Scaling a company is a problem of information flow

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

1. 29:51

   so I've had to change a lot you also mentioned how like there are different stages of the company that you've already experienced the company was one way when you were 10 people then it changed and now you're 125 people or such yeah could you talk about the two or three different stages that have happened and what broke and what then changed in you a lot of it by the way has to do with physical office structure so like for example when you're four people you fit in a room and then a big stage that happens is when you can't fit in a room anymore like you just physically can't fit in one room well now there's an information because a company is like a bunch of people sharing information and when you don't fit in a room now you have to like how do you make sure everyone knows what's going on the next physical infrastructure thing that changes is like you

2. 30:31

   can't see each other like even if you can't fit you might have one office space it's like one giant thing so you can see each other but then at some point the office gets big enough where you can't all see everyone uh so now not only like you don't know what everyone's up to but also it's just harder to exchange information it's really about information exchange uh like information flow is probably like the biggest thing as a company scales there's also so much going on at the company now that like i don't want everyone to know everything that's going on so you like you physically can't fit into your context window or your memory right so and it's just not efficient so you need to like you have to make sure that only the information that someone needs goes to them and that's a really fascinating problem and i think ai can totally help with this

3. 31:08

   you know like everyone uses ai to be way more efficient on an individual level i don't think we've seen AI like really accelerate productivity on like organizational level and that's like a really interesting concept like how do you like for example like basically we would love to have a bot where I have an AI assistant and it tells me all the stuff I need to know so as CEO the stuff I need to know is a lot more like high level sometimes it should go low level for an engineer it might be like you know this thing for for a go-to-market person might be this thing but it's like some subset all the all the information of the company right now we use Slack and it's just like so many things going on so we're trying to figure out this. Excel is also probably one of the most AI native companies on the planet. How do you guys use AI operationally? So maybe starting

4. 31:46

   like what you just hinted, sounds like you use an AI assistant and it gives you exactly like on a daily basis or an hourly basis, this is the information that Will needs to consume right now. I do have this, but it just... doesn't it's not it doesn't work but why because it's missing context like and that's everyone's realizing the same problem that like there's so much context like if people are talking with each other in the physical world it's not being recorded like and if you miss that then like your summary is just gonna be off if it doesn't have eyes if your ai is only in the textual world then you're missing someone's emotion and maybe someone's not having a good day like all these things matter right so we're very ai native but like there are huge limitations and i can't we're not yet in a world where i could just get a summary and that's all i need to know i still

5. 32:26

   need to constantly walk around the office talk to people get a feel i actually will look at people's faces and be like just like see like is something wrong like i overhear conversations i'm like constantly like just walking around and and getting a feel for for the day it's probably slightly reassuring to hear that even an excel like company has not solved everything with the