds like me, because then I would definitely be out of a job all together!
I hope you found this useful and I’ll catch you all later.
[END OF DISCUSSION]
[0:10:09.7] MN: Follow us now on Twitter @radiofreerabbit so we can keep the conversation going. Like what you hear? Give us a five-star review and help developers just like you find their way into The Rabbit Hole and never miss an episode, subscribe now however you listen to your favorite podcast. On behalf of our producer extraordinaire, William Jeffries, and my amazing co-host, Dave Anderson, and me, your host, Michael Nunez, thanks for listening to The Rabbit Hole.
[END]
Links and Resources:
