[00:00] (0.40s)
Travis, your thoughts on this category?
[00:02] (2.80s)
Anything come to mind for you in terms
[00:06] (6.48s)
you know, feature sets that would be
[00:08] (8.72s)
extraordinary here? I know you you like
[00:11] (11.04s)
to think about products and the consumer
[00:12] (12.56s)
experience.
[00:14] (14.88s)
It's really interesting. So, you know,
[00:16] (16.96s)
I've been spending, as you guys know,
[00:18] (18.48s)
I've been spending my time on real
[00:20] (20.56s)
estate and construction and robotics.
[00:23] (23.04s)
And so I' I've been out of the this kind
[00:25] (25.20s)
of consumer software game for a long
[00:27] (27.20s)
time. But it's super interesting over
[00:28] (28.72s)
the last six months. There have been a a
[00:32] (32.08s)
number of consumer software CEOs.
[00:36] (36.24s)
Like when I hang out with them or
[00:37] (37.60s)
whatever, they're like, yo, how are we
[00:40] (40.40s)
going to how are we going to keep doing
[00:42] (42.00s)
what we do when the agents take over?
[00:44] (44.88s)
Yeah. The paradigm shift is so profound
[00:48] (48.24s)
that the idea that you would visit a web
[00:50] (50.00s)
page goes away and you're just in a chat
[00:51] (51.68s)
dial. you know, you have an agent that's
[00:53] (53.20s)
just taking care of your flights for
[00:56] (56.08s)
So, I I kind of I I think there's a
[00:58] (58.56s)
leaprog over that. I think
[01:01] (61.60s)
it's just like you tell something, yo, I
[01:03] (63.76s)
want to go to New York. Can you you
[01:05] (65.84s)
know, I'm sort of looking at this time
[01:07] (67.28s)
range. Can you just go find something
[01:09] (69.36s)
I'm probably going to like and give me a
[01:10] (70.80s)
couple options?
[01:12] (72.48s)
Yeah. And it's just a whole you have an
[01:14] (74.80s)
interface and then you know is perplex
[01:19] (79.28s)
is this thing that you just showed on
[01:20] (80.48s)
perplexely is that the interface or do I
[01:23] (83.36s)
just have an agent that just goes and
[01:24] (84.80s)
does everything for me and is this the
[01:27] (87.52s)
start of that? I you know I just haven't
[01:29] (89.68s)
spent enough time. I I do know that
[01:32] (92.40s)
every consumer
[01:34] (94.48s)
software CEO
[01:36] (96.96s)
that has an app in the app store is
[01:38] (98.80s)
tripping. They're tripping right now.
[01:41] (101.20s)
And I mean big boys. I meet guys with
[01:43] (103.28s)
real stuff and sometimes I I'm doing
[01:45] (105.92s)
like almost like therapy sessions with
[01:47] (107.92s)
them. I'm like, "It's gonna be fine. You
[01:50] (110.00s)
actually you actually have stuff. You
[01:52] (112.64s)
have a mo you have real stuff that's of
[01:54] (114.64s)
value. They can't replace it with an
[01:56] (116.32s)
agent." And they're like,
[01:57] (117.44s)
"So, you're lying to them. You're doing
[01:58] (118.96s)
hospice care and you're telling them
[02:00] (120.24s)
everything's going to be okay, but the
[02:02] (122.00s)
patient
[02:03] (123.52s)
options on Robin Hood while he's like,
[02:04] (124.96s)
"Yeah, yeah, tell me more. Tell me more.
[02:07] (127.84s)
All these things.
[02:08] (128.56s)
There's certain things that are
[02:09] (129.84s)
protected. There's certain not things
[02:11] (131.12s)
that aren't. That's all.
[02:12] (132.24s)
Well, let's talk about that because the
[02:14] (134.00s)
you and I are old enough to remember uh
[02:16] (136.40s)
general magic. This vision was out there
[02:18] (138.56s)
a long time ago with personal digital
[02:21] (141.04s)
assistance and you would just talk to an
[02:23] (143.28s)
agent. It would go do this for you. This
[02:25] (145.28s)
feels like a step to that where it does
[02:28] (148.24s)
all the work for you, presents you the
[02:30] (150.72s)
final moment and says approve.
[02:33] (153.76s)
Like a concierge or a butler. Yeah,
[02:35] (155.68s)
I think what you're describing is what
[02:37] (157.92s)
we want. But I think more specifically
[02:40] (160.08s)
for today, Keith and Travis totally nail
[02:44] (164.32s)
it. Look, I think building a browser is
[02:47] (167.04s)
an absolutely stupid capital allocation
[02:49] (169.28s)
decision. Just totally stupid and
[02:51] (171.36s)
unjustifiable in 2025. Specifically for
[02:54] (174.56s)
Perplexity,
[02:56] (176.24s)
I think their path to building a legacy
[02:59] (179.12s)
business is to replace Bloomberg.
[03:01] (181.76s)
Everything that they've done in
[03:03] (183.84s)
financial information and financial data
[03:06] (186.72s)
in going beyond the model has been
[03:09] (189.28s)
excellent. As somebody who's paid
[03:10] (190.96s)
$25,000 to Bloomberg for many years,
[03:14] (194.72s)
the terminal is atrocious. It's
[03:17] (197.92s)
terrible. It's not very good. It's very
[03:20] (200.72s)
limited.
[03:22] (202.32s)
and anybody that could build a better
[03:25] (205.04s)
product would take over a hundred
[03:29] (209.44s)
billion dollar enterprise because I
[03:31] (211.04s)
think it's there for the taking. I wish
[03:32] (212.40s)
that perplexity would double and triple
[03:34] (214.56s)
down on that. And so when you see this
[03:36] (216.08s)
kind of
[03:37] (217.04s)
random spraw,
[03:38] (218.08s)
let's do it, Jimoth. Let's just go do
[03:40] (220.48s)
When you do the random spraw, I think it
[03:41] (221.92s)
doesn't work. But I just want to say
[03:44] (224.32s)
a browser is like the dumbest thing to
[03:46] (226.24s)
build in 2025 because in a world of
[03:48] (228.80s)
agents, what is a browser? It's a
[03:51] (231.52s)
glorified markup reader. It's like
[03:53] (233.76s)
handling HTML. It's handling CSS and
[03:56] (236.16s)
JavaScript. It's doing some networking.
[03:58] (238.64s)
It's doing some security. It's doing
[04:00] (240.56s)
some rendering. But it's like this is
[04:02] (242.96s)
all under the water type stuff. I get it
[04:06] (246.32s)
that we had to deal with all that
[04:07] (247.84s)
nonsense in
[04:11] (251.76s)
to try LIOS or Google for the first
[04:13] (253.92s)
time. But in 2025,
[04:16] (256.96s)
there's something that you just speak to
[04:20] (260.32s)
and eventually there's probably
[04:21] (261.60s)
something that's in your brain which you
[04:23] (263.04s)
just think and it just doesn't you're
[04:25] (265.12s)
thinking
[04:26] (266.64s)
I need a flight to JFK or at the maximum
[04:29] (269.92s)
today in a very elegant beautiful search
[04:32] (272.32s)
bar you type in get me a flight and it
[04:35] (275.36s)
already knows what to do.
[04:36] (276.48s)
Keith in some ways this is a step
[04:38] (278.48s)
towards that ultimate vision. So you'd
[04:40] (280.88s)
think it's worth it to you know sort of
[04:43] (283.52s)
perplexity to make this way point
[04:45] (285.44s)
perhaps if you look at it as a waypoint
[04:47] (287.28s)
between the ultimate vision which is a
[04:48] (288.96s)
command line and earpiece.
[04:51] (291.36s)
How do you get distribution Jason for
[04:52] (292.96s)
the 19th web browser in 2025? Well,
[04:55] (295.52s)
yeah, that is a challenge and I think
[04:57] (297.12s)
most people are speculating Apple, which
[05:00] (300.40s)
has a lot of users, might buy Perplexity
[05:04] (304.40s)
or do a deal with Perplexity and give
[05:06] (306.16s)
them that distribution because of the
[05:08] (308.80s)
Justice Department case against Google.
[05:11] (311.04s)
So, there's been a lot of speculation
[05:12] (312.24s)
about that. But Keith, what do you
[05:13] (313.92s)
Well, I don't think they'd buy anything
[05:15] (315.36s)
worth it. Like, what do what is Apple
[05:16] (316.72s)
going to get? We continue this failed
[05:18] (318.48s)
strategy of Apple,
[05:19] (319.76s)
right? Apple's missed every possible
[05:22] (322.00s)
window on AI and continues to miss it
[05:24] (324.24s)
and it has cultural I think the CEO has
[05:27] (327.36s)
challenges I think culturally they have
[05:28] (328.88s)
challenges I think they have
[05:29] (329.60s)
infrastructure challenges so it's it's
[05:32] (332.08s)
not an easy fix but buying perplexity is
[05:34] (334.40s)
not going to help like chas strategy is
[05:35] (335.92s)
actually pretty coherent one for
[05:37] (337.36s)
perplexity quap perplexity uh so I think
[05:41] (341.12s)
pick a vertical and own it strategy in
[05:43] (343.36s)
this case
[05:43] (343.68s)
not not a bad idea um especially because
[05:45] (345.84s)
you need unique data sources some of
[05:47] (347.44s)
those data sources may or may not
[05:49] (349.52s)
license their data to open AAI. So you
[05:52] (352.24s)
can do some clever things there, but um
[05:54] (354.48s)
I don't think there's any residual value
[05:56] (356.40s)
that Apple will get out of perplexity
[05:58] (358.08s)
except there's some product taste, but
[05:59] (359.84s)
what are you going to spend like a
[06:00] (360.96s)
billion dollars for product taste? I
[06:02] (362.40s)
mean Mark's spending hundreds of
[06:04] (364.00s)
millions of dollar hundreds of billions
[06:05] (365.44s)
or whatever he's spending these days and
[06:07] (367.44s)
you know Grock if anything Grock for
[06:09] (369.12s)
shows that Mark really doesn't need to
[06:11] (371.44s)
just spend money to build a whole new
[06:12] (372.88s)
team because everything they've done in
[06:14] (374.00s)
AI is also missed the boat. Well, I
[06:16] (376.48s)
mean, Keith, the way you phrase it there
[06:17] (377.92s)
almost makes it worth it for Apple to
[06:20] (380.32s)
throw a Hail Mary, have a team with some
[06:22] (382.48s)
taste because that's how they tend to do
[06:24] (384.48s)
things is something that is elegant. And
[06:26] (386.88s)
why not just throw your search to it,
[06:28] (388.72s)
throw 10 billion at what's
[06:32] (392.40s)
elegant would be if there's a bunch of
[06:33] (393.84s)
agents and just a chat box.
[06:35] (395.76s)
Seeing a bunch of visual diarrhea is not
[06:37] (397.84s)
elegant.
[06:38] (398.88s)
It's lazy.
[06:40] (400.64s)
Off on our on our little Bloomberg
[06:42] (402.48s)
clone, I'll give you naming rights. So
[06:44] (404.80s)
you can call it that.
[06:45] (405.44s)
Oh, you like you like it.
[06:46] (406.88s)
Polyhapatia. So, hey, can somebody can
[06:49] (409.76s)
somebody uh bring up the polyhapatia?
[06:52] (412.96s)
You know what's so funny?
[06:54] (414.16s)
It just rolls right off your tongue.
[06:56] (416.24s)
TK, listen. We were trying to do a
[07:00] (420.08s)
of companies and it maxes out at five
[07:03] (423.68s)
companies on a specific type of screen
[07:05] (425.92s)
where you're like you're trying to
[07:07] (427.04s)
compare stock price to EB and you're
[07:08] (428.96s)
like, okay, I can only choose five, I
[07:11] (431.12s)
guess. So, which five should I choose?
[07:12] (432.48s)
Lefont was on right like two episodes
[07:14] (434.16s)
ago. He was like I can't pull this up.
[07:16] (436.00s)
It's limited to six companies.
[07:17] (437.84s)
Dude, you it's So what do people use
[07:20] (440.00s)
Bloomberg?
[07:21] (441.20s)
They use it for the messaging. Now like
[07:23] (443.12s)
my team has traded huge positions via
[07:25] (445.44s)
text message on Bloomberg. So there is
[07:27] (447.20s)
something very valuable there.
[07:29] (449.20s)
But the core usability and the core UI
[07:32] (452.00s)
of that company has not evolved.
[07:33] (453.92s)
I have my contribution
[07:35] (455.36s)
and complexity is very good at that by
[07:37] (457.68s)
the way. It they they do a very good
[07:40] (460.08s)
I got a new domain named Travis. Let
[07:41] (461.76s)
this one just sink in here. This is my
[07:43] (463.44s)
way to weasel my way into the deal.
[07:45] (465.60s)
Begin.com.
[07:46] (466.96s)
begin.com.
[07:47] (467.92s)
You own that, don't you?
[07:50] (470.00s)
I'm just a little I snipe some good ones
[07:52] (472.24s)
once in a while. I got begin.com and I
[07:53] (473.92s)
got annotated.com. Those are my two
[07:55] (475.68s)
little domain. You're like you're like
[07:57] (477.12s)
one of these old people that show up at
[07:58] (478.56s)
those road show and then
[08:01] (481.76s)
road show and you're like, "Oh, I have
[08:03] (483.68s)
this thing that I bought 1845."
[08:06] (486.24s)
Guys, Jason Jason is Jason is the daddy
[08:09] (489.28s)
and god daddy. Okay, that's just what it
[08:11] (491.20s)
is. Your dad your dad.
[08:12] (492.32s)
That's what it is.