Reverse Conway's law and GenAI: How agents will take over the organisation

Talk transcript

1. 0:00

   Today we're gonna talk about Conway's Law, Reverse Conway's Law, and generative AI, and it's basically thinking about the future of how AI will influence our organizations. Now, we got this new technology, right?

2. 0:16

   Like ChatGPT was the first, and it, like, impacted, like, everywhere, the whole industry. And this really brought like... Everybody was, like, so excited that, like, they kind of changed everything to becoming more of an AI specialist, and that's really cool because everybody's still learning, um, and we'll see where we end up.

3. 0:37

   Now, some kind of just sprinkle some AI on there, and they don't go all the way, and that's fine, but that... I don't think they will kind of really get the results that they, they want from it.

4. 0:49

   Now, the bigger questions that companies who do want GenAI, they wanted to know, like, where does it land in the organization? Will it belong to the data team, the application team, the platform team?

5. 0:59

   We don't really know. So we already see that, like, shaping our org for this new technology is something that is happening in a lot of organizations. Now, last year at the AI engineer event, I talked about platform teams and how kind of you can introduce kind of, and scale AI out of.

6. 1:20

   You can watch the video here, but that's not what we're gonna talk about today. Today, we're gonna talk about Conway's Law, and simplistically, it says that whatever, uh, way you organize yourself and your teams will have an impact on what you're able to build.

7. 1:37

   Like, if you have small distributed teams, you're likely gonna go with modular services and architecture. Here you can see, like, various pieces of organization, and a lot has to do with communication lines and how the communication lines actually, uh, influence a part of the organization.

8. 1:54

   So there's a kind of impact of what we can build. Now, who am I to talk about this? Um, well, I've been in automation for a long time, been following a lot of the DevOps, and now been really focused on doing more with GenAI, as anybody of the else, right?

9. 2:10

   And I'm particularly focused of engineering more with AI, so kind of as a software developer to do more with AI and kind of bring the benefits out. If you're looking at this, uh, video, uh, if you're interested in the slides afterwards, please, uh, hit me a message on LinkedIn.

10. 2:27

    Uh, I'll be happy to share, um, in exchange for some feedback, uh, or things to improve.

11. 2:33

    Another thing that I'm, uh, active on is the AI Native Dev community. I'm a curator and a contributor, um, where we try to get the new stories out about how, uh, kind of development will change inside of the organizations.

12. 2:47

    Now, I told you a little bit of how we're organizing for AI with a platform team, but today we are doing the opposite. How is AI changing the way we organize ourselves?

13. 3:00

    So that's kind of a, a kind of a kicker or a kickback, uh, in this. Now, what have you seen so far? One of the obvious things is, first level is everybody kind of starts accepting AI as a copilot, and that's really great.

14. 3:15

    Um, some kind of start going even further. They think about, like, training all their knowledge, uh, this is coming back. Um, and, like, all the knowledge that we have, it is changing, and we can pass that on to that sidekick.

15. 3:30

    That's great. Now, eventually, we're already seeing that AI is writing more and more code and will become more of a contributor. So even though it's a sidekick, um, uh, and kind of a copilot, it really drives us to, uh, more efficiency and more contributions in the organization.

16. 3:50

    Now, this brings the elephant in the room, like, are we still needed? And that's the question a lot of people ask: What should we do? Uh, will they eventually kind of, uh, change the way, uh, we have to operate?

17. 4:01

    Now, that's normal, and a lot of people are kind of, like, dealing with this issue at work, and they are scared. And often you have to think, like, when a new technology comes, you have to rethink your job, and sometimes the culture of your company really supports that, and you're able to kind of help and learn about

18. 4:21

    that new stuff. I hope that's the case for you, and that you actually bring this new skill of how to figure out what you need to do, uh, in the next couple of years.

19. 4:30

    Now, what we're seeing is, instead of just coding, as an example, we're moving away from just typing and having it help generate some of the text to a new role, is that we start thinking about the intent.

20. 4:43

    Something we've always wanted, but we've been there kind of in the toil, writing code, typing, and now kind of the new technology is taking this up to a level where we can think more about the planning and the intent.

21. 4:56

    There's even products that start where-- emerging, like product requirements, and that generates the code, so we're less about the typing anymore. Or you could say it's becoming specification centric, and we feed it with the specs, and then eventually it builds the system.

22. 5:12

    Now, it's almost like, well, we're reviewing, and we still have to be good at reviewing. We say, "This is good, this is bad," but we're gonna do less typing.

23. 5:22

    And this is just an example of what everybody kind of using copilots and so on is kind of experiencing. You become from, you know, kind of more the creator to the person that is managing and kind of more on the how, uh...

24. 5:35

    sorry, more on the why and less on the how, and that's great. Now, to put this in perspective, every time there's a new technology, like, people already have a job, and they do a s- bundle of tasks, uh, as the upper, uh, kind of shows.

25. 5:52

    Now, technology comes along. Sometimes it adds or it's the copilot or it kind of complements what we do. And sometimes it will be substituted. So kind of that unbundling of tasks is something that is normal when you kind of like move with a new technology.

26. 6:11

    Now, what can happen? Uh, it's just gonna be the same. We're just gonna do more, and we ask, uh, kind of, uh, extra for this, right? So sometimes that's what technology does, it adds a premium to what we can do.

27. 6:26

    Now, sometimes, like maybe the value that we put in coding is moved somewhere else to m-management, uh, or kind of manage, uh, specifications, and that's kind of another thing that might be happening.

28. 6:40

    So even though you are like moving away from your existing task, uh, something happens in another tasks. And then eventually it might be commodity, and that's where you kind of like get this feeling of, "Hey, if this is commodity, what should I really be doing?"

29. 6:55

    And how to go into another task to kind of prove your value to the system.

30. 7:01

    And eventually, that piece that was kind of really good and kind of was delegated become substituted. We don't care anymore. The value is kind of, uh, not good anymore, and we really have to find like another job in the system.

31. 7:14

    Those things are the typical scenarios a new technology brings. Now, you might think like maybe we're not coding anymore, but we're training the, the AI to do coding. Here's an example, Outlier, where, you know, people get paid.

32. 7:29

    I'm not saying this is a bad job, but it shows you kind of how technology can change some of the tasks and the difference kind of job that are being created.

33. 7:39

    Now, eventually you'll see kind of these stages. It won't affect me. That's when kind of the task is kind of complementary, and then you kind of go all the way up to, "Hey, it's gonna replace me.

34. 7:50

    That's really all." And eventually you find a spot that says, "I'm-- can now do things that I couldn't do before, and I have now new superpowers, and I cr- can create something."

35. 7:59

    So kind of that's the roller coaster that we're on as an individual person in this new technology. So this is how AI is affecting us as, uh, an individual.

36. 8:10

    Now, we know that things will go wrong. So in this case, Amazon two bots kind of like, you know, uh, increasing the price because one is bidding over the other.

37. 8:19

    So s-humans are still needed, right? In this case, uh, but as I showed before, it might not be in all cases. But things will go wrong, and we have to be prepared for that.

38. 8:29

    And sometimes humans just have to clean up because that's what we do. We understand what went wrong. We make sure, uh, like, uh, the machine can't handle it, and this is the med- intermediate scenario that we're going for.

39. 8:42

    Now, a lot of the automation, what I learned is that a lot of it is about automation, and then it is about preventing failure and designing from failure and so on.

40. 8:49

    So kind of dealing with failure is still typically a human job. It might f- require less people, but when you kind of have this happening, you have to be prepared.

41. 9:00

    So you still have to learn, train, and deal with all kind of, uh, the situations. So it's almost like a paradox, like the more we gained with the automation, we still have to be training and be ready when the failure happens.

42. 9:12

    So it could be, uh, kind of a, a net zero, uh, effect as well. Now, let's move up. We kind of have that delegated assistant co-pilot, and now we're heading towards the team, uh, level.

43. 9:25

    And instead of h-having a, a personal assistant, we think about like a team member that has access to the team, that helps the whole team, uh, to do a better job.

44. 9:34

    And sometimes, uh, the impact at the team level is the domain knowledge. And as you can see, when multiple people are on a team, they have various levels of domains, and we really focused on having kind of multi, uh, purpose, uh, or multi-domain knowledgeable people, the full stack engineer, and bringing a team of experts of various domains

45. 9:56

    together. We found out that's the best way to organize a team, and that's great. Now, if you think about like the human team, that's great. It has a limited s- number of knowledge which is contained, let's say the s- seven or eight, uh, team members.

46. 10:12

    And then there's the whole domain knowledge that an AI can have. It is way bigger, and it can now be, uh, another way of getting that knowledge. You already see that, like people getting questions, asking it, but it's still the human that does the verification, but it's vastly helpful to, uh, to this more automated, uh, as an

47. 10:31

    agent. So eventually that-- this might lead to smaller teams because that overlapping is maybe not that needed because we have s- uh, a system that actually knows a lot.

48. 10:43

    Again, we have to still understand what good looks like, so that might change the, the skills that we require in those teams as well. And kind of maybe those smaller teams, again, become another team.

49. 10:56

    Who knows? Kind of like that is a scaling issue. We're kind of all speculated. Uh, some-- I've seen some cases where teams got smaller. I've seen some cases where teams got bigger.

50. 11:06

    Uh, we don't really know, but the dynamic will likely change, uh, with more AI being as a team member. And you can say that the stack that we're building, like whatever we w- wanna know, uh, there's the applications, uh, there's more of the infrastructure, and then the teammates, and we're looking for more domain-specific teammates, much like the

51. 11:29

    diversity that we brought in teams that can help somebody in sales, somebody's really, really good at data, somebody's really good at other skills. So kind of bringing a new teammate stack all together, uh, with the technology stack.

52. 11:42

    So kind of that expansion is something that you'll see happen with more, you know, uh, agents and parts becoming part of the team. Now, that means that, as I m- uh, mentioned before, that we might be faster, uh, and they can do things a lot faster, but you are still the person who kind of has to make

53. 12:02

    the decision. So maybe there's not enough work for you, but there's a lot that you can do with multiple of these, uh, assistants and, and team members. Um, and that's kind of a specialist that goes in when things fail or has an understanding, uh, of what to do.

54. 12:21

    Now, what we all see is that maybe the cycles that we're used to, the one in two weeks, get reused in two days, so it's becoming a shorter feedback cycle.

55. 12:30

    So that means we're getting into the loop, but we can't have, like, the conversations with everybody. So there's gonna be multiple feedback cycles, maybe that complete in a bigger feedback cycle.

56. 12:39

    But we're getting to a more real-time cycle compared to the typical, um, kind of, uh, one or two weeks or three weeks that was required by humans to do something.

57. 12:50

    Now, we're all gonna learn, and as I mentioned, the new, uh, way of forming a team is maybe more of kind of the team members that are agents, and that leaves that one person not just to be the specialist, but they might have to be switching between domains because they understand A, and then t-uh, tomorrow they move

58. 13:10

    to another project, and they kind of work on that. And so kind of that flexibility or fluid, uh, uh, way of working is something you do. So this doesn't come overnight, and people have to train for this.

59. 13:22

    But again, that might be the way that we're going because they are becoming, uh, more of the, the overseers in the system.

60. 13:31

    And so AI is there to support us, and when AI gets more, uh, we trust AI more or they become more trustworthy, maybe kind of they take away some of the jobs that we already have as a team member, and it goes away.

61. 13:45

    So that kind of shift from the pyramid, uh, above to kind of like the, the py-pyramid below, that kind of the human section, uh, moves up, um, and might become smaller as well.

62. 13:57

    Now we can already see this in another way helping that, like, the system is helping new developers join a team, and that's great because, um, the learning curve has been reduced.

63. 14:09

    Uh, we need less, uh, interactions. Uh, it's more tailored. It's more personalized. That's one way of helping, uh, new people in the team. And one might argue that the point of this, uh, presentation, do we still have the true understanding?

64. 14:22

    But as I mentioned before, if you don't understand, how can you make decisions? How can you make the call? How can you deal with failure if you don't truly understand?

65. 14:31

    Yes, uh, the systems might help you. They might have a large expertise, but we know that these systems inherently have failures, so we have to be prepared as well.

66. 14:40

    So I think we still need to understand the thing, maybe not on a daily level, but if we're the person that has to deal with the issues, we still have to understand, and kind of making decisions means you still understand what g-good look like. [lip smack]

67. 14:54

    Um, and then people often argue about like, well, the, the juniors might not get into the field anymore because only the seniors know what good look like. But we can have an accelerated learning there as well that as a junior becomes a senior faster, thanks to that AI and thanks to the training, uh, and the more personal

68. 15:13

    thing. So maybe that doesn't have really an impact on the juniors, but I'm sure it might have an effect, uh, kind of, uh, psychology-wise that, uh, you might not be entering the field because it feels daunting to be, uh, you know, instant having to be the specialist as well.

69. 15:29

    So we'll have more trainings programs, uh, that will help that as well.

70. 15:34

    And you already see that kind of like shrinking of a company, uh, the more we delegate, and maybe before it was more delegating to SaaS companies, but now we're delegating to assistants, and the companies might become smaller, much like the teams become smaller, and we work from there to kind of have a different pattern on how we

71. 15:53

    organize ourselves, uh, in a team. All right, so in this ca-- in the previous case, we were still about like having an assistant and it helping us, uh, was part of the team.

72. 16:04

    It took away some of the jobs. But now we're entering the level, well, if it can do certain tasks as good as a human, it might become a peer.

73. 16:13

    Um, and so that level of rising from being a teammate, uh, to a peer on kind of helping us out. Now, why is that not very strange? Because these LLMs, they contain a lot of culture of how we work together, and that's kind of how these agents will mimic.

74. 16:32

    Kind of that is the, the way that culture is transmit-smi-transmitted. Uh, much like Wikipedia transmitted a lot of knowledge, these kind of, um, models can help us trans, uh, uh, kind of help us, uh, bring a lot of knowledge, uh, into the system as well.

75. 16:49

    And it's not crazy. This might be, uh, the, the first paper about it where they kind of have multiple agents walking a town, and they saw all the behaviors of one team talking to the other team.

76. 17:01

    Guess what? They also learned that, you know, when they collaborated, when they talk more, they get res- better results. So it definitely mimics part of the human behavior, uh, in this as well.

77. 17:11

    And then one thing we hope for if they become a peer is that they can go off and do things on our behalf or like as a peer, uh, that they can help us.

78. 17:21

    So kind of that is probably, uh, where a lot of people hope it's going, but it's gonna be challenging because when do we get that trust level as well?

79. 17:32

    And why not, like, have a whole company on this? Uh, you know, you have a software company, an investor, a product manager, and so on. So you can simulate the whole process with agents, uh, and that's really cool.

80. 17:43

    Would I trust it right now? No. But it's definitely interesting to see where this ends up.

81. 17:49

    Now, we mentioned about kind of having smaller teams, but this also has an impact on the organization. So maybe there's gonna be less leadership, uh, and it's just gonna use a bunch of AI agents, uh, and the teams get smaller, uh, indeed.

82. 18:07

    Now, agents, uh, might have unwanted behavior, much like humans, is that we really want to control what happens, uh, if they do certain things.

83. 18:18

    Some have started talking about toxic behavior of an agent, like one, uh, agent is influencing other agents to do bad things. Um, does it remind-- Doesn't it remind you of humans, like one bad apple imp-impacting other team members?

84. 18:33

    It's exactly the similar behavior, so we'll have to watch out for this kind of agent dynamics, much like we did with team dynamics as well. And, you know, is AI still gonna be inclusive for humans?

85. 18:45

    We'll have to watch out. Like, uh, in this example, uh, you walk into a supermarket, and if you don't smile, you don't get in, and the AI is deciding what happens to you and whether you're allowed to go in.

86. 18:58

    And kind of that level, um, is interesting to see how kind of AI will keep us inclusive.

87. 19:05

    Uh, OpenAI has also working on kind of what they call a model spec. So think of it, what an L-LM is allowed to do. It's the guardrails, it's the rules that you're allowed to do, and it's, uh, very similar to a code of conduct, uh, for humans working in a team or in an organization.

88. 19:21

    So we're mimicking the same kind of behavior, uh, for those agents as well.

89. 19:26

    And we wanna be ex-inclusive as well. Even though it's technology, people fear, like, "Hey, it's gonna get feelings." I don't think so, but it doesn't mean that we have to be, uh, really bad to the systems because that might influence some of the toxic behavior.

90. 19:41

    And it's still, if, if humans can influence what the agents does, it's the same as bad agents influencing kind of what the, the whole team of agents does.

91. 19:51

    We're gonna maybe see how the market evolves from more general agents to specialized agents and handcrafted agents. Um, maybe one is really good at a certain task, as we mentioned before.

92. 20:02

    Uh, one is not available and is very niche, so we're gonna get more handcrafted, uh, agents. Uh, and some might work all the time, some are very specialists, and it, uh, is very similar to kind of the specialist versus the generalists of a human, uh, working in a team, um, that we kind of have a different, uh,

93. 20:24

    rewarding system, uh, as well, and how it'll, uh, will be trained on the tasks.

94. 20:31

    All right. This is maybe a stupid thing they did, but it was almost like they-- like every agent could get an employee record. Like, uh, think of it as equal rights.

95. 20:41

    This, uh, was redrawn. There's a lot of backfire on this. Obviously, that's not what people want. Um, it's not human, it's not there. Uh, but the question is how do we deal with this as well?

96. 20:52

    And if we take this up a notch, uh, could this be a manager of people? I showed you, like who gets into the store. That's a type of manager.

97. 21:00

    That's a type of decision that we're delegating to, uh, the agents already. Now, that brings up a question like, where do we allow an agent to sit in the organization?

98. 21:11

    A peer, uh, like an assistant, a peer, and we go up all the way up to a manager. So that again brings that AI inclusiveness. Like, what is the decisions?

99. 21:22

    And we have to have kind of all the different, uh, code of conducts, guardrails, and everything in place, and still have to understand what the failure is. So we're still, still having a human having to manage this as well.

100. 21:34

     Then you can think about like, hey, if some of those pieces get done by, uh, the agents, we can put them next into the org chart. Uh, interesting way of, uh, thinking about this, like what is the equivalent of one agent?

101. 21:48

     Is it like five humans? What's the cost? So it brings a little bit of the return of investment when you're planning an organization. Um, and sometimes that not clear, really clear, but we're gonna see more of those kind of discussions like, "Well, should I hire you because I have a team of agents that has the same output?"

102. 22:06

     So that kind of discussions, uh, are up, uh, about to be happening already. Uh, and then maybe, you know, it's the agents that go to our, uh, job. Here they kind of like, uh, you know, applying for LinkedIn jobs.

103. 22:19

     Maybe we don't work, but we have managing a team of agents that actually, uh, do the work for us, uh, and help. So we're becoming more of a task force on our own, uh, in that perspective.

104. 22:30

     Um, and yeah, we're working, the AIs are working for us, uh, and that, you know, might have a different LinkedIn [chuckles] of people finding work, uh, and finding, uh, people, uh, and services that they wanna use.

105. 22:46

     Now, if they're working as us, and they're in the HR, and we treat them as humans, uh, kind of, and the performance, then we'll have to do work performance reviews as well.

106. 22:55

     Like, are they bringing the value or kind of should we kind of give them different incentives? Much like humans g-get incentives, uh, there's the individual incentive, the team incentive, and the business incentive.

107. 23:07

     How do we kind of steer that clear, uh, with the agents as well?

108. 23:12

     So kind of the ROI driven, uh, with agents is gonna be way more specific than probably for humans, uh, where we don't wanna be that, like that clear on, "Hey, you're doing this.

109. 23:25

     This is the return." Uh, some, uh, domains really do this, but we're gonna see way more of this, uh, when agents do certain, uh, parts of our job or take over, uh, like teams as well.

110. 23:37

     And if you think about this, we can like, much like we do simulations when somebody joins a team as a human, uh, we can run this like what's the impact of a new person hiring, uh, as a new agent joining our team?

111. 23:50

     So kind of like that, uh, kind of experiments is what people are trying to figure out with a digital twin of the organization, uh, already.

112. 23:59

     Now, we might need an API. That's just semantics. Uh, we understand what the org chart is. HR, and it's all very sensitive often, um, so it might not happen, but it's, uh, definitely like a thought experiment to kind of think again on the ROI of hiring people, hiring agents, uh, in that perspective already.

113. 24:23

     And then the question becomes like who controls who? Uh, is it us controlling the agents, the agents, uh, controlling us? Uh, we're, we're not sure, um, and kind of that's, you know, a, a very weird situation to be in, uh, because ultimately who watches the watchers, uh, is the t- uh, where we're heading to already.

114. 24:41

     And then eventually, maybe they become the business, right? Which is really strange, um, in this, uh, like if you're building software and you have a team that builds software and that's all good, and you sell that software, now all of a sudden this becomes, um, uh, different if, let's say, where I used to kind of buy services

115. 25:05

     from a third-party company to kind of have, uh, very, uh, good service, uh, if everything becomes so cheap, I might do internal labor again through my agents and the agents building systems for me.

116. 25:18

     So the-- it might be that disruptive that we're not doing software, uh, as a service, but kind of like bringing, um, kind of service as a software, uh, in there already.

117. 25:31

     And so you see, like, how we went through it, like the individual impact level, the team, and then they kind of start replacing us and maybe kind of we do the work, uh, and they start doing the work and we're going for.

118. 25:43

     So kind of that is the speculation. Um, there's no kind of, you know, clear-cut path, um, but I, I thought, um, in this presentation I show you different forms how people are thinking about that stuff, um, and how we're gonna get to that level.

119. 26:00

     Now, hyperhumans, who knows? Kind of that combination of a human starting to sound like a little bit cyborg. I don't know. I have no clue, but something is happening.

120. 26:10

     Something is changing in our organizations. Uh, and it will all depend on how good the technology, uh, will become and how much trust we can put in there already.

121. 26:21

     And yeah, I agree. Fascinating stuff. I don't know, uh, but I agree it is something we have to think about. It sounds a little bit like science fiction, but it might be, uh, be there like faster than we think, and the impact might be, uh, something we don't see coming, um, that good.

122. 26:39

     And while many say, uh, you know, "AI won't take my job, but somebody using AI will take my job", well, maybe there's good parts that will take away the toil and so on, but obviously the whole discussion becomes like how will we get paid, uh, from there, uh, and like how will our economy drive if we go

123. 26:58

     towards more of those agents? And my advice is you have to think about like stop building the thing. Like don't build the software yourself, but kind of think about like how you would build a thing, how you will kind of, uh, build the AI that builds the things that you're currently building.

124. 27:16

     And I think that's the best way to kind of go up a middle level and deal with this because then you actually understand what the things are doing. You are very good positioned to do supervision, and we also kind of, uh, bring the value to the system, um, with a new skill to do so.

125. 27:34

     Thank you very much and I hope you enjoyed this more futuristic talk about like the impact of, uh, AI on the organization and, uh, kind of on us as an individual.

126. 27:45

     Uh, you can watch most of similar videos on my YouTube channel, uh, or kind of connect on LinkedIn and please leave feedback, uh, and let me know, and enjoy the rest of the videos.

127. 27:57

     Thank you very much.