Latest / In the Loop AI / This Week in AI: OpenAI Changes Everything, Neo Robot, Nvidia To $5T, Grokipedia

Transcript

This Week in AI: OpenAI Changes Everything, Neo Robot, Nvidia To $5T, Grokipedia · In the Loop AI

0:03
Ejaaz: Welcome back to the weekly AI roundup. It's been another huge week.
0:08
Ejaaz: NVIDIA became the first company to cross $5 trillion in market value,
0:12
Ejaaz: but they're closely followed behind by Apple and Microsoft to cross $4 trillion
0:17
Ejaaz: in market cap value. Some big numbers being thrown around.
0:20
Ejaaz: OpenAI had a pretty controversial week. They restructured their entire company,
0:23
Ejaaz: but they also revealed their secret plans for AGI over the next 10 years.
0:27
Ejaaz: Elon Musk decided to release an AI version of Wikipedia called Grokipedia,
0:33
Ejaaz: powered by their flagship model Grok. We'll get into that.
0:36
Ejaaz: And finally, there is a new humanoid robot available to you for the cheap price of $499 per month.
0:44
Ejaaz: Josh, there's a lot to get through to today.
0:47
Ejaaz: Starting off with all the open AI news, There's two main things to discuss here.
0:53
Ejaaz: They made two big announcements.
0:54
Ejaaz: One was their vision for the next 10 years on achieving AGI.
0:57
Ejaaz: And the second thing was on the restructuring. Starting with the AGI side of
1:01
Ejaaz: things, Sam Altman came on live stream with his chief scientist,
1:06
Ejaaz: Jakub, to describe what the vision of OpenAI is.
1:10
Ejaaz: And I think this is in response to a lot of people saying, Sam,
1:14
Ejaaz: like, do you have your best intentions at heart?
1:16
Ejaaz: You keep saying that, but then your actions are kind of producing products that
1:19
Ejaaz: might seem kind of consumery and not really big thinking, are you still focused on AGI?
1:25
Ejaaz: So this is kind of like clearing the plate here. And there's three main takeaways
1:28
Ejaaz: that I got from this that I found interesting.
1:31
Ejaaz: Number one, Sam has extended his deadline to achieving AGI.
1:35
Ejaaz: He says he's going to achieve it within 10 years, and it's going to be a slower
1:40
Ejaaz: transition than initially expected.
1:42
Ejaaz: He describes it as achieving AGI gradually over a number of different years.
1:48
Ejaaz: This is interesting to me because we're finally achieving a realistic timeline,
1:53
Ejaaz: in my opinion, versus some kind of like hype headline where we're just kind
1:57
Ejaaz: of like thinking it's going to arrive tomorrow.
1:59
Ejaaz: Number two, he's started giving solid milestones on this.
2:03
Ejaaz: He says in 2026, the deep learning that OpenAI is conducting internally will
2:08
Ejaaz: result in an AGI level science model, which will perform kind of at the caliber
2:14
Ejaaz: of an AI research intern.
2:16
Ejaaz: And then he says two years later in 2028, they'll have an AI researcher level model.
2:21
Ejaaz: So you go from intern to research grade performance, which is super cool.
2:25
Ejaaz: And the final point that really interested me was around compute.
2:28
Ejaaz: Josh, they've spent or committed around $1.4 trillion on data centers right
2:34
Ejaaz: now, which is roughly the equivalent of 30 gigawatts. It's a huge amount of compute.
2:39
Ejaaz: But what shocked me the most was he came out and said that his eventual plan
2:45
Ejaaz: is to have a factory built every week that produces a gigawatt of power.
2:50
Ejaaz: That's about $20 billion of compute per week.
2:53
Ejaaz: All really huge numbers kind of took my breath away. Josh, what's your take?
2:59
Josh: OpenAI is just doing what everyone kind of knows they were always going to do.
3:03
Josh: They very quickly realized they can't be a non-for-profit. They are now becoming
3:06
Josh: a for-profit organization.
3:08
Josh: In fact, it's a public benefit corporation. So it's a PBC, not an LLC or a Z-Corp
3:12
Josh: or anything, which means the intention is to be kind of aligned with public
3:16
Josh: interest, but basically it just blows the cap off of everything.
3:19
Josh: And this is a really big deal.
3:20
Josh: EJ, you mentioned that, or you mentioned to me actually earlier,
3:23
Josh: and this was news to me, that Sam Altman expressed interest in going public.
3:27
Josh: I think there's a big opportunity here to talk about how much money can be made by.
3:32
Josh: OpenAI going public and by kind of restructuring this. So the restructuring deal is interesting.
3:38
Josh: Are we ready to talk about that? Can we go into this like crazy chaotic restructuring
3:42
Josh: that's happening? Okay.
3:43
Josh: This is fascinating to me. So on screen, there is a disastrous diagram with a bunch of arrows.
3:48
Josh: A lot of it is kind of hand-waving, not important. The things that matter,
3:51
Josh: Microsoft owns a large percentage of the nonprofit.
3:54
Josh: The nonprofit quickly realized they needed to become a for-profit.
3:57
Josh: They had a capped upside on the previous version, but now they do not.
4:00
Josh: So Microsoft was limited previously to a 100x return on their investment.
4:04
Josh: Now there is no return on the investment or no limit to the return on the investment.
4:09
Josh: And this is interesting for a few things, Ejaz, because one,
4:12
Josh: I am very resentful of OpenAI because they are very clearly the fastest growing AI company in the world.
4:18
Josh: And I have no way to get exposure to them. That doesn't feel like it's aligned with the public.
4:22
Josh: Sam Altman is taking all the money for himself. He's throwing it at GPUs and I get no upside.
4:27
Josh: That doesn't feel right. this is open AI, you're a public benefit company.
4:30
Josh: Why are you not benefiting me as a member of the public?
4:33
Josh: Second to this is that Microsoft now seems to be the best way to get exposure
4:38
Josh: because there is this like,
4:40
Josh: there is this way of getting access to open AI when they IPO,
4:43
Josh: but open AI and Microsoft are now, or Microsoft's not the biggest stakeholder
4:47
Josh: in the open AI company at 27% versus the open AI non-for-profit, which is actually 26%.
4:53
Josh: So the largest form of exposure is through Microsoft. They have uncapped profits
4:58
Josh: and they are reporting earnings, I think, this week.
5:01
Josh: So Microsoft might be the best way to get exposure to OpenAI currently, but...
5:07
Josh: OpenAI as a whole has kind of disappointed me through this whole process because
5:10
Josh: it's all been very hand wavy and not really aligned with, I think,
5:15
Josh: the public intention that they stated, just because they've always been this
5:18
Josh: big for profit entity disguised as a nonprofit.
5:21
Ejaaz: Yeah, I completely agree with a lot of the points you made.
5:25
Ejaaz: And to just kind of give the listeners a bit of context here,
5:30
Ejaaz: OpenAI founded in 2015 as a nonprofit.
5:33
Ejaaz: It was set up as a nonprofit org with this ambitious vision of we are going
5:37
Ejaaz: to achieve AGI and we are going to provide it to the people.
5:40
Ejaaz: So everyone should have access to this super powerful technology.
5:43
Ejaaz: Then four years later, in 2019, they decided, we kind of want to make a bit
5:48
Ejaaz: of money from this and we need to raise a lot of capital.
5:51
Ejaaz: So in 2019, they set up a for-profit structure, but it was owned by the non-profit.
5:56
Ejaaz: So they could still get away with saying, hey, we're a non-profit.
5:59
Ejaaz: But it was capped, as you said, Josh, to any investor for 100x, just 100x returns.
6:05
Ejaaz: So a lot of feedback they've been getting and a lot of pressure they've been
6:07
Ejaaz: getting from future investors or current investors recently is,
6:11
Ejaaz: hey, we want to make more money on this deal.
6:13
Ejaaz: It's unfair that we're investing billions in you.
6:16
Ejaaz: Microsoft invested $13 billion and we want more upside than 100x.
6:21
Ejaaz: So Sam went to the drawing board and was like, okay, we can create a new structure
6:26
Ejaaz: called a public benefit corporation, which, as you said, now has uncapped gains. It's pretty crazy.
6:31
Ejaaz: So now you can just make as much money as you can on a normal private company
6:36
Ejaaz: that you were investing in.
6:37
Ejaaz: And the biggest point around this is on this diagram here, you'll see if you
6:41
Ejaaz: zoom in, if you squint, Microsoft, minority owner, they are now no longer a minority owner.
6:47
Ejaaz: Like you said, they are the majority owner, which is just crazy.
6:50
Ejaaz: But I'll tell you a few other reasons why Microsoft is a huge winner from all
6:54
Ejaaz: of this. It's not just money.
6:55
Ejaaz: It's not just the fact that their $13 billion stake is now worth $135 billion. What a crazy return.
7:03
Ejaaz: But they still maintain exclusive access and rights to all of OpenAI's research
7:09
Ejaaz: IP, which means that any kind of new model that they release,
7:14
Ejaaz: Microsoft basically has the first say and exclusive access to provide it to
7:18
Ejaaz: their partners, to their customers and any clients and enterprise clients that
7:21
Ejaaz: they wanna bring in in the future.
7:23
Ejaaz: Secondly, and this was a genius move, Josh, they set a standard.
7:28
Ejaaz: They said, listen, OpenAI, We will relieve all our exclusive access and IP access
7:34
Ejaaz: only once you've achieved AGI.
7:36
Ejaaz: But of course, the obvious question is, what the hell is AGI?
7:40
Ejaaz: How do you determine what AGI is? There's so much discussion around what the hell this thing is.
7:44
Ejaaz: So they said, we'll leave it to an expert party of third-party analysts.
7:49
Ejaaz: So it's a panel which is going to decide and determine OpenAI's fate going forward.
7:54
Ejaaz: So the winner of all of this, this restructuring, this OpenAI vision announcement,
7:58
Ejaaz: is actually Microsoft, and it's just crazy to see it.
8:02
Josh: Okay, so how do you navigate this as a participant who wants to get wealthy off of AI?
8:06
Josh: There is a few ways. I think for people who observe public markets,
8:10
Josh: you'll know the fact that Apple is sitting at all-time high,
8:12
Josh: Microsoft sitting at all-time high, Google, basically across the board,
8:15
Josh: everyone is sitting at all-time highs, which I believe is high signal.
8:18
Josh: Clearly, things are working. Clearly, a lot is being spent. Clearly,
8:21
Josh: there is some sort of a bubble, but the end is not quite near.
8:25
Josh: The big question that I have is like, come on, how do I get access to OpenAI
8:28
Josh: if you don't, if you're not an accredited investor, if you can't figure out
8:31
Josh: how to get private shares, the answer right now is just Microsoft.
8:34
Josh: Nobody in the world owns more shares, not even OpenAI themselves, than Microsoft does.
8:39
Josh: And Microsoft, like you said, EJS, they have access to their IP,
8:42
Josh: they get access to a revenue share, they get access to actual equity within the company.
8:46
Josh: And if you believe in the success of OpenAI at a scale that we do,
8:49
Josh: which is many, many trillions of dollars of market cap, 27% of that is not to be laughed at.
8:53
Josh: And I mean, Microsoft is a $4 trillion dollar company in the case that
8:56
Josh: open ai makes it to that level that is another nearly
9:00
Josh: trillion dollars in microsoft's pocket which
9:03
Josh: is a really big deal so it is unfortunate that
9:06
Josh: it gets diluted out through the form of having to buy the same
9:09
Josh: company that makes microsoft office which is just like an abomination my
9:12
Josh: personal opinion at least but um i think that is the way to kind of navigate
9:16
Josh: this forward until we do get the eventual ipo from the open ai team like sam
9:20
Josh: allman discussed so that's kind of how i'm thinking about navigating it is like
9:23
Josh: i want access sam's being a little a little closed closed with that um but microsoft
9:28
Josh: is probably your your path to success there that's.
9:30
Ejaaz: So interesting i i read it kind of the uh in
9:34
Ejaaz: a different way josh which is the gloves are now finally off
9:37
Ejaaz: open ai doesn't have the constraint of having to lop as a non-profit organization
9:41
Ejaaz: it's quite clearly a for-profit um the foundation owns equity stake in this
9:45
Ejaaz: so i think sam will probably move quite quickly on an ipo because he needs so
9:50
Ejaaz: much capital if he wants to get to a gigawatt factory per week, like he says.
9:55
Josh: That's probably going to be the biggest IPO ever.
9:57
Josh: Right. Like, I think right now, OpenAI and SpaceX are kind of neck and neck
10:00
Josh: for most valuable private company in the world.
10:03
Josh: And when that company goes public,
10:05
Josh: my God, the amount of volume that is going to be traded on day one.
10:08
Ejaaz: I'll be buying.
10:10
Josh: Shout out to the early employees who got stock in OpenAI. My gosh,
10:14
Josh: you are about to buy very, very, very, very luxe houses.
10:17
Josh: But I guess that's it for OpenAI. This is a really fun topic that I've been
10:20
Josh: dying to talk about, which is human order robotics.
10:22
Josh: I love human order robotics. And the first one has hit the market,
10:26
Josh: or at least so they say. say.
10:28
Josh: We're not entirely sure where this stands. Do you want to introduce everyone
10:31
Josh: to the OneX Neo humanoid home robot? What the hell is it? What does it do? How does it work?
10:38
Ejaaz: Yes. So if you remember on a previous episode, actually last week,
10:41
Ejaaz: we spoke about the new figure humanoid robot where it can enter your home,
10:45
Ejaaz: it could do a bunch of things, but it can also do a bunch of like other things
10:48
Ejaaz: like manufacturing jobs, service jobs and all that kind of stuff.
10:50
Ejaaz: This is the latest humanoid robot to come out of a company called OneX.
10:56
Ejaaz: And my first impressions of it Josh can I can I can I be honest with you can I be honest I'd
11:00
Josh: Love for you to be nothing but honest.
11:01
Ejaaz: It looks like a grown-up Teletubby
11:06
Ejaaz: I don't mean that in a demeaning way. Let me skip ahead to show you like a clip of this thing.
11:11
Ejaaz: It's kind of like in this soft, it's wearing a soft suit. It's got a very pleasant
11:16
Ejaaz: looking face. It's got these kind of like glow up ears.
11:19
Ejaaz: It looks like an adult alien Teletubby, but it's comforting to be around.
11:24
Ejaaz: And in terms of like what it can do, it's this humanoid robot that can live
11:28
Ejaaz: in your home and it can do all the things you would kind of expect a houseware robot to do.
11:32
Ejaaz: It can clean your dishes, it can put away the groceries, it can lift things.
11:36
Ejaaz: I think it has like a lifting capacity of 55 pounds, which is pretty crazy for
11:40
Ejaaz: something that weighs 60 pounds.
11:42
Ejaaz: I don't know how those mechanics work. I'm not a physicist, but that is pretty insane to see.
11:46
Ejaaz: And this video is like a really high production thing, Josh.
11:50
Ejaaz: I don't know how much they invested in this, but if the robot that we eventually
11:53
Ejaaz: get in our homes in 2026, for the people that are ordering this for 500 bucks
11:57
Ejaaz: or buying it outright for $20,000, which is a hefty price tag,
12:02
Ejaaz: if it gets anywhere near something like this, I think it's going to completely change the game.
12:06
Ejaaz: I would personally get something like this in my home. Josh,
12:09
Ejaaz: have you put down the payment on this? Absolutely. What are you doing here?
12:13
Josh: No, not at all. Again, here's the tale of two wolves where I am a techno optimist
12:17
Josh: who is rooted in reality.
12:20
Josh: And I have been around long enough. I have seen enough of these videos that
12:23
Josh: look very beautiful and lovely and very welcoming to have a humanoid robot.
12:28
Josh: But EJ, how many people have humanoid robots in their house? A total of zero.
12:31
Josh: It just doesn't exist. No one shipped them to market. A few things that are
12:35
Josh: important to know about this robot. It is, like you mentioned, $20,000 or $500 a month.
12:39
Josh: And it does the following. It is able to do basic tests like laundry,
12:43
Josh: possibly the dishes, possibly vacuuming.
12:47
Josh: We're not entirely sure. And the problem with that is that if it can't do the
12:51
Josh: task, you have to use this thing. I believe they call it expert mode.
12:54
Josh: Now, expert mode is the giant red flag waving in the sky for me because expert
12:59
Josh: mode is the highest signal that this is not ready to enter the market.
13:05
Josh: Expert mode means if it's not able to do the thing you want,
13:08
Josh: you can schedule a time with the Neo team and someone from the headquarters
13:12
Josh: or wherever they're based will get into this little VR suit and manually walk
13:17
Josh: the robot through your home to do the task for you.
13:20
Josh: So instead of having a maid come to your house, you get some tech engineer,
13:24
Josh: put on some goggles, and he is walking through your personal space doing the
13:28
Josh: task that you want to do, that you've scheduled ahead of time in order to make this happen.
13:32
Josh: And I believe that is what part of that monthly $500 a month thing is,
13:36
Josh: is you're actually, you're paying hours of labor for humans to do the emulated
13:42
Josh: version inside of this humanoid robot.
13:45
Josh: So that to me is a really big red flag. The second red flag that I want to serve
13:48
Josh: before I'll pass back to you, AJ, is the fact that this is a pre-order.
13:52
Josh: You're not buying a robot. You are buying the rights to a future robot.
13:55
Josh: And there is no guarantees that this robot will ever exist at scale,
13:59
Josh: nor will it exist to the extent that it's shown in this video.
14:02
Josh: And it's one of those things, and we've seen this with It's like,
14:05
Josh: show me, don't tell me. And I'm seeing a lot of these really lovely demos.
14:10
Josh: But until this thing is actually produced and until this thing gets shipped
14:13
Josh: and until it arrives at people's houses, I am going to be sitting here a little
14:17
Josh: skeptical because a lot of the companies like...
14:21
Josh: I'll use Tesla as an example. They have a proven track record of building really
14:24
Josh: badass manufacturing, hardcore engineering stuff.
14:27
Josh: These startups don't really have any track record of manufacturing hard things.
14:30
Josh: And a humanoid robot is an incredibly difficult thing to manufacture.
14:34
Josh: So when they say pre-orders and delivery in 2026, well, I assume the best case
14:38
Josh: is like December 31st, 2026.
14:40
Josh: I would be shocked if they deliver these at scale before that.
14:42
Josh: But I'm wondering if you have if you have a different take on that.
14:45
Ejaaz: Sadly, no, I am not going to pay the equivalent
14:48
Ejaaz: of half a month's rent uh well
14:51
Ejaaz: maybe not in new york but some in some other state um to
14:54
Ejaaz: get some third party that is wearing a
14:57
Ejaaz: metamorph suit and acting like a shitty uh version
15:00
Ejaaz: of avatar um the movie to navigate my robot and clean my clothes i would rather
15:06
Ejaaz: spend that money to get some actual house help to to come in and help me out
15:10
Ejaaz: with all the things that we're seeing on this video so far okay i have one simple
15:14
Ejaaz: rule if you are going to advertise a next generation humanoid robot that is autonomous,
15:20
Ejaaz: it should be autonomous.
15:23
Ejaaz: And don't sell me something that it needs to be teleoperated.
15:26
Ejaaz: It makes sense in the case of something like self-driving cars.
15:30
Ejaaz: Waymo, I know, has a tele operation thing. That makes sense because there's
15:33
Ejaaz: a lot of safety and stuff involved.
15:34
Ejaaz: But for something that is consumer-friendly, for something that is meant to
15:37
Ejaaz: make my life a hell of a lot easier in my most private space ever at home,
15:42
Ejaaz: it should be fully autonomous.
15:43
Ejaaz: And let alone the home is like the most private place, right, Josh?
15:47
Ejaaz: Like I have all my valuables here. I don't want some dude in a third world country,
15:51
Ejaaz: sorry, navigating around my house and picking stuff up and putting it down. What if they break it?
15:55
Ejaaz: What if they do something wrong? Who's liable for that? Do I get insurance with this thing?
15:59
Ejaaz: I have no idea. And then tacking on a $20,000 price tag to buy this thing outright.
16:04
Ejaaz: If you are buying this outright for $20,000, I'm going to go ahead and say that
16:08
Ejaaz: you are kind of like an idiot at that extent.
16:12
Ejaaz: Like, I do not advise doing this.
16:14
Ejaaz: And then the second point you made, Josh, I think is the more prescient point, which is,
16:19
Ejaaz: this thing hasn't been built yet. This hasn't been produced at scale.
16:23
Ejaaz: This thing isn't going to come into your home. It's not even guaranteed to come
16:26
Ejaaz: into your home until like sometime in 2026.
16:29
Ejaaz: By that time, you're going to have other robot companies like Unitree,
16:32
Ejaaz: like Figure, come out with their actual robot, which is being built and scaled.
16:35
Ejaaz: I have no idea where 1X is as a manufacturing company.
16:39
Ejaaz: You've mentioned several times actually on our show, Josh, that one of the bullish
16:44
Ejaaz: cases on Tesla's Optimus robot is the fact that Elon is such a manufacturing nerd.
16:50
Ejaaz: He's so good at scaling. And I just don't see this or want to trust this with a robot startup.
16:55
Josh: Yeah, I don't want to call people dumb for spending $20,000 because if I had
16:58
Josh: a disposable $20,000, I would certainly be an early adopter.
17:01
Josh: Maybe we can refer to that.
17:03
Josh: One of the weird edge case questions that I have is like, okay,
17:06
Josh: you pay $500 a month, then what? Like, what if I don't want it anymore?
17:09
Josh: Is there like a repo man that comes? Does the robot walk out of my house?
17:12
Josh: How does that work? The teleoperation thing, it's a serious problem.
17:16
Josh: Um, granted you can schedule it, but I imagine myself like, Oh,
17:19
Josh: you walk out of the shower and here is like some, some dude from like wherever he's from.
17:22
Josh: He's just like, like, hi, I'm just cleaning your windows.
17:25
Josh: It's this really weird, like intimate distraction.
17:28
Josh: Um, and I think the open-endedness of that expert mode relative to the set amounts
17:34
Josh: of chores that the robot can do leaves a lot to be desired or just a lot of
17:39
Josh: unknowns where if I do spend all this money, I don't actually know what it's capable of doing.
17:43
Josh: They say it could do a few things, but there's a lot of edge cases.
17:45
Josh: Everyone's houses are different.
17:46
Josh: Just a little uncertain on this robot. But again, happy people pushing this
17:49
Josh: forward. I think this is great progress. I love the vision.
17:53
Josh: If the vision in the video becomes a reality, that is a life that I really am
17:57
Josh: excited about. That is a reality that is awesome.
17:59
Josh: So like shout out to the One X team for trying. They're pushing the ball forward.
18:02
Josh: I am not going to knock it too hard because I just admire their effort in making
18:07
Josh: this a reality. And like everything else, we'll see.
18:09
Josh: Ship the first robot. Let's see how it performs in homes.
18:14
Josh: Let's we'll see how it goes but this is,
18:16
Josh: I guess that's it for the robot stuff. Now we get into the cyberspace,
18:20
Josh: the world of Grok. Oh, hell yeah.
18:22
Josh: And Grokopedia has launched this week, EJ. So why is this a big deal?
18:25
Ejaaz: This is a huge deal for me because I grew up on Wikipedia and I used it to help
18:32
Ejaaz: me research and write a bunch of essays at high school and at college.
18:36
Ejaaz: I still used it up until a few days ago where I would like help me clarify something
18:41
Ejaaz: about like learning about some archaic thing, which I wouldn't be able to find
18:44
Ejaaz: anywhere. It is the internet's resource knowledge.
18:47
Ejaaz: Millions of people use it every single day.
18:51
Ejaaz: But there's one issue, which is I don't know whether this source of truth is actually the truth.
18:57
Ejaaz: And it's been something that we've been increasingly faced in society as things
19:02
Ejaaz: like social media become enveloped around the world, like everyone's making
19:05
Ejaaz: TikToks, everyone's learning their news from TikTok, from tweets.
19:09
Ejaaz: And so it becomes really murky in terms of like figuring out what that truth is.
19:13
Ejaaz: So Elon came out a few months ago, and it's
19:16
Ejaaz: crazy to say that this happened a few months ago and he said I think
19:20
Ejaaz: AI will be the best source of truth and so
19:22
Ejaaz: what I'm going to do as part of building out Grok which is xai's
19:26
Ejaaz: flagship AI model is create an AI
19:29
Ejaaz: version of Wikipedia so he built that and he probably aptly called it Grokipedia
19:35
Ejaaz: so what we're seeing in front of us right now is a demo of Grokipedia looks
19:40
Ejaaz: like a simple chat shewiti prompt bar except that you can search for everything
19:44
Ejaaz: and anything that you typically would in Wikipedia.
19:46
Ejaaz: And when you do so, you get this really neat little concise summary and intro
19:51
Ejaaz: on whatever topic, one we're seeing on the screen is the Tesla Cybertruck.
19:55
Ejaaz: And then it goes on to kind of dig into its early life, similar that you'd see
19:59
Ejaaz: on Wikipedia, how it was originated, and its site sources, Josh,
20:05
Ejaaz: throughout the entire thing.
20:06
Ejaaz: That to me is the most important part because typically in Wikipedia,
20:11
Ejaaz: they do this thing which is kind of hidden behind closed doors.
20:14
Ejaaz: Where they restrict the number of sources you can use as legitimate sources for your articles.
20:21
Ejaaz: And if you look at that list, and if you compare it to the black list,
20:25
Ejaaz: which is the very, very long list of sources which you can't quote,
20:29
Ejaaz: it starts limiting what can be factual and what can be opinionated.
20:34
Ejaaz: And it starts skewing a lot of people's opinions as to what is the truth or what is not.
20:38
Ejaaz: Grok, Grok, Grokpedia and Grok on the other hand is completely unbiased.
20:42
Ejaaz: It pulls sources from anywhere and everywhere and it does its own analysis to
20:46
Ejaaz: vindicate whether the source is legit or whether it's not.
20:49
Ejaaz: It does this in real time and it processes hundreds of millions of articles,
20:53
Ejaaz: sources, primary sources, secondary resources every single day.
20:57
Ejaaz: So I'm super excited about this because I'm, it's going to be my new research
21:01
Ejaaz: tool going forwards in terms of fact checking.
21:03
Josh: The purpose of Grok is to seek truth. That's kind of their, their end goal.
21:06
Josh: That's their stated goal. The same way that a lot of these other companies like
21:09
Josh: OpenAI was to have open source AI.
21:11
Josh: Grok's is just to seek truth. And I think with Grok5, they're really making
21:14
Josh: an attempt to lean into that because a lot of these traditional models are trained
21:18
Josh: on the data of the internet, which is highly opinionated and highly biased.
21:21
Josh: And what the Grok team, the XAI team in particular is doing is they're figuring
21:25
Josh: out ways to filter through the noise and find as close to source truth as they can.
21:30
Josh: And we've kind of seen this work with on the X platform with community notes,
21:34
Josh: how generally speaking, they are the closest thing we have to source truth because
21:38
Josh: it takes all these opposing opinions and kind of aggregates them and comes to a conclusion.
21:42
Josh: Grokipedia is showing us community notes at scale. And it's kind of,
21:46
Josh: it's the training set that is going to be used on the next Grok model,
21:50
Josh: which is Grok5, but they're kind of offering it to the public as a public good.
21:53
Josh: And not only is it a public good, it's an interactive public good.
21:56
Josh: So if you go to Grokipedia, you can actually suggest a new page to be generated
22:00
Josh: because they're all AI generated.
22:02
Josh: And Grok will take the liberty on itself to actually filter your request and
22:05
Josh: see if it's worthy or if it's not.
22:07
Josh: So if Grokipedia deems your request worthy enough, it will actually go and do
22:11
Josh: all the work in order to generate a new Wikipedia page or a new Grokipedia page.
22:15
Josh: And I think that kind of collaborative nature is fun, not only for Grok to get
22:19
Josh: better training data, but for the public to kind of get their own way.
22:23
Josh: And what's funny, Ejaz, is I've seen a lot of larger influencers on X posting
22:28
Josh: their Grokipedia page with a thank you letter, because for so long their public
22:31
Josh: perception has been skewed in a way that wasn't necessarily true.
22:35
Josh: And Grokipedia allows that to change, where now there is a source of truth about the person.
22:40
Josh: And when you see slander about someone online, you can check,
22:43
Josh: you can reference check this and it at least is trying its best to be truthful.
22:47
Josh: So that to me is why Grokopedia is, is this exciting new, important revolution
22:51
Josh: in technology that we have as a public good. Thank you to the X team. Yeah.
22:56
Ejaaz: Well, Josh, one of those people that posted a thank you about Grokipedia was
23:00
Ejaaz: the founder of Wikipedia himself, Larry Sanger, one of the founders of Wikipedia.
23:06
Ejaaz: And he goes in this tweet, OK, I finished reading my first long article about
23:11
Ejaaz: Grokipedia, the one titled Larry Sanger. That's him. That's his name.
23:15
Ejaaz: And he goes on to describe that for a version one for an AI based Wikipedia,
23:20
Ejaaz: it's actually really good. and it got a lot of things correct.
23:24
Ejaaz: Most importantly, the major things, which are the most controversial things about his biography.
23:29
Ejaaz: And he goes on citing a few examples.
23:32
Ejaaz: He does have a few criticisms, the one being that it gets a few minor mistakes
23:37
Ejaaz: that kind of roll in to become a big problem throughout the article.
23:41
Ejaaz: But again, he says, hats off to you, Elon, at the end.
23:45
Ejaaz: For everyone, this is amazing. I'm looking forward to seeing 1.0 and 2.0 going forwards.
23:50
Ejaaz: I think what a lot of people ended up doing when this product released,
23:55
Ejaaz: Josh, was to test a few controversial topics.
23:58
Ejaaz: Now, I don't want us to be politically affiliated at all on this podcast,
24:03
Ejaaz: but just to give a few examples, there was a side-by-side comparison of RFK
24:08
Ejaaz: Jr., which is a big political figure in the US.
24:11
Ejaaz: And you notice that one is very explicitly unbiased and less opinionated versus
24:16
Ejaaz: the other, Grok being the more unbiased version.
24:20
Ejaaz: You've got a side-by-side comparison of Donald Trump, and we had a kind of
24:29
Ejaaz: Unbiased adjudicator here in this, Josh, in this experiment where they gave Gemini 2.5.
24:34
Ejaaz: I don't know why Gemini 2.5 is the unbiased AI model of choice, but apparently it is.
24:39
Ejaaz: And it ended up deciding that Grok was in fact the more unbiased model.
24:43
Ejaaz: But there's also a bunch of really cool features that you can do in this as well, Josh.
24:48
Ejaaz: You mentioned that you can like submit to Grok an article that you want to get
24:51
Ejaaz: generated and have it instantly done.
24:53
Ejaaz: You can also fix mistakes in Grok as well. You can kind of highlight a request
24:58
Ejaaz: or highlight a thing that you think is wrong, submit it to Grok,
25:01
Ejaaz: and it analyzes it in the back end.
25:03
Ejaaz: The coolest part about this is you're not reliant on a bunch of
25:06
Ejaaz: specific moderators, human moderators that go to bed at night,
25:09
Ejaaz: that are awake at different times to get back to you.
25:12
Ejaaz: This is just done by an AI all in real time.
25:15
Ejaaz: If there's something that you don't understand, you can also just highlight
25:17
Ejaaz: it similar to how you would do on your iPhone and click like search and it brings
25:22
Ejaaz: you the Grokipedia page of that thing.
25:25
Ejaaz: So overall, I think this is a net improvement in what Wikipedia was.
25:29
Ejaaz: Do I think it's the best thing?
25:31
Ejaaz: No, but I think it's going to get a lot better over time. And I'm super excited
25:34
Ejaaz: about this. This thing is open source. Anyone can access it.
25:37
Josh: Super cool. Bullish on Grokipedia. And it's actually not the only open source
25:41
Josh: public good that we are getting in this week's episode.
25:44
Josh: There's a second one that goes by a very weird convoluted name of X402.
25:49
Josh: And I want to just briefly outline X402, give you like the 10 second little elevator pitch.
25:54
Josh: You can imagine X402, this kind of, it turns the web into a giant vending machine.
25:59
Josh: So you just like picture the internet as a vending machine.
26:02
Josh: You can walk up to it and you could try to like grab something off of the shelf,
26:04
Josh: whether it be a file or a video or a song.
26:07
Josh: And if it costs money, the website will say, 402, pay this tiny amount.
26:12
Josh: And you could pay it instantly. It's a very tiny amount. Like we're talking fractions of pennies.
26:16
Josh: You could pay instantly using crypto stable coins like USDC.
26:20
Josh: And then boom, you get the thing. And the amazing perk of this is you don't
26:24
Josh: need an account. You don't need a password.
26:26
Josh: You just pay for the thing you want and you walk away. Just like a vending machine.
26:29
Josh: You want to watch a new video.
26:30
Josh: Oh, here's like, give me 10 cents or 10%, one
26:33
Josh: tenth of a penny and you get access to this video and you can take it off
26:35
Josh: the wall and x402 it's just this open platform kind
26:39
Josh: of like http in fact they use part of the http protocol for
26:42
Josh: this um that allows the internet to implement this across
26:45
Josh: the board so anyone who wants to integrate it
26:48
Josh: can integrate it you can kind of imagine i think a lot of people are familiar
26:51
Josh: with linux how it's just open source if you want to integrate you can that's
26:54
Josh: what 402 is and it's this really awesome thing that allows ai agents in particular
26:59
Josh: to start engaging with the internet without needing an account without needing
27:03
Josh: a password, without needing to prove that they are humans.
27:06
Josh: So you have this big explainer up. Do you want to add anything to that explanation?
27:10
Ejaaz: No, I think you did a great job. What I will add is, I love the vending machine example.
27:16
Ejaaz: But yeah, to emphasize, this is a new web standard. So you don't have to go
27:22
Ejaaz: through a middleman like Stripe or PayPal.
27:25
Ejaaz: This is something that you can spin up and integrate into your website or app
27:29
Ejaaz: today, right now. And what's cool about this is it unlocks a bunch of really cool use cases.
27:35
Ejaaz: So in the previous world, before X402, you would need to set up a PayPal account,
27:41
Ejaaz: you would need to set up a checkout system, or enable subscriptions via your own API.
27:47
Ejaaz: So if someone wanted to get access to your product, you would query your API,
27:51
Ejaaz: and you could either offer them two options.
27:53
Ejaaz: Here's all the information in the product that you want for free,
27:56
Ejaaz: or you need to subscribe to my thing, which is kind of like,
27:58
Ejaaz: think of like a newsletter or a news media corporation in today's world.
28:02
Ejaaz: You've got to pay a subscription to get access to the Financial Times, for example.
28:05
Ejaaz: Now, you can go to the same app and website and just pay per use,
28:10
Ejaaz: which is the coolest part for me.
28:12
Ejaaz: Like, Josh, do you know the number of times that I've gone to read an article
28:15
Ejaaz: on tech or whatever, and I've just been paywalled?
28:18
Ejaaz: I would love to be able to pay like 50 cents or whatever the amount is one time
28:23
Ejaaz: and get access to it instantaneously a protocol like this allows you to do it
28:28
Ejaaz: and any human can get access to it but the second part is something that you
28:32
Ejaaz: mentioned josh ai agents
28:34
Ejaaz: I'm firmly of the belief that the future way of digital commerce is not going
28:39
Ejaaz: to be via humans or facilitated by humans.
28:41
Ejaaz: It's all going to be done digitally through agents.
28:44
Ejaaz: You need a standard like this, an open protocol that facilitates payments in
28:49
Ejaaz: a matter of microseconds to exist for that world to exist.
28:53
Ejaaz: So this is a major, major evolution.
28:56
Ejaaz: People probably won't be as excited about this until they start using the product
28:59
Ejaaz: themselves. They probably won't have any idea that it's happening in the back end.
29:03
Ejaaz: But as a former and current crypto nerd, this is super cool.
29:08
Josh: Yeah, well, here's why people should be excited about it because initially it sounds horrific.
29:11
Josh: I'm like, wait, I don't want to pay to read an article or to watch a video.
29:15
Josh: And that seems like a terrible, terrible, terrible user experience.
29:18
Josh: But the alternative has actually been much worse.
29:21
Josh: It is the advertising model that gets in the way.
29:23
Josh: It's the subscription model that costs $30 a month just to read a couple of articles.
29:27
Josh: And what this does is it kind of offloads a lot of the financial burden of a
29:33
Josh: company onto like an open source platform.
29:36
Josh: So if you don't want to, if you wanted to go read that article,
29:39
Josh: EJ, you wouldn't need to pay $20 a month to the New York Times.
29:42
Josh: You could just pay a couple cents for that specific article and get a chance
29:46
Josh: to get the content that you want and move on with it.
29:48
Josh: And I imagine that the net of this probably equals out to be lesser.
29:54
Josh: Than what it would be in the traditional sense of things. So that hurts a lot
29:58
Josh: of traditional companies where if you are making a killing off of $20 a month
30:01
Josh: for people to read five articles, this is probably going to be a little scary.
30:05
Josh: But for the rest of the open internet that wants to lean into this,
30:08
Josh: that wants to have the AI agents do things on our behalf, if I could give my
30:11
Josh: agent a wallet with $100 in it and say, here's your budget for the next couple
30:17
Josh: of months, go get me all of the information you need,
30:19
Josh: track down all the sources you need to aggregate that for me,
30:21
Josh: that is like a huge, huge unlock.
30:24
Josh: And most importantly, you just don't need an account or anything.
30:26
Josh: It just lowers the friction to use the internet. So this, by all means, great progress.
30:30
Josh: Big fan of X402 and just excited to see what comes from it. Normally,
30:33
Josh: when we get these protocols, the amount of innovation that happens as a second
30:38
Josh: order effect of it is just massive.
30:39
Josh: So this has a real chance at really shaping the way the internet looks and how
30:44
Josh: we engage with it over the next couple of years.
30:46
Ejaaz: Okay, going on to our final topic and arguably the most controversial on the docket this week.
30:54
Ejaaz: We want to put data centers out in space. This is something we mentioned last
30:59
Ejaaz: week where a company called StarCloud is planning on building an AI compute
31:04
Ejaaz: data center and launching it into space.
31:07
Ejaaz: Apparently, this is going to make it a hell of a lot more efficient.
31:10
Ejaaz: Apparently, beaming compute down onto Earth is way more environmentally friendly
31:15
Ejaaz: because it's out in space.
31:17
Ejaaz: There's nothing around it. It can just emanate heat, right?
31:20
Ejaaz: But Josh, I know you have a lot of strong opinions on this.
31:23
Ejaaz: I have this tweet lined up to give an update as to what exactly they're doing
31:27
Ejaaz: and how they're achieving it.
31:28
Ejaaz: And I have to say, and by the way, this tweet thread was produced by an AI model,
31:33
Ejaaz: Perplexity, so shout out to them.
31:35
Ejaaz: They give a very compelling case, Josh. There's a few things.
31:38
Ejaaz: They talk about the fact that these satellites are probably,
31:41
Ejaaz: this hardware is probably not going to be as heavy as people are claiming it to be.
31:45
Ejaaz: So it actually makes it kind of efficient or cost worthy enough to launch some
31:50
Ejaaz: of these things in space.
31:52
Ejaaz: It's going to be the first ever AI model that is going to be launched and trained
31:56
Ejaaz: and fine-tuned out in space. That's another one.
31:59
Ejaaz: And also they talk about the biggest criticism that they've faced,
32:02
Ejaaz: which is around how these data centers in space are gonna emanate heat.
32:07
Ejaaz: There is no way to do that in space. Space is a vacuum.
32:12
Ejaaz: And they talk about this neat little convection layer that they're gonna like
32:16
Ejaaz: kind of film the entire or cover the entire data center with.
32:21
Ejaaz: I don't know if that's true. That's kind of like the theory around it.
32:24
Ejaaz: But most importantly, the economics are pretty insane.
32:27
Ejaaz: On Earth, they claim that it's gonna cost you around $167 million to set up
32:32
Ejaaz: the same kind of thing that they're trying to do in space.
32:35
Ejaaz: But in orbit, it's only going to cost them $8.2 million.
32:39
Ejaaz: Josh, to any kind of businessman looking at these figures, that is an investable
32:43
Ejaaz: worthy company. Tell me why I'm wrong.
32:46
Josh: This is outrageous. I can't believe we're talking about this.
32:49
Josh: See, this is proof that you can't really trust AI agents, really.
32:54
Josh: So the cost structure here is pretty outrageous. So like $167 million versus $8.2 million.
32:59
Josh: Clearly, there's something wrong there because the cost per kilogram to orbit
33:02
Josh: at the moment is $2,000 to $10,000 per kilogram.
33:08
Josh: Five or 40 megawatts is an astronomical amount of infrastructure required,
33:14
Josh: particularly because in order to cool a five megawatt data center in space,
33:21
Josh: because there is no atmosphere,
33:22
Josh: it requires 16 square kilometers of radiators and heat dissipation.
33:28
Josh: And if you want a 40 megawatt data center up there, that is a very gigantic
33:35
Josh: piece of space junk that is not really proven in technology.
33:40
Josh: There's a lot of things like shielding that is really difficult because there
33:43
Josh: is a lot more solar flare energy. If a single bit gets flipped in any of these
33:46
Josh: training runs, it's very difficult to recover.
33:48
Josh: You have to do the whole thing over again. So radiation is a problem.
33:52
Josh: The cooling is a problem. The cost, I think they must be assuming that the cost
33:56
Josh: per kilogram to orbit is like starship final form when it gets down to like
34:01
Josh: 10 to 30 kilograms per orbit.
34:02
Josh: The reality is it costs several orders of magnitude more to do that today.
34:06
Josh: And what this isn't taking into account either is the decreasing cost of electricity over time.
34:12
Josh: Since the early 1900s industrial era in the United States, we have not really
34:17
Josh: needed to produce that much more energy.
34:19
Josh: Like our energy has been kind of gradual, but it's never gone up the exponential
34:22
Josh: curve in the same way technology did.
34:23
Josh: We are very much at the exponential curve right now where that is changing.
34:27
Josh: And I think to assume that energy costs are going to remain this high is fairly
34:31
Josh: outrageous because clearly the goal is to get energy costs as close to zero as possible.
34:36
Josh: There is no change to this trend. And this whole thing kind of seems outrageous.
34:40
Josh: Now, granted, I love the idea that they are experimenting, that they want to
34:44
Josh: try to make these fun outer space orbits. But it just seems kind of hand-wavy.
34:48
Josh: One of their bragging points is, oh, this is 100 times more compute power than
34:52
Josh: the largest GPU that's ever been in space.
34:54
Josh: And that's because we don't ship GPUs to space. Like this is just the first
34:58
Josh: big one we've ever shipped. And it's one of them.
35:00
Josh: And real data centers now need millions of these coherently operating together.
35:04
Josh: So to me, it's like, it's a fun science experiment, but I wouldn't take this
35:08
Josh: very seriously because there are lots of long time scales and technology breakthroughs
35:11
Josh: required in order to make that number realistic.
35:16
Ejaaz: All right. Okay. Listen, Josh, all very valid points, but I have one more comment to make.
35:22
Ejaaz: You criticize the fact that this company is suggesting to build out like, what is it?
35:27
Ejaaz: A 14 kilometer satellite or data infrastructure to actually build it.
35:31
Josh: That's just the physics of what it takes to cool it.
35:33
Ejaaz: Right. Okay. To cool this, right? Isn't the whole point of space that there's a lot of space?
35:39
Ejaaz: Why can't they just launch whatever the hell they want? We have infinite space
35:42
Ejaaz: in this thing? Why can't they just do that?
35:44
Josh: It turns out like space isn't super infinite and particularly on low earth orbit.
35:48
Josh: I imagine there's going to be a large battle for low earth orbit.
35:51
Josh: Why low earth orbit? Well, it is low earth. It's close to earth.
35:55
Josh: And because of a thing called latency, we want these things,
35:59
Josh: we want to be able to communicate with these things faster.
36:01
Josh: And unfortunately, the speed of light does have a cap and things cannot,
36:04
Josh: we haven't figured out how to make things travel any faster than the speed of light.
36:07
Josh: So if your data center is further away, it takes much longer to communicate
36:11
Josh: back with Earth. And that becomes a problem.
36:13
Josh: What I prefer is instead of putting these like giant 16 square kilometer like
36:19
Josh: pieces of things into this precious space, space in space, well,
36:25
Josh: wouldn't it be cool to just throw like a lot more of these Starlink V3 satellites
36:29
Josh: down there so they can beam direct to cell service back to us?
36:32
Josh: I think there's a lot more practical utility of that low Earth orbit versus
36:35
Josh: throwing these like gigantic things that require a lot of space.
36:39
Josh: You just, if you saw, we talked about the SpaceX launch last week.
36:42
Josh: Those satellites are very small. They don't take up a lot of space,
36:45
Josh: but they are very high impact.
36:46
Josh: These on the other hand are gigantic. And yeah, it doesn't strike me as very
36:52
Josh: exciting as a lucrative way of using our limited real estate up there.
36:57
Ejaaz: Yeah, I appreciate the boldness of the idea. This team did come out of Y Combinator,
37:02
Ejaaz: so I'm presuming that it is quite fitting to the batch and cohort that they're graduating from.
37:07
Ejaaz: Y Combinator always launches pretty loony ideas, which end up sometimes becoming world-changing.
37:13
Ejaaz: But we are yet to see whether StarCloud will prove their thesis of launching
37:18
Ejaaz: Jensen Huang's NVIDIA GPUs into space.
37:21
Ejaaz: Speaking of Jensen Huang, $5 trillion market cap. It is an absolute crazy week.
37:28
Ejaaz: It is earnings week this week, guys. Stock market is at all-time highs.
37:32
Ejaaz: We mentioned Microsoft and Apple. Apple, super randomly, coming out of nowhere.
37:36
Ejaaz: I thought AI had beholden them. I thought they had completely skipped the entire phase.
37:40
Ejaaz: Somehow they surpassed a $4 trillion market cap. Maybe they're underpriced.
37:43
Ejaaz: We'll talk about more of that potentially next week.
37:46
Ejaaz: But that is all that we have on our docket today.
37:49
Ejaaz: I just want to say from me personally, thank you so much for the support all
37:53
Ejaaz: of you guys have shown us. So far, we had like one of our most viewed videos
37:57
Ejaaz: ever last week. And we're having an amazing week this time.
38:00
Ejaaz: All the comments and feedback that you guys are giving, all the likes,
38:02
Ejaaz: all the subscriptions, all the ratings that you're giving us on RSS,
38:06
Ejaaz: Spotify, Apple Music, wherever you're listening to this is hugely, hugely appreciated.
38:10
Ejaaz: If you haven't done any of that already, that's okay.
38:13
Ejaaz: Please do it now. It takes a few seconds.
38:16
Ejaaz: And Josh, do you have anything else that you want to share?
38:18
Josh: Oh, I have to remind you, we've had a big week in the world of YouTube,
38:21
Josh: which is awesome. There's been a lot of new people joining and subscribing.
38:24
Josh: Just a reminder that there is an audio version of this available on Spotify
38:27
Josh: where you can also watch the video as well as just audio only on RSS feeds,
38:31
Josh: basically anywhere where you get podcasts. Spotify is my favorite way to watch the show.
38:35
Josh: Sometimes I'll rewatch our episodes and I'll go on my Spotify and it's nice
38:38
Josh: because it's a pocket player that does audio, but it also has video built right in.
38:43
Josh: So if I'm on the treadmill and I want to watch, I could just watch it
38:45
Josh: on my phone or if I just like lock my phone because unless you
38:48
Josh: have youtube premium you can't lock your phone and still listen it still
38:51
Josh: plays in the background so spotify is my preferred way of watching but yeah just thank
38:54
Josh: you all so much for the support and for sharing with your friends and just being
38:57
Josh: here for every episode as we travel along the journey and reach all-time highs
39:01
Josh: across the board particularly in stocks this week so big earnings season coming
39:05
Josh: up i'm sure we'll get into some of that next week but that's been another great
39:08
Josh: week on limitless and thank you all for watching as always we'll see you guys next week.