What AI Changes

00:00:03: Hi folks, this is All

00:00:05: Things

00:00:06: Product with Petra Ville

00:00:07: and Theresa Kortz.

00:00:09: And we're so happy

00:00:10: you're here!

00:00:17: Petra right before we started recording I was on LinkedIn... ...and wrote the screed about what has not changed because i'm really annoyed by all the people that say everything is changing.

00:00:30: a

00:00:30: lot is changing.

00:00:31: let's just start there will acknowledge that.

00:00:33: but it would be fun for us to dive into this topic AI changes and what it doesn't change because it seems to be the hot topic of the day.

00:00:42: And I think we might have some surprising hot takes,

00:00:47: should they just share?

00:00:48: Some of the heartaches that i have been discussing recently.

00:00:51: yeah

00:00:53: who from the top off my head?

00:00:54: random coaching conversations hallway conversation water cooler conversations where do you have hallway

00:01:01: in?

00:01:01: watercolor conversations?

00:01:03: oh still here in hamburg at meetups and something like that.

00:01:06: So in Hamburg, the product community still tends to hang out with each other!

00:01:11: Nice yeah so it's really nice.

00:01:13: there are more Meetups than ever by the way new ones popping up even.

00:01:18: I appreciate smaller ones maybe but still nice and cozy.

00:01:23: i couldn't complain.

00:01:25: yes one thing that I was sitting with for a while, was this hey our engineering department is now so much more efficient.

00:01:36: Cost of developing something went down so much but you know we still have just massive bottleneck called testing.

00:01:45: So now the stuff is piling up at the end of our task works and We still struggle so much with testing what we got which led me to this kind of, and it was just a conversation that I heard.

00:01:58: It wasn't the conversation that i participated in but okay... Was that tying back with our last episode by the way?

00:02:09: Was an engineering organization that understood what good looks like in the past already?

00:02:17: Because how can it be that now everybody is super happily developing features with the help of AI or even agents do without the help from the engineer, however they do.

00:02:26: But everything piles up and is untested...

00:02:30: Was this about testing or was it code review?

00:02:36: Both!

00:02:37: Okay because I heard

00:02:39: a lot teams

00:02:42: reviewing all the codes that AI creates.

00:02:44: Yeah, like you go through this cycle of engineers are no longer writing the code and a lot of organizations one engineer writes the code before it's released.

00:02:53: another engineer reviews the code.

00:02:55: And so now with agents writing in the code there is an up pile-up of code review until you get agents to do that.

00:03:02: There

00:03:02: was one aspect but I was actually wondering if they did test driven development Or if there was never a case in this organization and they're always in, so retrofitted all their testing after developing the actual features or something like that.

00:03:20: So there wasn't interesting thing.

00:03:22: And it's one thing.

00:03:23: for where AI is not changing any of the problems off their tech organization obviously It just emphasizing The issues that may have already had.

00:03:33: Yeah, this one is really surprising to me because I actually think One of the things that agents are really good at.

00:03:38: Is writing automated testing

00:03:40: That it still

00:03:41: requires like human oversight To make sure they're right test.

00:03:43: yeah written.

00:03:44: but i'm wondering Like?

00:03:45: It's just an organization that didn't already have Automated

00:03:49: which you mean.

00:03:49: it sounded like exactly right.

00:03:51: like.

00:03:52: do They Have qa people that Are literally testing every teacher?

00:03:55: yes,

00:03:56: they were talking about QA People.

00:03:59: and And then there It's humbling to some extent, to listen to those conversations.

00:04:05: Actually I had the conversation with Arne afterwards because we're currently starting a process of curating next year product at heart.

00:04:12: and then again as I said to Arne it is always so interesting to hear these conversations Because us belonging so much this product bubble And this product thought leader bubble Sometimes you forget that companies are still out there.

00:04:26: Maybe manual testing.

00:04:28: people are manually testing stuff before it then gets deployed in a weekly release, maybe even or something like that.

00:04:37: Maybe

00:04:37: weekly?

00:04:40: Yeah I did talk with the team recently.

00:04:43: they're actually two of the best product managers um That i've worked with and they shared.

00:04:49: They started to share some delivery concerns And A lot was rooted around The fact their Team Was reluctant To use AI to write code.

00:04:57: They had literally no automated testing.

00:05:00: What they shared with me was, they would do a sprint and release new features but as part of those feature releases the whole bunch old stuff will break.

00:05:10: Yeah And so it's like...they were reluctant to give their engineers new stuff because that would just break all this old stuff.

00:05:16: It'd break core functionality in there app.

00:05:20: I remember asking them Do you guys have any automated testing?

00:05:24: And they were like, no we don't do anything.

00:05:28: Interesting!

00:05:30: Yeah so

00:05:31: I don't think this is a function

00:05:32: of why

00:05:33: people end up in organizations that don't have

00:05:36: the chip yarding?

00:05:37: I

00:05:37: just don't

00:05:37: know what AI changes...I think maybe AI is exposing what was already broken.

00:05:43: yeah i would totally subscribe to that.

00:05:46: and for example that's another hot take And it actually made me gasp for a second.

00:05:53: It was somebody saying on the stage that they think we now finally can ditch all... all the scrum that we ever used or something like that.

00:06:02: So, all of our frameworks to organize work basically are no longer important and at first I thought this is a bold statement for somebody on stage in front.

00:06:15: you don't know who's in audience right?

00:06:17: And maybe it such an organization so close.

00:06:21: finally!

00:06:22: have a good rhythm and good rituals for what they actually do, no matter if it's with the help of Scrum or Voodoo.

00:06:29: Or I couldn't care less but there always will be this ritual that you need.

00:06:36: We always plan to work even in times of AI.

00:06:41: Of course again AI can help us to plan our works But still somebody has to be aware of planning their work being a thing and reviewing the work being a thing, and demoing the work.

00:06:54: And it's impact being a things and measure measuring to success of the word forever being a think.

00:07:01: last but not least we just recorded an episode on that as well is kind off look at your work and you're efficiency and then thinking about how can we improve our working relationships for this sake of be more efficient in the neck?

00:07:17: For example they they no longer talk about iterations.

00:07:21: They gave it even a different name and I was like, oh my god now we...they still work in one week chunks.

00:07:28: but because they just declared that scrum is no longer thing you of course don't call it iteration anymore!

00:07:33: They found the new word for that.

00:07:34: um and this is for example something that i think AI isn't changing as well.

00:07:40: so this planning to work reviewing to work demoing to work making your work more efficient no matter if you use a scrum framework or any other framework for organizing, but somebody has to take care of organizing the work with the help of an large language model maybe.

00:07:55: Go for it!

00:07:57: But still that will remain

00:07:59: unchanged.".

00:08:00: So they argued scrum is dead.

00:08:02: let's put aside this scrum part.

00:08:04: really.

00:08:04: what they argued was like planning and aligning like a line in this hand for what we're going to accomplish and then measuring whether we did that, maybe doing retrospectives.

00:08:17: Yes some of these rituals that we associate with scrum do they argue.

00:08:21: why?

00:08:21: Um

00:08:25: yeah to some extent.

00:08:27: so they double-clicked the bid on estimation meetings refinement meetings and that estimating things was maybe always a bit of an overhead, never good idea.

00:08:40: And so now in terms of AI it's finally over.

00:08:43: With that part I do agree because all the cross-functional teams have certain level of AI efficiency or literacy work more on product itself or software these days.

00:08:57: not much understanding and putting into specification But even then you still will need to be asked To put things on time horizons at least.

00:09:12: Yeah, so I don't know it's.

00:09:15: It was well but there was what they were arguing with.

00:09:17: So finally these rituals are all.

00:09:19: They always have been overhead and now they're no longer necessary because There would be way too much over head in times of AI and faster development cycles.

00:09:28: I mean, i argued to fifteen years ago that estimates are a waste of time.

00:09:31: so...i agree with that part.

00:09:33: Yeah!

00:09:34: I think thats the point and i agree yeah..with as well but the rest.

00:09:38: But

00:09:39: there is no difference between like estimating story by story task-by-task which frankly it's just pure fiction from the beginning.

00:09:50: Yes but estimation

00:09:51: for alignment btw its always something that i found helpful.

00:09:55: And saying This week we're working on this chunk of work and our goal is to complete it like that Is always needed regardless?

00:10:03: Like I don't think AI has changed.

00:10:05: That's like are We just saying go do some more can get as far As you get and well, we'll talk about what you did later.

00:10:11: like yeah I'm not sure i get It.

00:10:13: look how Do you even Work with other people if You're Not Having a Conversation About Who's Doing What by When?

00:10:20: Yes, exactly.

00:10:21: So that is why I always when people are saying like scrum is dead or safest dad and would I hope safe to be there forever?

00:10:29: But still you need a process even as a product person.

00:10:39: A very small nucleus of an agreement and a process for one team.

00:10:45: And it's only this one team that works.

00:10:47: by then, yeah maybe all the teams are shrinking.

00:10:49: I get the idea!

00:10:51: Maybe there aren't fewer engineers in their team?

00:10:54: Maybe they're smaller than

00:10:55: them?!

00:10:55: It is super because not just six people have to rely on how they work.

00:10:59: They don't need three hour long retrospectives every other week.

00:11:03: cool but i dont think scrumstead or any processes Defining framework is that

00:11:12: what I think it's funny about this one.

00:11:14: Is that?

00:11:15: For people, they're using agents to build.

00:11:18: They're like we are seeing the cycle where like early on when people were building with agents People started talking about planning documents and first plan with the agent And then let the agent Once the plan is finalized, let the agent build and you'll get better quality.

00:11:35: I even wrote a blog post about this right?

00:11:36: Like then we get to the point where You have multiple agents building And so people talk about like.

00:11:42: there's this.

00:11:45: If you hear about people talking about loops and graphs these are all ways To communicate requirements So the agent knows when it's done.

00:11:54: and like with a graph, It is way to let multiple agents coordinate their work.

00:11:59: The agents don't step on each other's toes which by-the-way all of this Is same as rituals and ceremonies we do With our teams To build software We're just now doing for agents.

00:12:10: For someone say Like This isn't needed or its dead.

00:12:13: Its actually more needed Because while number Of people involved might be going down The number of an Agents involved Going up And throughput of work is going up, which means like for anybody to have any idea what is happening you need some sort of way of saying who did what by when.

00:12:33: Yep and yeah it's actually even supercharged!

00:12:37: You're right.

00:12:37: I haven't thought about it from that angle.

00:12:41: so stating all these work agreements are dead how we organize work instead?

00:12:48: That's here to stay i think as well

00:12:51: okay.

00:12:52: So mine

00:12:53: are all around discovery habits, right?

00:12:56: Like I just hear people say you know we don't need to do any discovery anymore because delivery is free.

00:13:02: We can just launch our product and hope it works or here like the other day i read an article about how story based interviewing doesn't work anymore...we can't ask people about their past behavior because AI's brand new when no one has used it before this.

00:13:19: When

00:13:19: I read this, even my mom was very decent AI experience.

00:13:23: Yeah like when i read these things what it tells me is that people misunderstood the habit in the first place.

00:13:29: so yes you can build something and just put it in the world but when it falls short of your expectations You're not going to know why.

00:13:37: because you didn't do the work To enumerate the assumptions It then depended upon?

00:13:41: You didn't systematically test those assumptions.

00:13:44: So literally all you learn Is it didn't work You don't learn how to fix it, you don't know how to iterate on.

00:13:50: What do differently next time?

00:13:52: You're just throwing spaghetti at the wall.

00:13:55: and same with interviewing like story-based interviewing was never tell me about your.

00:14:01: my product.

00:14:02: It's not About The Product is Not About The Thing That Your Building.

00:14:05: Its About The Customer.

00:14:06: Tell Me About The Last Time You Experienced This Problem This Need And It Doesn'T Matter That The Solution Is Brand New.

00:14:15: Your customer is not brand new.

00:14:17: They're still a human, they've passed lived experiences and

00:14:20: real life experience

00:14:21: to learn about those things.

00:14:23: so

00:14:24: I am the

00:14:25: worst one

00:14:26: how you would because this You notice customer delight moments these aha moments These features that you have never thought about but then you see them struggling or than you hear them share a story And then the entire team has just huh?

00:14:40: This actually something that we could solve for.

00:14:44: Where people want to get those ones from if they're not talking to humans?

00:14:49: They just wanna ask

00:14:51: the LLM.

00:14:55: Yeah, so convenient...

00:14:58: Yeah!

00:14:58: I mean here's what i'm glad about like fine go ask your LLm.

00:15:02: use that to build your product.

00:15:03: good luck.

00:15:04: um we're going through a period of time right now.

00:15:07: i'm just thinking About The fact That Like Bending Spoons Just By AirTables They Just Announced We're Buying Miro.

00:15:18: A week or two ago from the time of our recording, wrote a post on LinkedIn about how I don't think SaaS is dying because AI.

00:15:26: If your SaaS company is dying it's because you're making bad decisions for your customers and that has nothing to do with AI right?

00:15:33: And so i wrote about air table in particular...I didn't name AirTable on Linkedin but here I am!

00:15:38: I'm now naming them on our podcast.

00:15:44: They are the heart of my knowledge management system in my business.

00:15:49: About two years ago, their performance cratered!

00:15:54: I wait—I'm not exaggerating—forty-five seconds to wait for a view to load on my

00:16:00: browser.".

00:16:14: Focus.

00:16:14: they got distracted.

00:16:16: They started building a lovable a replete and AI agent that builds things for you.

00:16:20: And they stopped investing in their core product, and there core product deteriorated very quickly.

00:16:27: This is not AI killing sass this is people at SASS companies making stupid decisions.

00:16:34: Well

00:16:36: now Ben and Spoonz, I fully expect them to quadruple or ten x my price.

00:16:41: So i am furiously working to get off of the air table!

00:16:43: Yay thank you.

00:16:44: um yeah

00:16:46: You know that I just did that with harvest so welcome to... Yeah like

00:16:52: it's so easy to blame whatever is new AI didn't kill AirTable.

00:16:59: The

00:16:59: air table team getting distracted from their core value killed the air table.

00:17:04: I can make another argument raising way too much venture capital kill to irritate.

00:17:09: Yes, yeah that happened to so many startups.

00:17:10: in the first they

00:17:11: were a very successful growing company.

00:17:14: The problem is they weren't growing at a rate that justified the value.

00:17:19: They raised money at

00:17:20: yep.

00:17:21: This

00:17:21: is a problem they created for themselves.

00:17:23: And I can make the same argument for Miro, Miro brings in six hundred million dollars a year and revenue.

00:17:30: that's very successful business!

00:17:32: If you talk to customers.

00:17:33: lots of them are very happy.

00:17:36: so why do we have to sell?

00:17:38: Because they raised way too much money at a seventeen point three billion dollar valuation how are you valued at seventeen point thirty billion when you have six hundred?

00:17:49: Yeah, let me do the math for you.

00:17:51: That means you think this company is gonna make at least six hundred billion million dollars a year For the next thirty four years.

00:17:58: Yes like come on We're like This is getting

00:18:02: and we will see the exact same thing happening with most of The niche AI products.

00:18:09: yeah And I don't think they all have raised way too much money as well.

00:18:18: This is not killing sass.

00:18:20: I think it's the venture capital mindset of one winner takes the market Which by the way if you look at any market, that is not true.

00:18:27: The only market i can think Of right now where thats even remotely True Is search.

00:18:32: and You know what google has faced a lot of litigation for antitrust issues over That.

00:18:37: but like let's Look At um air tables business we have Air table We Have coda?

00:18:47: Miro we have mirror.

00:18:48: We have mural, we had fig jam.

00:18:49: We had Jamboard.

00:18:50: right like all those companies were doing.

00:18:52: fine It is not.

00:18:54: the VCs create this ecosystem where they convince founders one company will win and everybody else will fail.

00:19:04: Yeah,

00:19:06: it dramatically reduces the chances that a company will be successful.

00:19:11: yeah And I find this wild in these times especially Because the consumer will be less tolerant of one mediocre solution.

00:19:23: And with more players in the market, everybody was able to find The One Solution that fitted their needs best.

00:19:29: but now with AI In the picture and Julie Sue just wrote a very lovely newsletter about it and how she is doing hyper personalized software for her business?

00:19:39: So do I with my harvest solutions?

00:19:41: so entire CRM invoicing estimate stuff is all heavily personalized now with the help of Cloud Code.

00:19:52: And, of course are we the tech bubble?

00:19:55: Are we the forefront?

00:19:56: Are We The Early Adopters?

00:19:57: Sure!

00:19:58: But will people become more and more capable of doing that in building stuff for themselves?

00:20:03: Yes...and then a market where you try to have one player catered for everybody's needs?

00:20:10: it's just a wild bet isn't?

00:20:15: Okay, so you just opened a giant can of worms because your starting to make the argument that AI is- And

00:20:20: we already had twenty minutes.

00:20:23: No no I'm not arguing.

00:20:24: it does that at

00:20:25: large scale.

00:20:25: What what you said is a trend.

00:20:27: i do think Technologists technology companies will build more Single point solutions internal Because they will benefit from them being completely customized.

00:20:42: yeah

00:20:43: I don't think the average consumer is going to do this.

00:20:46: I Don't Think The Average Business Is Going To Do This, i Think What's More Likely To

00:20:50: Happen?

00:20:51: We're Gonna Have Lots Of Smaller Players Building Much More Targeted Solutions For Niche Customer Segments.

00:21:00: Yes

00:21:01: Hopefully

00:21:01: Right.

00:21:02: Yeah That'S what I would love to See.

00:21:04: That's what I would love to see too.

00:21:06: But then the

00:21:07: bet is a weird one, The Bet of all these shareholders saying like there was won and winner takes in every market.

00:21:13: basically

00:21:14: Yeah i mean im sort hoping that venture capital just falls apart but we'll se who it happens.

00:21:19: Which opens the can of worms on capitalism?

00:21:22: Are not going ther today Not tonight?

00:21:24: No!

00:21:25: I wouldn't say venture capital falling apart opens the Can of Worms of Capitalism.

00:21:29: Capitalism is fine.

00:21:30: I think venture capital has whole different beast.

00:21:33: I do, here's what i will say.

00:21:40: This is going to change the way software is built and that if you're building software one thing some new ideas around personalization in platforms, in products.

00:21:58: Rather than like I don't think the average restaurant owner is going to build their own point of sale system because they can do it with Claude Code.

00:22:06: Agreed!

00:22:06: But i think a point-of-sale provider Can now almost like Notion...I actually Think Notion Is A Huge Leading Edge Innovator In This Space.

00:22:16: Notion was built around this idea of building blocks.

00:22:19: You have a database as a building block, you have a wiki page.

00:22:23: is the building block?

00:22:24: I don't use Notions so i dont know all their building blocks but ive heard about it.

00:22:27: The way they build there product Was around these ideas and that allows them to mix and match those building blocks in infinitely customizable ways.

00:22:37: This pattern we will see more often And hopefully we will see SAS providers figure out how to offer hyper personalization because AI makes it just now possible.

00:22:49: Yeah,

00:22:50: that is what changes and as we call the episode What Will Remain Unchanged in Times of AI?

00:22:57: I think these are all The things We've learned In the past.

00:23:03: so this Is like Get your processes in order, get your rhythms in order.

00:23:09: Have you rituals still going?

00:23:12: Maybe change them maybe shorten the cycles.

00:23:16: may be ready for hyper personalization but a lot of the basics are not changing.

00:23:21: at least

00:23:22: talking to customers Keep

00:23:24: talking to

00:23:24: your customers.

00:23:26: You said you're

00:23:28: retiring from saying this?

00:23:29: I know, I can't.

00:23:30: it turns out!

00:23:32: It's a reflex.

00:23:33: so thank-you Teresa.

00:23:35: Thanks Petra that was a spicy one.