Jerod Santo: Which means free Bitcoin mining. Yaaay! [laughter]

Divya Sasidharan: And then you’ll also get recommendations on YouTube things you should watch. They’re like “Watch this! It’ll tell you.” I think the one he was telling me about - he was like “Oh, did you know that Earth used to be an alien mining ground?” He was like, “Yeah, they found these giant tree stumps, and it’s obvious… Who cut them down? They’re too big, and trees don’t grow like that, so they must have at some point, and someone must have been around to cut them down.” I was like, “Whatever…”

Stefano Maffulli: [00:38:10.23] Yeah. And you want to have a simple way. Some licenses even have more specific wording around defining what source code actually means. The GNOME GPL is one of those. There are very clear descriptions and prescriptions about what needs to be given to users in order to exercise those freedoms, their freedoms as a user.

For AI it’s complicated, because there are a few new things for which we don’t even have – there are no court cases yet… I keep repeating the same story - when software came out for the first time [unintelligible 00:38:47.18] research labs, they started to become a commercial artifact that people could just sell. There was a conscious decision to apply copyrights to it. There was not a given fact that it was going to be using copyright, and copyright law.

That decision was a lucky one, honestly, or it was well thought out - I don’t know which of the two - because copyright as a legal system is very similar across the world. And building the open source definition, the free software definition, the legal documents that go with software for open source software and free software, those legal documents built on top of copyright means that they’re very, very similarly applied pretty much everywhere around the world. The alternative at the time were conversations about around treating software as an invention, and therefore covered by patents. Patent Law is a whole different mess around the world; the whole different applications they have, the whole different terms… Much more complicated to deal with.

So for AI, we’re pretty much at the same stage where there are some new artifacts. After you train a model, and that produces weights and parameters that go into the model, those models - honestly, it’s not clear what kind of legal frameworks apply to those things… And we might be at the same time in history where we could have to imagine and think, and maybe suggest and recommend what the best course of action will be. Whether it makes sense to treat them as copyrightable entities, artifacts, or nothing at all, or inventions, or some other rights, or exclusive rights.

And the same goes into the other big conversation that is happening already, but for which there is no – I don’t have a clear view of where it’s gonna end, is the conversations around the right to data mining. And if you follow the conversations around ChatGPT being sued by the New York Times and Getty Images, Stability AI [unintelligible 00:41:01.08] and GitHub being sued by anonymous etc, etc. a lot of those lawsuits hinge on what’s happening, why are these powerful corporations going around and crawling the internet, aggregating all of this information and data that we have provided, uploaded… We society. Some commercial actors, some non-commercial actors. We have created this wealth of data on the internet, and they are going around claiming it, and basically make it proprietary, and building models that they have for themselves. And on top of that, you can already start seeing “Oh my God, they’re gonna be eventually making a lot of money out of the things that we have created.” Or even more scarily - sometimes I think about this myself… I’ve been uploading my pictures for many years without thinking too much… So there is another base out there; I’m sure that someone has built another base out there with my pictures as I was aging… And now these pictures of me can be used, could be used by an evil and evil government, or an evil actor to recognize me around the streets at any time… [unintelligible 00:42:12.24] So is that fair? Is that not fair? Those are big questions, and there is no easy or simple answer.

Adam Wiggins: [01:21:49.13] Because your kids are all grown up and out in the world doesn’t mean you don’t still love them and have emotional reactions to things transpiring in their lives, right? Yeah, for sure, that that one was tough for me to see, because it was part of our original vision, or somewhere in that early vision, that it’s quick to get started. And it’s not about it being free exactly. It’s that you don’t need to make too many decisions to start a new project, and get it on the internet. And that’s something that – actually, I think open source was the thing that really kind of opened my mind to that originally, which is… I remember years ago when you wanted to - or at least my early experience with SQL databases was things like Informix, where you needed to go get a license before you could even use the software. So if you have an idea for a project late at night, you’re not just going to like spin it up out of nowhere. There’s this whole ritual that is involved. Now there’s the money, of course, and then that leads you into “Okay, is this project really important enough to spend the money?” and like needing to frontload that decision. So for sure, being able to like just get something up and running quickly and frictionlessly, and the free is part of that… So yeah, I was quite sad to see that.

I have a lot of sympathy for the business decision, because having been on the inside of that machine, even all those years ago - I’m sure it’s changed a lot since then, but it is a huge amount of… “Work” doesn’t even capture it. “Responsibility”, “stress”, I’m not quite sure. But being responsible, or a steward in some way for all these applications is just this huge job, and there’s really a lot of clutter with the free apps. And I probably, looking back, would have done it a little differently. Actually, we did make some big changes… This was kind of right around the time I left… Actually, I’ll give a shout-out to my colleague, Peter van Hardenberg, who’s now running the Ink & Switch research lab… But he was one of the ones that helped lead the project of setting up these hobby dynos, and something where there was like a little bit more restrictions on the free stuff… So if you’re like “Okay, yes, you get started with a quick side project, but if you really are going to make it be a longer-term thing, and you’re getting value from it, you should pay some amount. It doesn’t have to be a lot; just a little bit.” And that’s a way – it’s less about the money in the sense of like paying for the runtime or whatever, and it’s more about just sorting out what’s really important to people and what’s not. And that sorting is important, because you end up with all these junk apps that were like one-off experiments, or someone just kicking the tires, or whatever, but you can’t tell those apart from the ones that are important to people. And so it makes just the running and the maintenance of the platform a lot harder.

And then, you throw in the whole like spam and abuse side of things, which we always knew in the beginning would be a challenge… But of course, this is long before there was such a thing as cryptocurrency mining. So in today’s internet - I mean, I don’t know what the trust and safety team or whoever it is there that needs to deal with this, but I can only just can’t even begin to fathom how much work it is to kind of filter and manage just this totally free, open-ended runtime.

So I wish there had been a different solution; I’m sure they’ve thought about it and came up with what made the most sense, because I can see something there being a big problem for the business, and that problem impacting other users and customers, right? Badly-behaved apps in some ways sucking up all the time and attention of the team that’s running the thing, and then they have less time to support the people who are really getting value from the platform. So I see why the free apps would be a problem. But at the same time, it was a core part of the vision, which is that it’s completely frictionless to get started, and this takes that away.

Adam Stacoviak: Yeah, you’re doing some cool stuff in the public too, especially with Spiral. We’ll talk about some of the stuff that you’re doing in the open, the Bitcoin wallet of course, mining Bitcoin, the hardware aspects… But this aspect of embracing open source software, embracing really the public aspect of getting that feedback loop which I think is pretty interesting. So feel free to pepper that in as we get closer to it, but I know that you mine Bitcoin; do you do anything developery today? Like, if you’re hacking today, how would you describe a hack that you might do? Not so much today specifically, but today in terms of a timeframe.

Adam Stacoviak: Yeah, you're doing some cool stuff in the public too, especially with Spiral. We'll talk about some of the stuff that you're doing in the open, the Bitcoin wallet of course, mining Bitcoin, the hardware aspects… But this aspect of embracing open source software, embracing really the public aspect of getting that feedback loop which I think is pretty interesting. So feel free to pepper that in as we get closer to it, but I know that you mine Bitcoin; do you do anything developery today? Like, if you're hacking today, how would you describe a hack that you might do? Not so much today specifically, but today in terms of a timeframe.

Adam Stacoviak: I was thinking about the last 100 years, because I’ve found this interesting… My hometown where I grew up was a coal miner town. I grew up in Pittsburgh, Pennsylvania… And very common for the idea of a company house to be a thing. A coal mining plant would establish its routes, have its company there, they’d have a company store, they’d have a company house, the entire neighborhood essentially is owned by the company, and the employees get a house for free, or paid for by their wages/work…

Joseph Jacks: …and in order to upgrade the software, you would have to order, and they would have to ship you a new series of floppies, from which you would need to kind of like physically install, and you would own those physical like pieces of technology that would give you an upgrade. And then the pricing model was all built around that, too. It was perpetual licensing, where you would like basically buy that version, and then you would have maintenance that you’d pay for the version, and in order to upgrade a new version of the software you’d have to physically spend a new upfront amount for the new version, and then do that over and over again.

[34:18] Most of the – I would say probably 99% with some interpolation synthesis, mixing and matching of ideas… Most of the categories in software, including SaaS, which have become a little bit more like fractal domain-specific… Cal.com has almost like one feature, and wow, it’s a billion-dollar company. That just happens because of economies of scale, and humans using more and more technology. But most of these ideas were implemented in the seventies, eighties and nineties, even in the mainframe days. Scheduling, booking things… I mean, this is like calendar functionality.

I was recently watching the oral history of Mike Markkula, who was the first CEO of Apple. He was a guy that Sequoia founder, Don Valentine, introduced to Steve Wozniak and Steve Jobs to basically help them with the business plan, and like run the company. And Mike was also a really brilliant programmer, and he wrote Fortran and basic programs that would actually run on the early Apple computers. And the collection of those programs came with a credit, sort of a copyright credit at the bottom, which was Johnny Appleseed. You might remember this from the early Apple computers, this Johnny Appleseed name. So Johnny Appleseed was Mike Markkula. And Mike in the computer history museum describes this – it’s on YouTube, you can watch it. It’s a really great, multi-hour oral history. Mike actually personally wrote the programs for the applications that would ship in the early Apple computers. So including the calculator, calendar app, bookkeeping app… A lot of these basic – a very basic, primitive word processor…

And so I just think back to like the most fundamental, basic things that humans do as part of their work or their lives have been re-implemented and re-implemented so many times in the history of technology, going back to literally the fifties and sixties, that when we have new ways of distributing and innovating in the core technology, what happens is you just reinvent and you re-implement what happened, what came before, with shinier and newer, sexier, more sort of societally-relevant and interesting abstractions. And then those things end up taking off. And oftentimes people are given a choice. There’s sort of like the free version of the thing, maybe it’s open source and it’s like less sophisticated, but you can totally control it… Then there’s the proprietary version of the thing, which is much more polished and sexy and packaged. This is so fascinatingly – for the first time, and sort of skipping to the AI stuff, we’ve seen this incredibly shockingly surprising almost acceleration of equivalency. This actually happened this week - I mean, the latest LLaMA models are state-of-the-art. And by all accounts, the benchmarks are pretty much indistinguishable, and the capital required to produce that is similar. Mark Zuckerberg literally did an interview a couple of days ago saying they’re spending tens of billions of dollars building LLaMA, and they’re giving it away completely. And in fact, they’re making – Yann LeCun was just tweeting about this the other day… They’re making their licensing even more permissive. You can do distillation, quantization… You can use the outputs from LLaMA models to train other models… It’s almost like an Apache or MIT-level permissiveness. The artifact is very different for a neural network that has massive amount of information compression. Not in the Pied Piper sense, by the way, but in the neural network, AI, tokenization of –

Jerod Santo: Gatsby’s was even a little bit weirder, because it wasn’t just that, but it was like, “As your site grows, Gatsby gets slow. If you pay us, we’ll make Gatsby faster.” That was even weirder. So it was like, it should already be just the way Gatsby works… But it required this very complicated caching solution that was all in their proprietary code, which would just make Gatsby work the way it was supposed to, I guess, in the first place… Anyways, it was a strange thing.

A lot of these I think it’s too early to tell. Deno KV is out there, Deno is still alive and well, but are they making money off of it? Who knows…? Next is a little bit different, the way that its relationship is with Vercel, in terms of it not being – maybe it’s the same. I don’t know, it seems a little bit different. It’s kind of like the cart and the horse, which one came first. Or that’s the chicken and the egg. I don’t know, I’m mixing all my metaphors. But I tend to agree. I wonder how attractive this is to enough people for it to move the needle, I guess, for Fred and the Astro team… And it seems like it’s an experiment. And sometimes you have to experiment when you’re trying to find ways for monetizing that makes sense. And so I wish them luck with this. I would tend to just reach for something that’s just open source, and maybe I sign up for a service for it, like in terms of a database. But I get it, where it’s batteries included, it’s like tightly integrated with Astro if you’re already using Astro, it’s the Easy button, generous free tier, so I can kind of get started with no money down… The way they’re planning on charging is based on storage, so it does do kind of that dynamic scaling, to where when you’re not using it, it scales down, serverless style, like Neon and others… So that’s cool. But then they charge you per gigabyte of storage, basically. That’s the idea. So in Nick’s case, if he was using it, and all these hackers came by on his PHP script, and just shoved comments into his AstroDB like crazy, he’d be paying out the wazoo to these folks. But if you’re a better programmer than Nick, you might be able to avoid that fate.

Kris Brandow: Yeah. Alright, so I think that’s a pretty good overview of the value propositions that conferences have had historically. I think the big kind of headline I’ve had in my mind around this for a while is “Is it valuable for me or for any attendee to fly to a place to mostly sit in an auditorium with a whole bunch of people, potentially risking my health for a talk I’ll be able to see on YouTube in three weeks?” Not literally three weeks, but roughly three weeks.

[18:12] And I think for a lot of people the answer to that is probably no, but I do think the other part of conferences that has already existed, which is kind of what you brought up, Angelica, of this - I get to interact with all of these people I wouldn’t be able to interact with otherwise. I think that is a thing that is very unique to conference spaces.

I think my own experience at GopherCon this past year, and GopherCon in previous years, and other conferences, has always been this very strong enjoyment of what has in the past been this kind of shadow track of the conference, but more and more conferences are now making like an official track, which has been aptly named the hallway track. And for those of you out there who have not been to a conference, or don’t know what a hallway track is, it’s basically literally people, instead of going to a particular talk, one of the talk tracks, stand in the hallway, or in the sponsor area, or dining den, or wherever, and have chat with each other, and get to know each other, and do more interpersonal interaction, instead of watching the talks. Some people do this for just like an hour of the whole conference, some people spend their entire conference doing this. But I think that when it comes to value propositions, this is the largest that I think conferences can offer in this current era that we’re in, of this interactivity, that you just really can’t get anywhere else. Like, yeah, we do have video calls… Once again, they’re not the same. There was a difference in quality of personal interactions when we had – I’ll just keep using GopherCon, since this is a Go podcast… But GopherCon online, versus GopherCon in-person this last year, right? There was a very large difference in those interactive aspects of things.

So with that, I want to move to the second part of the podcast today, where we talk about this idea I have that I’ve called hallway conf… Which is not a name of an actual conference, but a generic name for a kind of flavor or type of conference. So there are some kind of big features that I’ve been thinking about, that I think – these would help transition us from a regular track-based/talk-based conference into a hallway-based conference. So I’m just gonna go through a couple of these, and then we can talk, and panel, feel free to choose individual things you want to talk about; it doesn’t have to be on this list. But I think the biggest thing, or one of the biggest things is I think instead of having a few large theaters for talks - I think most conferences have like two or three tracks; some are single-track, but it’s usually you have one room you can fit the entire conference in, and then you have some other rooms you can fit roughly a third or half of the conference. So instead of having two or three of those, having a larger number of much, much smaller theaters. And to tack on with that - you know, you’re flying the speakers out; I think what you said, Angelica, is absolutely accurate, that people really want to talk to the speaker. They don’t just want to hear what they have to say. I can hear what they have to say on the YouTube recording. Yeah, I’m at the conference because I want to be able to talk to them and interact with them. And now we have the technology that we can easily pre-record all of the talks. So instead of having the speakers give live talks, have them give pre-recorded talks, and then show those talks in the aforementioned theaters.

And the nice thing about this as well is that now you can schedule a talk to be aired more than once, which means that someone doesn’t have to choose between two live talks they want to see. You can be like “Oh, I want to see all these talks”, and you could figure out a schedule so that you can see the talks that you want to have.

And since we still have the speakers there, this kind of third part of this main change is to have a speaker stand, where speakers would have shifts, or time slots where they would be in this den. Say maybe two-hour blocks, or one-hour blocks, where people who have watched their talks can come and talk to them about their talks, and have these nice little kind of smaller interactive spaces for them. And since we have all the talks recorded, you could also do something where, for attendees, you can put the talks online early, so that people can watch them, and have watched them, and then spend most of their time talking to different speakers about the things that they said in their talks.

Christine Yen: An artifact of entering this round in a more position of strength meant that we could do things like put constraints on the timeline… So raising this round, we were able to constrain it to really only three or four weeks of uncertainty before we signed something and moved on. And I will say that in those three or four weeks, my brain was just completely fragmented. I was capable of responding to things and charging forward and kind of executing on what needed to be done, but I basically apologized retroactively to everyone in my life who was not part of the fundraising process, or not relevant/tangential to it, because every free moment my brain got, it was seeing if I got an email, or if I could go back and tweak something with the deck, or incorporate this piece of feedback I got.

I will say that personally and physically, COVID has helped me establish a lot more healthy habits, working from home, which has been a surprise silver lining in the pandemic… And I think all of that was necessary and helped, even in this fragmented fundraising period; making sure that I was walking the dogs and thus getting my exercise, and eating relatively healthy. It’s easy, and it is sometimes still glamorized to skip those things in the founder world, and I am glad to be at the phase of my life where I recognize maintaining this meat machine of mine is a necessary part in my brain working well.

Nader Dabit: Some of the DAOs allow you to just join for free, some of them ask you to mint some type of token… But either way, you have to have, for the most part, some type of token to get in. And the reason that that’s the case is to kind of give ownership to everyone that’s there, and also provide them a way to liquidate that ownership.

So for a token, you can basically sell that, either on some digital exchange, or maybe like an NFT would be like OpenSea, or some type of NFT marketplace… But yeah, most of the time you can get a token or buy a few ERC20 tokens, or work for those tokens by participating in the Discord, and then you can participate in that way.

For the Developer DAO, which is the DAO that I have created for developers, the token is completely free. All you have to do is just go to the website and mint the token. And you do have to pay the Ethereum transaction costs. When you pay that cost, you get that token sent to your wallet.

And really, the reason that there is any type of cost at all is to kind of have some type of way to prevent civil resistance… Because if this token is gonna be worth money at some point, what’s to stop someone from minting all 8,000 of those for themselves? Like, how do we equally distribute ownership between a large number of people? Well, there is some type of way that we can do that with needing someone to mint.

But what I’ve been doing is actually setting aside a few thousand of my own dollars, and just minting those and sending them to people around the world that can’t afford even the $20 minting cost. But that way I know that that person is real, and that they’re kind of in it to be part of the community, and they’re not just in there to kind of speculate.

Nick Fitzgerald: That’s it, right? So what compilers do is within the memory they’ll lay out where structures are, the same way they would lay it out in native memory. But if I’m a Rust program and I wanna talk to a C program, C doesn’t really understand Rust data structures. Rust can kind of talk at the C level, but it’s a little bit painful… And now we’re talking about interacting with JavaScript, but [unintelligible 00:30:22.15] but JavaScript running inside SpiderMonkey, a JavaScript engine that’s on WASM… So the structures are just totally foreign, and communicating is really difficult. But this is where another WebAssembly proposal comes in called Interface Types. Interface types are basically a type grammar, sort of similar to Web IDL, where Web IDL defines “Here’s records, here’s different string types etc.” Interface types defines a similar type grammar, and then the idea is you’ll be able to communicate like this.

Interface Types isn’t quite an IDL, in that the plan is eventually you’ll be able to kind of have dynamic adapter functions that allow you – say you’re receiving a string over this Interfaces Types boundary that lets modules communicate… And Rust’s strings are represented as UTF-8; so it’ll wanna just lower that into a flat UTF-8 array. But JavaScript strings are much more complicated. In SpiderMonkey I think there are – I forget if it’s 7, 9 or 13, but it was one of those odd numbers like that. Different representations of a string. Maybe the coolest one is called a rope. A rope is basically a tree of little strings. The idea is if you concatenate two strings and your string representation is just a buffer, you kind of have to potentially grow a new buffer that’s the size of both of them, and then you have to copy the strings in, and it’s really expensive. That’s [unintelligible 00:31:57.27] operation.

[32:01] But if you have a rope, what you can do is you can just say – you know, it’s kind of a tree, so you have a node that’s just “I am the concatenation of this one string and this other string, and creating that is order one.” So it’s very cool, but it’s very complicated. But interface types kind of will eventually allow you to define your own ways to kind of lower the platonic ideal of a string down into a rope, or something like that. Kind of like arbitrary computation for translating these types on either side.

So that’s kind of like the furthest vision. But right now we are defining just what’s called a canonical ABI, which fixes the representation; you have to use a string buffer, or something like that. There’s one representation for each type.

So with just a canonical ABI it is kind of just like an IDL, but this it’s open to that next step once we ship the first phase… So this is gonna allow all of these modules to talk to each other. And each of these modules - what’s really key about interface types is that they’re kind of [unintelligible 00:33:04.14] So if you think about npm modules, when you use an npm module, it gets all the same permissions and capabilities that your application has. And this is a problem; we’ve seen these supply chain attacks, where some generic markdown library or something - I don’t think it’s actually happened with a markdown library, but… You know, I just do something very innocent, and then actually I’m reading your SSH keys from disk, and I’m sending them off to some server, or I’m mining Bitcoins, or whatever… And so it’s not great. We talked a bit earlier in the podcast about capabilities and how a WebAssembly module can’t do anything unless you explicitly give it something to do. So Interface Types kind of preserves that ability between different WebAssembly modules. It says “Just because I can read to the disk and I’m talking with you and I’m using your markdown library doesn’t mean you can talk to the disk. All you can do is take this markdown [unintelligible 00:34:06.00]

Dave Cheney: I’ll give you another example of where something could be overused. In my first talk at dotGo I wanted to talk about Rob Pike’s functional option pattern, and how I applied it in the design of some code that I was working on. Two years later a number of people had said to me “Look, I really like that talk, but I’m not sure that to other people I can propose these ideas. It’s too weird. Like, this idea of anonymous functions that return anonymous functions - it’s too weird.” That was kind of the impetus for another talk about “Hey, free functions are a thing. We should use them.”

But it’s a realization that that is probably – like, if everything in your codebase is functions that return functions that call functions, that’s a really big innovation in tech. You’re mixing that with heavy channel use and heavy goroutine use - that’s a pretty pro level move. That would be one of those cases where you seem to have used all the features of the language. Isn’t that gonna be a little bit hard to follow?

Dan Abramov: [01:28:18.20] Yeah, I mean, you can do it this way, it’s just a painful way to migrate. I think they now have it in the docs that the recommended way to migrate is you just take your entire pages directory… You know, in the pages directory, the way it was structured is you have this get server side props, so get static props. This is the part of the file that executes on the server only. And then you have this page component in the same file. So previously you were mixing server and client code in the same file.

And so in the app directory, the recommended migration is that you take this page component, you move it to another file. So that is the start of your existing tree. You put use client at the top of just that file; you don’t need to add it to any other files. You just put it into your entry point. And then your get server-side props - that is what now becomes a server component. So server components are really – like in Next.js, they serve the same function as get server-side props in the old Next.js, as like Astro templates. It’s just like this part that executes first. And so you really need ‘use client’ only in client components that are imported from a server component. It’s the place where things turn into JSON. It’s where, if you pass props down, they will become JSON, get passed over the network, and then they will be revived.

So I think that is how to think about migration, is that you don’t – I think people think of it as a per file, kind of intuitively. They think “Oh, I need to open all files, and put ‘use client’ on the ones that are client.” But it really, it’s more like you have the entrypoint into your import tree, so you have your page component or something liek this, you put ‘use client’ on top of that, and that’s actually enough. And then maybe later you kind of pull some things out of it.. Maybe you have some data fetching deep down, that you’re like “Oh, actually, I’ll just do this fetch above in the server component, and I’ll pass the result as a prop.” And so you kind of gradually can pull some things out of it, out of the client part to the top, but you don’t need to add directives everywhere.

Nikita Shamgunov: Yeah, there’s a lot of known algorithms, and they’re mostly caching algorithms. So it’s already happening today a little bit in Postgres; there is a buffer manager, a buffer pool… I’m maybe mixing SQL Server and Postgres terminology here, because my background is SQL Server. But the architecture is similar, where the buffer pool has a counter for every page, and it refreshes the counter of the pages touched… And then the algorithm kind of sweeps the cache and decides which pages haven’t been touched for a while, and then evict them from the cache.

Here we’re adding another tier, in the remote storage. We also track pages, and you see which pages have been touched recently, and which have not been, and then you offload those pages onto S3. There is a caveat, however; S3 does not like small objects, and a page is 8 kilobytes. So we need to organize those pages into some sort of data structure that will bucket those pages together, so when we throw those pages onto S3, we throw a bunch of them together in a chunk. That data structure is called an LSM Tree, and that’s the implementation of LSM tree that we built from scratch in Rust, and that’s integrated with S3 offloads older data to S3. It’s kind of like several use cases. One use case is a very large database; if you have a very large database, chances are large portions of that database are never even touched. So over time, some of that data - maybe it’s the data from like, I don’t know, five years ago, and you don’t really need it, but you’re keeping this there because it doesn’t cost you much, and it’s better to have them for occasional use that not have at all, or put them in a different system.

And the other use case is you have a big fleet of databases; a lot of them are scaled down to zero, because you just have them for occasional usage, and now if you keep them hot, that will start to add up both on the compute side, and on the storage side. Storing all that data into SSDs is a very different economics than storing all that data in S3 in a compressed form. So these are the second place, where integration with S3 can drive much better economics.

Chris Anderson: And I kind of had to agree with them on both counts. So I come back, and I had to bribe them with ice cream, and I come back and I thought “How could that have gone better?” Clearly, we needed a cooler robot and a better flying plane. And I thought “Well, what if the robot was the plane? What if it would just fly itself? That would be cooler.” So I googled “flying robot.” If you google “flying robot”, it turns out that you get drone. That’s what a drone is, it’s a flying robot. I’m like “Got it. Huh. What’s a drone?” If you google “drone”, it’s like “it’s an aircraft with an autopilot.” Like a plane with a brain. I’m like “Yeah, okay. What’s an autopilot?” If you google “autopilot”, it’s like “sensors, and compute, and software…” I’m like, “That’s kind of what’s in the LEGO Mindstorms box.” It came with gyroscopic sensors, accelerometers, magnetometers, a Bluetooth connection which would connect to a GPS… I’m like “I think it’s kind of the same stuff that’s in this box.” So around the dining table we stuck together the bits, and had a nine-year-old program the code… And we had a LEGO autopilot. And then we stuck it in the plane, now retrieved from the tree.