WEBVTT

NOTE Sentence-level transcript of https://www.youtube.com/watch?v=LZuWZRze3MU

NOTE One cue per sentence. Cue ids are the line anchors on /transcripts/LZuWZRze3MU.html. A cue ends where the next begins, or 2 s after its last word.

s1
00:00:01.309 --> 00:00:03.309
[music]

s2
00:00:12.760 --> 00:00:14.040
Good morning.

s3
00:00:14.040 --> 00:00:15.960
I know a lot of you in this room.

s4
00:00:15.960 --> 00:00:17.120
It's great to see you.

s5
00:00:17.120 --> 00:00:20.920
Welcome to the voice track at AI Engineer World's Fair.

s6
00:00:20.920 --> 00:00:24.560
For those of you who don't know me, my name is Quinn La Holman Cramer.

s7
00:00:24.560 --> 00:00:26.000
I work at a company called Daily.

s8
00:00:26.000 --> 00:00:31.040
We make developer infrastructure for real-time audio, video, and AI.

s9
00:00:31.040 --> 00:00:37.080
And we're the team behind Pipe Cat, which is the most widely used framework for building voice agents today.

s10
00:00:37.080 --> 00:00:39.720
Pipe Cat is open source and vendor neutral.

s11
00:00:39.720 --> 00:00:48.320
It's used by companies like AWS and Nvidia and Anthropic and thousands of startups and scale-ups and enterprises.

s12
00:00:48.320 --> 00:00:58.480
And today I'm going to talk about what kind of agents we're building today, including voice agents, but not just voice agents, and what I'm interested in building next.

s13
00:00:58.480 --> 00:01:06.080
And I'm going to try to put all this in the context of the roughly 80-year history of digital computing so far.

s14
00:01:06.080 --> 00:01:07.600
So, we've got a lot to cover.

s15
00:01:07.600 --> 00:01:09.760
We're going to go fast.

s16
00:01:09.760 --> 00:01:19.920
But we're going to start in 1945 with an essay called As We May Think, written by an engineer, an academic, a civil servant named Vannevar Bush.

s17
00:01:19.920 --> 00:01:26.960
Bush deeply understood technologies ranging from analog computers to photography to radio to radar.

s18
00:01:26.960 --> 00:01:31.600
As We May Think is a extraordinary piece of writing.

s19
00:01:31.600 --> 00:01:45.040
The essay predicts the development of, among other things, document display on a screen and document scanning and OCR and speech-to-text and text-to-speech and programming languages and hypertext and search engines and data networks.

s20
00:01:45.040 --> 00:01:53.800
Something like the GoPro camera, something weirdly like the Amazon Kindle store, and voice interfaces and brain computer interfaces.

s21
00:01:53.800 --> 00:02:03.240
And I've been thinking a lot about As We May Think lately because Bush wrote this essay right at the very beginning of the computing age.

s22
00:02:03.240 --> 00:02:09.600
And I think it feels to most of us like we're working right at the beginning of a new age, the intelligence age.

s23
00:02:09.600 --> 00:02:13.200
So, what will we build?

s24
00:02:13.200 --> 00:02:19.040
Well, at the moment we're building agents and we're having a lot of fun doing it.

s25
00:02:19.040 --> 00:02:28.000
And a lot of the AI engineering work we're all talking about this week is focused on building a full coherent software stack for AI agents.

s26
00:02:28.000 --> 00:02:38.360
Here is Satya Nadella talking a couple weeks ago on a crossover episode of the No Priors and Latent Space Pod about the challenges of building agents in 2026.

s27
00:02:38.360 --> 00:02:39.040
That's sort of

s28
00:02:39.040 --> 00:02:39.400
That's right.

s29
00:02:39.400 --> 00:02:49.000
So, so in some sense you kind of want to harness to define the models, the the data, uh and the tools.

s30
00:02:49.000 --> 00:02:51.800
And so that you have a loop across those three.

s31
00:02:51.800 --> 00:02:56.680
And so what we are trying to first of all make sure is each of our products that we build, right?

s32
00:02:56.680 --> 00:03:04.120
Whether it's GitHub Copilot or the security copilot the stuff we showed with M dash or even the discovery for science, it doesn't matter.

s33
00:03:04.120 --> 00:03:14.920
All of them are multimodal harnesses um with tools access so that you can do this progressive uh disclosure of tools even so that they're token efficient.

s34
00:03:14.920 --> 00:03:20.040
Uh and then you're feeding it with very rich context.

s35
00:03:20.200 --> 00:03:31.400
So, if you were here last year at AI Engineer World's Fair, you could draw a through line from the things we were talking about last year to loops and tool calls and context engineering and the stuff we're focused on this year

s36
00:03:31.400 --> 00:03:33.800
to some emerging ideas.

s37
00:03:33.800 --> 00:03:37.880
Uh you can hear that in Nadella's clip just there.

s38
00:03:37.880 --> 00:03:47.560
I think of this is kind of agents plus plus, like multi-model harnesses and software copilot embedded in every single piece of software, organization level harnesses.

s39
00:03:47.560 --> 00:03:54.200
So, how do we go from agents to agents plus plus to the next thing beyond agents.

s40
00:03:54.200 --> 00:04:03.920
Well, the last time we had this kind of massive change in how we write software and what we write software for and to do was the early days of the World Wide Web.

s41
00:04:03.920 --> 00:04:06.120
And I was around for the early days of the World Wide Web.

s42
00:04:06.120 --> 00:04:15.200
I was a baby programmer in 1995 and the thing we talked about all the time in 1995, the way we talk about agents today, is web pages.

s43
00:04:15.200 --> 00:04:27.480
I spent a lot of time writing HTML by hand and building web server software in C and indexing and search software in C and authoring tooling and management infrastructure for web pages in Pearl.

s44
00:04:27.480 --> 00:04:34.160
I was as excited about HTML in 1995 as I am about agents today.

s45
00:04:34.160 --> 00:04:37.800
And the web page is still with us and it's still important and useful.

s46
00:04:37.800 --> 00:04:44.280
But today we talk a lot more about web applications and native mobile applications than we talk about web pages.

s47
00:04:44.280 --> 00:04:57.680
So, just like we went from web pages to full-blown web and native mobile, clearly we're going to chart a path to a new fully AI native software that comes after agents and agents plus plus.

s48
00:04:57.680 --> 00:05:02.480
So, let's keep going back in time to in order to think about this future.

s49
00:05:02.480 --> 00:05:14.600
Here's a timeline Vannevar Bush lays out and as we may think, he talks about the abacus, which was both an immensely useful device for doing practical everyday mathematical calculations

s50
00:05:14.600 --> 00:05:22.560
and also an incredibly important theoretical tool that led to ideas like numeric place value and the concept of zero.

s51
00:05:22.560 --> 00:05:33.640
And Bush talks about the massive jump from the abacus to the state-of-the-art electromechanical keyboard calculating machines that he had in 1945.

s52
00:05:33.640 --> 00:05:48.040
And then he posits that we're about or he is about to witness and help create an equally large leap to what he calls the arithmetical machine.

s53
00:05:48.040 --> 00:05:56.080
And then he goes a step even further than that and he invents or designs in that essay a device he calls the memex.

s54
00:05:56.080 --> 00:05:59.600
And we have a little bit of an advantage over over Bush in 1945.

s55
00:05:59.600 --> 00:06:02.200
We've seen 80 years of computing play out.

s56
00:06:02.200 --> 00:06:04.000
So, we can modify his timeline a little bit.

s57
00:06:04.000 --> 00:06:15.000
We can go from the abacus to the stored program computer to 40 years later the personal computer and 40 years after that this AI agents era that we're all collectively

s58
00:06:15.000 --> 00:06:18.960
helping to invent and create and bring into being.

s59
00:06:18.960 --> 00:06:28.800
So, the question for me is what did we build to go from those very first digital computers in the 1940s to the personal computer in the 1980s?

s60
00:06:28.800 --> 00:06:36.200
Well, in the 1950s the big job was to figure out more effective ways of transmitting human intent to these new computing machines.

s61
00:06:36.200 --> 00:06:37.880
We built the first programming languages.

s62
00:06:37.880 --> 00:06:40.400
We wrote the first compilers.

s63
00:06:40.400 --> 00:06:51.480
And the the theoretical underpinnings here were figuring out how to combine the elegance of mathematical formalisms with something a little bit more like natural language.

s64
00:06:51.480 --> 00:06:56.680
And then building on that in the 1960s the challenge was to make these machines interactive.

s65
00:06:56.680 --> 00:07:01.840
Make these machines capable of a two-way dialogue with humans.

s66
00:07:02.120 --> 00:07:08.880
The '60s also saw the birth of graphical programming with systems like Ivan Sutherland's Sketchpad.

s67
00:07:09.240 --> 00:07:12.800
And the '60s were an amazing era for science fiction.

s68
00:07:12.800 --> 00:07:19.760
Even though almost nobody had access to a computer the computer became a big part of the popular imagination.

s69
00:07:19.760 --> 00:07:24.160
The idea of a computer really resonated with people and ideas matter.

s70
00:07:24.160 --> 00:07:29.360
For example, here is the idea of the computer in Star Trek.

s71
00:07:29.360 --> 00:07:31.320
Put her on record.

s72
00:07:31.320 --> 00:07:33.320
Recording.

s73
00:07:34.360 --> 00:07:36.360
Come.

s74
00:07:36.880 --> 00:07:39.640
Captain's log supplemental.

s75
00:07:39.640 --> 00:07:45.880
Engineering officer Scott informs warp engines Can be made operational and re-energized.

s76
00:07:45.880 --> 00:07:49.360
Computed and recorded, dear.

s77
00:07:49.880 --> 00:07:51.960
Computer, you will not address me in that manner.

s78
00:07:51.960 --> 00:07:53.720
Computer.

s79
00:07:53.720 --> 00:07:57.000
Computed, dear.

s80
00:08:00.880 --> 00:08:04.360
I love the background sound of punch cards going through a punch card reader.

s81
00:08:04.360 --> 00:08:11.120
So, like you know the computer is working even though it's talking to you about what it's actually computing.

s82
00:08:11.120 --> 00:08:22.960
There were of course a bunch of other talking computers in in science fiction of the '60s and the next year after this, the Kubrick movie that was a interpretation of Arthur C. Clarke's 2001: A Space Odyssey had the HAL 9000 computer.

s83
00:08:22.960 --> 00:08:29.600
This is a much, much more dystopian view of a talking computer than the Star Trek computers.

s84
00:08:29.919 --> 00:08:38.880
And by the 1970s, computers had become powerful enough that the next big job was designing abstractions that could scale to much larger amounts of data and much more powerful

s85
00:08:38.880 --> 00:08:40.159
computing substrates.

s86
00:08:40.159 --> 00:08:46.240
We got relational databases, which introduced new theoretical underpinnings for data manipulation.

s87
00:08:46.240 --> 00:08:50.720
And we got declarative languages, which leveraged those new theoretical insights.

s88
00:08:50.720 --> 00:08:55.880
And programming languages in general continued to evolve in what to me at least are really amazing ways.

s89
00:08:55.880 --> 00:08:59.600
We got Smalltalk and object-oriented programming in the '70s.

s90
00:08:59.600 --> 00:09:03.120
And all of this set the stage for the personal computer in the 1980s.

s91
00:09:03.120 --> 00:09:05.440
The Macintosh shipped in 1984.

s92
00:09:05.440 --> 00:09:08.080
Windows 1.0 shipped in 1985.

s93
00:09:08.080 --> 00:09:12.280
And Microsoft's mission statement was a computer on every desk and in every home.

s94
00:09:12.280 --> 00:09:17.000
And incredibly, Microsoft delivered on that mission statement.

s95
00:09:17.000 --> 00:09:24.920
And we got a computer on every desk and in every home because these new personal computers delivered real, amazing, tangible benefits.

s96
00:09:24.920 --> 00:09:29.120
Take VisiCalc, for example, which was the first spreadsheet program.

s97
00:09:29.120 --> 00:09:42.520
A truly new abstraction for doing computation, numerical computing, two-dimensional, interactive, so durable and so useful that probably most of us in this room use a direct descendant of VisiCalc

s98
00:09:42.520 --> 00:09:47.480
regularly, Google Google Sheets or Microsoft Excel or whatever.

s99
00:09:49.200 --> 00:09:53.720
Or put another way, this was a spreadsheet in 1957.

s100
00:09:54.200 --> 00:09:57.560
And this was a spreadsheet in 1985.

s101
00:09:57.560 --> 00:10:06.440
And I think a lot about VisiCalc these days too because I think VisiCalc is an example of how transformative new technologies can be in the way of delivering a capability

s102
00:10:06.440 --> 00:10:11.680
that used to require a lot of specialized people and specialized knowledge and making it generally accessible.

s103
00:10:11.680 --> 00:10:20.120
And I think VisiCalc is a potentially a counter-argument to the argument or the fear or the concern that AI is going to lead to mass unemployment.

s104
00:10:20.120 --> 00:10:22.960
Because VisiCalc didn't put accountants out of business.

s105
00:10:22.960 --> 00:10:27.640
Instead, it made much much much much more accounting-like work possible.

s106
00:10:27.640 --> 00:10:38.480
And it made new categories of work possible that we couldn't even really conceive of when a spreadsheet or doing a screen's worth of calculations as we think about it today

s107
00:10:38.480 --> 00:10:41.920
took a roomful of people.

s108
00:10:41.920 --> 00:10:54.160
So if we were here in the Moscone Center in 1985 and these two interfaces, the Macintosh System 2 and Windows 1.0 were state of the art, what would we have said the world would look like in 10 or 20

s109
00:10:54.160 --> 00:10:57.560
or 30 or 40 years?

s110
00:10:57.560 --> 00:11:01.520
Well, we actually have a really great example of a prediction from that time.

s111
00:11:01.520 --> 00:11:10.400
Like as we may think another famous document in the history of human computer interaction, concept video from Apple made in 1987 called Knowledge Navigator.

s112
00:11:10.400 --> 00:11:14.160
This is very much worth tracking down online and watching all of if you haven't seen it.

s113
00:11:14.160 --> 00:11:18.200
I'm just going to play about 20 seconds from the middle.

s114
00:11:22.800 --> 00:11:24.839
You have three messages.

s115
00:11:24.839 --> 00:11:28.920
Your graduate research team in Guatemala, just checking in.

s116
00:11:28.920 --> 00:11:36.440
Robert Jordan, a second semester junior, requesting a second extension on his term paper and your mother reminding you about your father's

s117
00:11:36.440 --> 00:11:40.120
surprise birthday party next Sunday.

s118
00:11:40.800 --> 00:11:55.760
So, the video shows a foldable tablet, a touchscreen interface, [clears throat] a conversational voice assistant with a really strong personality, access to both global and personal information, real-time video generation, real-time computer vision, seamless video call integration,

s119
00:11:55.760 --> 00:12:02.280
delegation of complex tasks for autonomous execution, and what we might call today continual learning.

s120
00:12:02.280 --> 00:12:06.240
And it's really, really clearly influenced by AS we may think, but it's also quite different.

s121
00:12:06.240 --> 00:12:17.280
It really is updated for 40 years of progress, and it really does sort of presage this AI agent era we're in now in a way that Vannevar Bush's Memex didn't and maybe couldn't.

s122
00:12:17.280 --> 00:12:25.160
The Knowledge Navigator video divides our timeline, I think, quite neatly in half, and hold that thought cuz we're going to come back to it.

s123
00:12:25.160 --> 00:12:37.040
The 1990s were about the network, first local area networks and dial-up, and then the internet and the web, and with the benefit of hindsight, I now think that the single

s124
00:12:37.040 --> 00:12:41.880
most important thing about the web was that it was multimodal from the very beginning.

s125
00:12:41.880 --> 00:12:52.600
More even than the GUIs of the 1980s, the web anticipated that text and audio and video and data were not different things to be used in different programs, they belonged

s126
00:12:52.600 --> 00:12:53.520
together.

s127
00:12:53.520 --> 00:13:03.400
And in a real sense, the web was an attempt, and a conscious attempt on the part of a lot of people building the web to make that Knowledge Navigator video real.

s128
00:13:03.800 --> 00:13:14.800
Then in the first decade of the new millennium, the big job was to make all of this computing stuff mobile and continually connected, to put this new multimodal networked computer

s129
00:13:14.800 --> 00:13:21.920
in your pocket, literally, to give a supercomputer to everybody in the world that they could carry around in their hand.

s130
00:13:21.920 --> 00:13:29.800
And as with the 1960s, there was an efflorescence of like futurism on screen in the first few years of the new millennium.

s131
00:13:29.800 --> 00:13:41.160
And I think it was because computers you could carry around with you and cameras everywhere and a kind of Moore's law for pixels making screens super cheap really gave us a chance

s132
00:13:41.160 --> 00:13:45.800
to think through what we thought the future would look like in a new way.

s133
00:13:45.800 --> 00:13:52.560
A lot of stuff we could almost but not quite build was cohering in the minds of people working on these machines.

s134
00:13:52.560 --> 00:14:00.880
And the best and most famous Hollywood computers from that era were created by John Underkoffler [clears throat] for the films Minority Report and Iron Man.

s135
00:14:00.880 --> 00:14:04.040
Here's Minority Report from 2002.

s136
00:14:05.520 --> 00:14:07.720
It's no longer there.

s137
00:14:07.720 --> 00:14:08.440
Time frame?

s138
00:14:08.440 --> 00:14:09.840
13 minutes.

s139
00:14:09.840 --> 00:14:12.040
Hey Chief, investigator from the Feds here.

s140
00:14:12.040 --> 00:14:14.160
Yeah, I don't need some twink from the Fed poking around right now.

s141
00:14:14.160 --> 00:14:15.240
John, I wrote it down on your calendar.

s142
00:14:15.240 --> 00:14:16.440
I left you a message at your house.

s143
00:14:16.440 --> 00:14:18.800
Check in with the Favors ahead of Ford and see if the neighbors knew where they went.

s144
00:14:18.800 --> 00:14:19.600
Check all relations.

s145
00:14:19.600 --> 00:14:20.600
Check the neighbors and relations.

s146
00:14:20.600 --> 00:14:21.320
But John

s147
00:14:21.320 --> 00:14:21.800
It's actually

s148
00:14:21.800 --> 00:14:22.960
Just get him some coffee.

s149
00:14:22.960 --> 00:14:23.954
Tell him some stories how I save

s150
00:14:23.954 --> 00:14:23.960
[music]

s151
00:14:23.960 --> 00:14:25.720
your ass every day and you can't do without me.

s152
00:14:25.720 --> 00:14:26.440
I got coffee.

s153
00:14:26.440 --> 00:14:27.200
Thank you.

s154
00:14:27.200 --> 00:14:28.200
Danny Witwer.

s155
00:14:28.200 --> 00:14:29.960
Twink from the Fed.

s156
00:14:29.960 --> 00:14:30.640
Oops.

s157
00:14:30.640 --> 00:14:32.640
Gone.

s158
00:14:32.760 --> 00:14:39.560
So the gestural interface in Minority Report was implemented on screen as special effects, but it was actually based on John's PhD work at the MIT Media Lab.

s159
00:14:39.560 --> 00:14:42.440
In a real sense, this was real technology.

s160
00:14:42.440 --> 00:14:48.440
John had brought the UI out of the small screen and into the world with us in a bunch of really interesting and lovely ways.

s161
00:14:48.440 --> 00:14:56.920
John also consulted on Iron Man, which is a very different view of the future than you Minority Report, which was Spielberg working in like the American Kubrick dystopian tradition.

s162
00:14:56.920 --> 00:15:00.920
Iron Man is really squarely in that Star Trek goofy futurist tradition.

s163
00:15:00.920 --> 00:15:05.320
But I think you could see the common elements in the UI depicted on screen.

s164
00:15:05.320 --> 00:15:08.480
It's still from the same era.

s165
00:15:12.080 --> 00:15:13.640
Wake up, Daddy Sean.

s166
00:15:13.640 --> 00:15:14.390
Welcome home, sir.

s167
00:15:14.390 --> 00:15:15.320
[music]

s168
00:15:15.320 --> 00:15:17.240
Congratulations on the opening ceremonies.

s169
00:15:17.240 --> 00:15:18.320
They were such a success.

s170
00:15:18.320 --> 00:15:20.285
As was your summit hearing.

s171
00:15:20.285 --> 00:15:20.400
[music]

s172
00:15:20.400 --> 00:15:26.240
And may I say how refreshing it is to finally you in a video with your clothing on, sir.

s173
00:15:27.920 --> 00:15:29.760
You!

s174
00:15:29.760 --> 00:15:31.880
I swear to god, I'll dismantle you or short your motherboard.

s175
00:15:31.880 --> 00:15:34.800
I'll turn you into a wine rack.

s176
00:15:37.520 --> 00:15:43.680
I co-founded a startup with John in 2006 to make the Minority Report interface into a commercial product.

s177
00:15:43.680 --> 00:15:49.105
This is our demo reel from 2012, 6 years into that work.

s178
00:15:49.105 --> 00:15:51.105
[music]

s179
00:15:53.690 --> 00:15:55.690
[music]

s180
00:17:10.760 --> 00:17:20.680
This long project to build the multi-modal, multi-device, multi-screen, multi-player, ubiquitously connected computer is still what I'm working on 15 years later.

s181
00:17:20.680 --> 00:17:30.760
In 2010s, we built out the cloud, which laid the groundwork for the infrastructure and data centers and data capacity we would need to scale up AI training and inference,

s182
00:17:30.760 --> 00:17:33.400
which brings us to now.

s183
00:17:33.400 --> 00:17:34.880
We're building agents.

s184
00:17:34.880 --> 00:17:38.160
And we're starting to think about agents plus plus.

s185
00:17:38.160 --> 00:17:47.680
But I think we can also start to think about the next thing, the AI native software that is to agents what today's internet is to the web pages of 1995.

s186
00:17:47.680 --> 00:17:49.480
And one way to think about the story is this.

s187
00:17:49.480 --> 00:17:54.200
We went from the calculator to the computer to the personal computer to the global cloud computer.

s188
00:17:54.200 --> 00:18:04.360
And now we actually have the ability to build the Memex and Jarvis from Iron Man and knowledge navigator from that 1987 video for real, completely working.

s189
00:18:04.360 --> 00:18:09.080
And by building those things, we'll figure out what we want to build next.

s190
00:18:09.080 --> 00:18:17.160
A couple of weeks ago the team at Tavis released a reimagined knowledge navigator video this time entirely built on real and available technology.

s191
00:18:17.160 --> 00:18:26.120
I'll just play another 20 seconds of this, but like the original knowledge navigator video, this is worth tracking down and watching in full.

s192
00:18:31.360 --> 00:18:32.480
Good evening, Hassaan.

s193
00:18:32.480 --> 00:18:35.440
I adore that houndstooth jacket you're wearing today.

s194
00:18:35.440 --> 00:18:38.880
Anything I can help with or would you like to review tomorrow's schedule?

s195
00:18:38.880 --> 00:18:40.040
Thanks so much, Tom.

s196
00:18:40.040 --> 00:18:44.200
Yeah, let's review tomorrow's schedule and see how busy it is.

s197
00:18:44.200 --> 00:18:45.960
Opening your calendar now.

s198
00:18:45.960 --> 00:18:48.400
Here is the quick version since it is late.

s199
00:18:48.400 --> 00:18:50.240
Tomorrow morning is slammed.

s200
00:18:50.240 --> 00:18:55.720
Investor meeting at 9:00, then back-to-back one-on-ones and internal meetings until 1:00.

s201
00:18:55.720 --> 00:18:59.080
So, the full 4-minute video is one take, completely real.

s202
00:18:59.080 --> 00:19:09.360
And when you watch it, it really does feel both like the knowledge navigator video from 1987, familiar but built on real technology, and like something brand new.

s203
00:19:09.360 --> 00:19:20.120
And I'll just close with a massively multiplayer game project I've been working on with some friends as a canvas to really think about what AI native software can be.

s204
00:19:20.120 --> 00:19:25.000
This game is built from the ground up with LLMs as the core of every interaction.

s205
00:19:25.000 --> 00:19:32.880
At every moment in the game, there are hundreds of inference calls happening and we couldn't have built anything like this even a year ago.

s206
00:19:32.880 --> 00:19:35.280
Oh, sorry.

s207
00:19:36.480 --> 00:19:39.013
Welcome to Gradient Bang, a multiplayer game

s208
00:19:39.013 --> 00:19:39.120
[music]

s209
00:19:39.120 --> 00:19:42.960
that showcases real-time agent orchestration.

s210
00:19:42.960 --> 00:19:50.080
Gradient Bang demonstrates several patterns for AI sub agents such as asynchronous non-blocking context compression.

s211
00:19:50.080 --> 00:19:51.520
Okay, make a note for later.

s212
00:19:51.520 --> 00:19:54.800
We are going to eliminate Heliotrope from existence.

s213
00:19:54.800 --> 00:19:55.360
Noted.

s214
00:19:55.360 --> 00:19:57.353
Long-running sub agents that share context.

s215
00:19:57.353 --> 00:19:58.080
[music]

s216
00:19:58.080 --> 00:20:00.840
Eagle is on five trade loops.

s217
00:20:00.840 --> 00:20:03.903
Hawk and Raptor are on five exploration loops

s218
00:20:03.903 --> 00:20:04.000
[music]

s219
00:20:04.000 --> 00:20:04.760
each.

s220
00:20:04.760 --> 00:20:06.720
Your fleet is busy.

s221
00:20:06.720 --> 00:20:08.000
Progressive skills loading.

s222
00:20:08.000 --> 00:20:10.440
How much does your average ship cost?

s223
00:20:10.440 --> 00:20:11.108
Ships range quite

s224
00:20:11.108 --> 00:20:11.320
[music]

s225
00:20:11.320 --> 00:20:12.280
a bit, Captain.

s226
00:20:12.280 --> 00:20:14.320
Dynamic user interface generation.

s227
00:20:14.320 --> 00:20:16.080
Show my task history.

s228
00:20:16.080 --> 00:20:17.640
Certainly.

s229
00:20:17.640 --> 00:20:19.840
Uh hide the map.

s230
00:20:19.840 --> 00:20:20.520
Okay.

s231
00:20:20.520 --> 00:20:20.934
And conversational

s232
00:20:20.934 --> 00:20:21.280
[music]

s233
00:20:21.280 --> 00:20:21.760
voice.

s234
00:20:21.760 --> 00:20:23.640
No, I don't want to exchange it.

s235
00:20:23.640 --> 00:20:26.480
I just want to sell it for cold, hard cash, please.

s236
00:20:26.480 --> 00:20:26.829
I'm afraid

s237
00:20:26.829 --> 00:20:27.040
[music]

s238
00:20:27.040 --> 00:20:31.120
the galaxy doesn't allow you to be shipless and hitchhike.

s239
00:20:31.720 --> 00:20:37.680
So, I went long after wrap-up, but I will say that the first version of this new draft talk was an hour.

s240
00:20:37.680 --> 00:20:41.400
So, I have a lot more things I'm super excited to talk about with all of you.

s241
00:20:41.400 --> 00:20:44.080
So, if you are interested in this stuff, come find me.

s242
00:20:44.080 --> 00:20:45.840
We have a booth on the show floor.

s243
00:20:45.840 --> 00:20:53.760
I'm online everywhere and I'm excited to build agents cuz agents are awesome, but also to build the next next thing, too.

s244
00:20:53.760 --> 00:20:54.994
Thank you.

s245
00:20:54.994 --> 00:20:56.994
[applause]

s246
00:21:12.024 --> 00:21:13.600
[music]

s247
00:21:13.600 --> 00:21:15.600
Mhm.