Hank: The key to happiness is low expectations is what I always tell everybody. 

Gonto: Exactly I think it's the same of a product. The key to happiness is low expectations. The key to having a great company is to provide low expectations. 

Hank: Okay, we're talking about Devin, Dinners, and Raycast today. And to kick off on Devin, we had a couple things. You had an interesting article that you saw. I watched. Like a 40 minute primogen video about it and of course this thing's been going on for a few months. It's kind of interesting to get an update on where it is.

Where do you want to start? 

Gonto: Let's start with do you remember the week when Devin went out? Everybody was blown away there were a lot of memes about the CEO of Devin winning some math contest I don't know if you remember but he was a genius and made doing math out of the blue and then the week after that there was a hacker news video that was showing how the launch video was full of shit and that you can't actually do anything about it.

After that, I heard that they worked on improving it and then I know they recently launched and now people are using it. And yeah, what's, what's your take? 

Hank: So they launched that video and then everyone was like, it's a lie. It's a scam. How did they raise from these VCs? VCs are stupid, and whatever people just say, this is kind of a whole side note because Primogen also went off on the VCs about how dumb they are and all that.

This is Founders Fund, which is like Peter Thiel's thing. Brilliant engineer. 

Gonto: He’s not stupid.

Hank: He's not stupid. The other main backers are the Stripe founders, the Collison brothers. Not stupid. Another one of them was like, Elad Gill, brilliant engineer as well. There was another guy who I think had good credentials. And the founders also are, well…a note about VC, because I see this all the time, and sorry I'm taking a son of a rabbit hole here.

But I've just got a rant for all the engineers listening. Engineers always criticize VC from a perspective of how could they fund that product? VCs don't fund products. They fund founders, at least the best ones do, and they fund the founders who they think have enough grit and intelligence to like, work their way into a successful product.

And this is, this is a case of that, I think. You know, one of the, I think the CTOs scale AI engineer, like they all have credentials in ML and AI. 

Gonto: CTO, CEO, all new AI, they were really good at math, they were research scientists. I think it was a good bet, like as you said, you bet on the founders, you bet that they'll be able to do what they say they will, and maybe they will, maybe they won't, but I think it's a good bet because remember that for a VC, for a company to do well, it has to be worth more than one or two billion, like it has to become a unicorn.

And for them, if a company is okay ish, which is like 100 mil, they don't care. So they have to go for big swings, and this was definitely a big swing. 

Hank: This is the biggest swing, and now this brings me into Prime's video. Like, Prime had a great critique, which made me think, Oh, I gotta talk about this. Prime had this great critique on, Hey, Devin, why are you putting it out there that this is going to just straight up replace software engineers and all that?

When instead, you could say, “this will make all your engineers 10, 20, 30, 40 percent more effective, more productive.” Because any company would pay 500 a month to make their 150 to 300, 000 a year engineer 20 percent more productive, right? 

Gonto: But I don't think they could have done that in this case, because  Devin's product is not that.

Devin's product idea is it will go on its own and then eventually send the PR so you merge it. And that to me is not really making engineers 20-30 percent faster. Like in this case, I think they claim they chose was the right one. It's maybe just that the product is not on par with the claim. 

Hank: Yes, they had to put the full vision out there.

So I'm actually with you. And I think if you go out the gate saying we're here to make engineers 20, 30, 40 percent more productive, you will lose. You won't get the founders fund funding, like founders fund led the round and you won't get all that hype and it's going to hurt for a few years. 

Like Devin's going to not hit any of those original claims, but because they set their vision and the bar, they have a shot at actually hitting that. That's totally different. And you think about who were they actually marketing to? They're not marketing to engineers who they want to eventually replace. They're marketing to the VCs, to the people who fund Founders and have the biggest vision.

Gonto: I still think that's a mistake. Like, I think they could have done it better. Like I would pitch this as you don't need to do all of those like shitty and boring bug fixing. Devin is a, it's your companion damn junior engineer. Who can fix maybe some bugs, fix this thing, fix this other. And then I think, like, it would still be interesting, it would still be interesting for engineers.

And it would still be interesting, I think, for VCs and similar. Because it is starting with junior engineers to eventually move to something else. To me the problem is also that they couldn't accomplish that. Like, there were two tweets about it, but, um, Answer.ai basically wrote a blog post about this, about saying how the problem with Devin wasn't that he didn't do what he was supposed to do.

It was the amount of time it made them lose to understand where they've been blocked, why, and what the fuck happened. And to me what's mind blowing is, how bad does a product have to be so that two companies, because I saw two blog posts, two companies go above and beyond to actually post about how bad using a product is.

Like, it's mind blowing. Like, doing a case study is hard. And most companies are like, okay, this is not good. You just ignore it. But doing a blog post, analyzing the data, writing it, it's so much effort. Like, why would they do that? Like, what's your take on that? 

Hank: To me, it's an engineering blog, you know, and when there's a lot of hype and attention on a thing and you're excited to try it, you want to write about it.

A huge portion of engineers will do that. There are lots of great engineering blogs out there. To me, this just feels like such an attractive case of, hey, this is a moment for us to capture some hype, to show that we're trying new things, and also maybe to, you know, ride the wave. You've talked about this before, whenever there's kind of like a wave of attention on something, if you can ride that a little bit, it's great.

So to me, that just feels like this. 

Gonto: How about you say that there's two things that come to mind. One is, this is a great way to recruit people because it's like we're using the latest and we're trying anything, even if it's shit, but we're trying everything. And the other thing, to me at least, is either they lost a lot of time by doing this.

So if they lost a lot of time by trying to practice, by seeing what was fucked up or not working, etc. They were just pissed that not only did it not save time, but it made them lose time. The other thought that I had was, is it related to pricing? Like, this is 500 a month, which is not expensive to be honest, if you compare it to a lot of other services, but because it's replacing or giving you this junior engineer, you get the perception that it's more expensive than other services when actually it's, it's not.

Hank: It's probably still cheaper than most interns, like summer interns you'd get, and it's going to get better over tim, you know, at an interesting rate. So, let me ask you this. With what we know now, if you were tasked with marketing Devin from the outset, what changes would you make? 

Gonto: I would honestly do like a mea culpa blog.

I would start with, “we're sorry we fucked up.” Like we pitched that it was going to replace engineers and everything. And we think we can still do that in five, six years. But right now it is a dumb junior engineer. And that's still something that you want to use and something that you need. And then you go over some of the claims of what was broken to say either how you could have done it correctly or how you're working on fixing it.

But I would treat this as like a post mortem when the service goes down, like your server goes down and you do a whole post mortem on what's happening. Like that to me is the best because you own it, you double down on your vision, but you say it's in the future and you change your messaging for the present, which might actually make people expect less of it and therefore value what it's done, which maybe it's not that bad.

It's just a comparison of reality with expectations. 

Hank: What if we could go back in time? How would you change it from the outset?

Gonto: I think you still pitch the vision, but again, I do think you need to pitch like, this is a dumb junior engineer and this is how you use it. Like, I think I saw their blogs yesterday and I was like reviewing a bit and I don't think it's clear enough on how to use it, what tasks are good, how you give it to them and what to do.

And because of that, I think people were left alone and like, you try for whatever you want. And eventually that's what didn't work. Um, yeah, I thing one. Thing two was at least in the blog post they say that they worked well for two cases, three cases, but it didn't work at scale. So maybe have they tried at scale or like maybe they could have started with smaller companies instead of somebody like let this be or I don't know.

What would you have done? 

Hank: I think actually a good model to follow and maybe people would disagree with this would be Tesla's self driving marketing. Because it started as like, okay, we've got advanced dynamic cruise control. We're gonna add lane assist. And the whole time, Elon is saying, Eventually, this car, you're gonna get in it, and it's gonna drive you from A to B.

And we're not there yet, but, they set expectations fairly correctly at the outset of that Tesla marketing, and I think it could do the same. Hey, here's our grand vision. We can replace engineers up to some level. And Devin can be a member of your team. You can have multiple Devins, whatever. Here's where we're starting.

And here's our plan to get there. And of course, that plan is gonna be accelerated against what the reality's going to be. But that's how I would've approached it. And I think the video, you know, You could be honest about like, Hey, we've edited out some parts with this and all that. Because then you don't blow the trust.

Because that's the hardest thing to get back now, is putting the tube back in the toothpaste. Like, if you did a mea culpa vlog now, it still doesn't undo all the trust you've broken, but it would help. But now people are going to dismiss Devin for years, even when it gets good. Even when it's, you know, as good as a junior engineer.

People will say, oh that whole thing was a scam, and it will take them years to catch up. 

Gonto: Yeah and I love the idea of the Tesla. I think the Tesla did well and that’s where I think Elon will have a lot of visions. Some will work, some will not, but because some of them worked, people don't remember the ones that didn't. As long as you mention them as a vision and not something that is happening in the next year or six months or something like that. 

Hank: And actually Elon's done this in another place too, like SpaceX is maybe even a better example.

He's like, our vision is to get to Mars, but we're going to start here. And then we're going to land and catch rockets and do all this stuff. And like, we're all amazed. If he had told us we were going to Mars, by 2020, we would all hate SpaceX now, but  he was careful about that. Maybe that's the lesson here.

Gonto: Everything is about what's the difference between expectations and reality. Not only applies to work, also applies to life. I think it applies to everything. 

Hank: The key to happiness is low expectations, is what I tell everyone.

Gonto: Exactly. I think it's the same of a product. The key to happiness is low expectations. The key to having a great company is provide low expectations.

Hank: Let's talk about dinner. 

Gonto: Yeah. So there was a tweet from Paul, um, Klein, and he talked about how they did like a dinner for folks in this space. And what he was saying is that most people are doing it incorrectly. And what he talks about these do not do like first interest or icebreakers. Do research on who's coming and assign the seats.

And then just make it useful for people instead of your sales speech. I really liked his approach, but I think you can do even more than this. I would love to hear about you. Like, have you done dinners? What's your take on doing dinners? What's your take on Paul's advice? 

Hank: Yeah, I think it's generally good advice.

I think you can do intros in a good way. I mean, the picture he posted, there's like 20 plus people there. So, it makes it difficult. And, yeah, I think there are things you can do. I think generally it's good advice. And what's interesting is, they do lots of events and meetups. And in San Francisco, there's always something to go to.

So if you're inviting people to something, you don't want them to regret going to your thing instead of another thing. And you want to be interesting. You want people to build relationships at your thing. And you want to think about, what's the vibe and the relationships that people have while they're at your event? But yeah, I know you had more thoughts.

Gonto: Yeah, and I agree with those. I think like on dinners, what I would say is doing, as Paul was saying, like a more casual dinner without a pitch, without anything else typically works really well. What I've seen that really works is actually inviting, not just prospects, which most people do, but also prospects and customers, and I've done something where I sit like.

Two prospects on the left, one customer on the right, then two prospects on the right, one customer on the left. And the idea of that is that your customers are actually selling to your prospects. And this is good in every way, on the sense like great customers get a free one, two michelin star dinner, prospects gets to the same.

And then if your customer sells to the prospects. You can offer them something like a free trip to a conference, some free discount for their company or something like that depending on what they care about.Bbut I think that's thing number one. And then what paul doesn't talk about that I think is important is: how do you use the dinner for something else?

And something I've done that is a bit creepy, but works really well is: let's say it's 20 people, in the 20 people I would sit four people from the company inside. So then among 20 people, there's typically four to five conversations. And the idea is that you have one of your employees in every one of those conversations and they're taking notes.

Like they take notes from what was chatted about, what was interesting, what was learned, who said something, et cetera. And then after the dinner ends, you can do two things. One is you follow up with everybody, not to sell them, but rather to give them thoughts on what was chatted about. Not just in your area of the table, but actually in all of them, what did they learn and why?

And then people like to increase their ego. So then on those people, you can ask them like, Hey, you had a really great quote here. Can I use it? Or something like that. And people will say yes. And then you can share some of those quotes in LinkedIn, in Twitter, in social media. And people will associate then that person to your brand, which I think is fantastic for your brand if it's a known person.

So I think those are the type of things that I think you can do to make it better. Last thing that I would say is never do sales speeches, I agree with Paul. The only thing that might be interesting is bringing somebody that people admire. You're doing an AI dinner and you're inviting Greg, like the CTO of OpenAI.

Like, sure, people will be happy for him to pitch or talk in the beginning something. Um, but I think it's a different format, but that's my only rant. It's like, people only think about the dinner. People do not think about what extra things they can get from the dinner, like blog posts. Like sending summaries, like quotes, et cetera. And they're equally or actually more important than the dinner itself, just based on the reach you can have. 

Hank: Yeah, the point of the dinner is to build rapport and relationships. So if you don't follow up on that, you've totally blown it. But then I like your points about preparation. So, I mean, if I were to put it succinctly, you know, I like bullet lists, so if I thought of this in terms of bullet lists and expanding Paul's, say one, brief your team, you know, tell them, Hey. This is your job, this is where I want you seated, and have a piece of paper out, or your phone, and take notes. And tell people, oh, I'm just taking notes, I thought that was awesome that you said, because then they don't think you're rude and texting or whatever.

Two, you can arrange seating. You know the guests that you invited, and you should be able to look them all up, and all your employees should know something interesting about them. And in fact, I like your idea of like, out of every four or five people, there's one employee. You should give that employee the assignment of like, Hey, find these four people and have them sit next to you.

Tell them, Hey, you four, you're part of my like little group here, sit next to me. And this is why I thought you four would be interesting to chat together. You know, give, give them a reason why you're doing it. And then, three, follow up. You have to, you have to send the follow up, use the notes, have a debrief.

Not just, oh that was an awesome dinner, good job guys. There should be a meeting about the dinner that happens afterward, and who's following up with who. Hey, is that person ready for a pitch, or do you just want to send them a nice note, or tweet at them or something casually? 

Gonto: I think it's important not to sell to everybody that came to the dinner. It's, you sell to those that are ready. For the rest, you follow up with conversations that happened, you invite them to another dinner, you build rapport in relation to what you're saying. But I think understanding that is key. 

Hank: I will say, on the pitch, you don't have to pitch, but you do have to tell people who you are and what you do.

And why you brought them there. If I were Paul at Browserbase, I'd say, okay, as a reminder, Browserbase does this and you're all interesting. And I thought we would have a nice dinner because you're all interested in this. And so that's enough for me to spend my VC money on you and not get you know, slammed by the board, and that's all I'll say. No pitch, enjoy. Right? And then you seem very magna… magnanimous and you know, it's not too salesy. 

Gonto: So I like that idea. I've never done it like that. How I've done it a lot of times is when I assign these employees in these groups, or four or five peoples, there's always opportunities to share how you do things or how you operate in that, you can actually explain a bit of how the company works, just so that people understand your anecdotes.

And it is a bit more organic way, as part of the conversation, to also include what you do and include the topic. 

Hank: Yeah, I like that. 

Gonto: Let's move to our last topic, which is the Raycast video or ad. Um, do you want to talk about it, Hank? 

Hank: Yeah, so I saw this from a couple different places. Raycast did this really well produced ad, video, whatever, they put on Twitter.

And it shows this woman basically just struggling to focus. And IlLoved it. I think a lot of people loved it. And I think the reason is because everybody who works from home or has ADHD or has any sort of anxiety or imposter syndrome which is like, you've just checked the box on pretty much all the engineers who will see this video. They all can relate to this woman.

And then she has this beautiful solution. She pushes the command to start Raycast and then the room descends into darkness and it's just her and her work and it gets done and it's beautiful and it's such a relief. Like there's this nice emotional arc that everybody relates to. It's beautiful. 

Gonto: I agree and they did it all in a minute and what I liked about it is it's an emotional video. Like it's…the video works because of empathy because this is holy shit is exactly how I feel and it focused and I can't. So in that sense, I also thought it was fantastic. 

I do have one critique on it, which was interesting I loved it so much actually that I sent it to a bunch of people. Some of them used Raycast, others didn't. But from a lot of the people that I sent this their main question was that “wait how is this different from Do Not Disturb from Mac?”

The Mac already has a focus mode. And that's true, and that's not explained in the video. So it's, people saw the video, they loved it, and they thought it was a competition or the same as the focus in Apple, which is not the case, because the focus on Apple will disallow notifications for, from arriving to you, but it will not block apps or websites.

What Raycast does is, you cannot open the app, or you cannot open the website, so it's a lot more. And I think it is a missed opportunity on the sense that they did a fantastic emotional video, people related to it, but then a lot of people discarded it without even looking at the docs just because they thought it was something that it wasn't.

So maybe in this case, adding a bit of logic, a bit of the features, which I'm sure they didn't add on purpose, I think would have helped, um, in the end. So that more people will have tried it out. 

Hank: Yeah, there's a world that could have incorporated that into the ad. 

Gonto: Exactly. 

Hank: If she had tried, you know, the Mac focus mode or whatever it's called. I'm a Windows user, sorry. And that failed for her somehow. 

Gonto: Exactly.

Hank: Then that would complete that. What I was also thinking about was, if this had been designed by a different team, I think what most people would have done, and maybe there is a Raycast video of this somewhere, they would have talked about the importance of focus, how we've all heard the stat of if you get distracted, it actually takes you 45 minutes to get back into focus and everything's distracting you.

Like they could have just made a video ad that lists features and talks about that stuff. And it, it could have been effective, but not nearly as emotionally impactful. Right. And this is something we've talked about. 

Gonto: This was much better than a feature video. I would just have found the way as you mentioned on incorporating this, mostly because all of, actually for now, all of, Raycast users are Mac users because there's no Windows version yet. So understanding that would have been important.

Thank you everybody for coming today as always. If you have feedback, let us know and if not, we'll see you next week. Thank you!